README20.TXT Copyright (C) 2001 by Robert S. Larkin This is a summary of the changes in Version 2.0 of the DSP-10 Software, UHFA.EXE. This document needs to be used along with the readme16.txt" file that went with 1.6x or 1.9 versions of the program. From Bob Larkin, W7PUA - Preliminary Issue of 24 Jan 2001, Updated to 2 Feb 2001. THANKS - There are many people that have contributed to the DSP-10 project during the period of this "Weak-Signal Revision." I can't hope to acknowledge all of these generous individuals, but also, I can't ignore the special help in beta testing and document reading that has come from KD7TS, W7SLB, W7SZ and W7LHL. Important Note: Two requirements on the software versions are a- The Version 2.0 of the UHF3.EXE must be used in the DSP. b- Loading UHF3.EXE now requires either the Analog Devices Windows loader or the new version 1.03 of EZFAST from Dwight Elvey. INSTALLATION - This is DOS, known as "life before Wizards!" Therefore, installation consists of making sure that the proper versions of files are in the directory designated for DSP-10 files. This directory may need to be on a hard drive if you are using EZFAST to load the DSP, because of problems with slow transfer rates of some floppy systems. The following files need to be replaced or added to install version 2 into a DSP-10 running an earlier version: UHF3.EXE The DSP program UHFA.EXE The PC program UHFA_43A.RND Random number table (not in earlier versions) EZFAST.COM If you are using EZFAST, it needs to be V1.03 The configuration file UHFA.CFG will be updated to include the new configuration variables, and their default values, the first time that UHFA.EXE is run. This will pick up any customized settings that applied when you were running an earlier version. If Version 2 is your first program for the DSP-10 you should review the installation notes in README16.TXT, as well as making sure that the above files are in place. To check the versions of UHF3.EXE and UHFA.EXE while the DSP-10 is running, use the Scrl-F3 command to bring up bottom line type 2. This has mostly a "hex dump" of status bytes. The UHF3.EXE DSP program has its version in the first hex value on the left. It should be "20" (still in hex). The UHFA.EXE PC program version is shown on the right side of the line and should be "Ver2.00." For general information, the UHF20.ZIP program unzips to the following files: README16.TXT Text file with lots of info. This is still mostly valid and until a single unified document is created, this remains the base document. README20.TXT Text file that explains the differences from Ver 1.6. Use it with the above README16 for the whole story. This file! UHF3_20.EXE The DSP program - NOT a PC executable, see the readme16.txt. UHFA_20.EXE The PC DOS executable control program. GNU_GPL.TXT The software license. Required in directory. EGAVGA.BGI Graphics driver required by UHFA.EXE. Required in directory. UHFA_43A.RND Random number table used with several of the weak-signal modes. ********************************************************************* KNOWN BUGS (still send any that you may find to W7PUA boblark@proaxis.com). 1-If you are using the EME-2 or LTI modes, followed by Scrl-Alt-F4 for shut-down, you may come up in "16-bit" data when you go to other modes. This is revealed by the right half of the spectral display being blank. 16-bit is correct for EME-2 and LTI, but is only a seldom used option in the other modes, such as CW. Restore 8-bit data by Scrl-F2. 2-LTI transmit does not update the Moon data or the Alt-A Box. It is not really finished yet, but is basically working well and quite useful! To get moon Az-El during transmit, hit the Home key to drop out of transmit for a second or so. Go back to transmit. If you were in a pound-sign loop, it will return to that loop. LTI receive updates everything, as it should. 3-Data does not enter into the CW buffer after doing a loop in LTI (or using CW and then going through the other modes). LTI and CW share common buffers, but the pound-sign loops don't get written to the screen when you change modes. The work-around is to clear the CW/LTI transmit buffer with Shift-Del when you change to either CW or LTI. 4-In EME2 mode, if the bottom diagnostics line (Scrl-F3) is not "Moon:", the moon data will flash on-and-off. Changing the Scrl-F3 will not bring up any other diagnostic mode. To stop the flashing and select the "Moon:" option, either change to a different mode, or watch the "E" below the frequency box. When it is white (default colors) it is properly selected. 5-During startup you may see a message flash by about "Retry -1 to Download DSP Pointer Table." This is normal, and a single retry is not a problem. This will be dealt with soon! ********************************************************************* A FEW NOTES ABOUT FREQUENCIES Several conventions have been used in the DSP-10 to make sure that both ends of a contact can be on the proper frequency! The rules for making this possible and easily done are in the following paragraphs. The description given applies to the display style with a waterfall spectral display (Alt-Y, y). The alternate "slide-rule" display shows a more restricted set of information. The frequency box on the left side displays the transmit base frequency. For CW, this is the actual transmit frequency. For SSB, this is the frequency that a carrier would be on if it existed. For FSK modes like LHL7, PUA43, EME2 and LTI, it is the base frequency to which the FSK value is added. For these latter modes the FSK value is displayed just above the Transmit Frequency Box as "FSK=xxx." If the FSK mode involves randomization, the FSK value will change to indicate the current value. If the path does not involve either EME Doppler correction or randomization, the receive frequency is displayed as follows. For SSB, the displayed frequency in the Transmit Frequency Box is still the carrier frequency. For CW, the Transmit frequency Box is the frequency that will produce a tone to the ear that is equal to "CW Offset" (displayed next to "CW" in the Mode line). For FSK modes as listed above, the base receive frequency is displayed in the Transmit Frequency Box and the tones are heard at their FSK values, since the receiver is in "USB". If the receiver frequency has been shifted by RIT, or the path involves EME Doppler corrections, or the mode involves randomization, such as is available in PUA43, EME2 and LTI, the receiver frequency is displayed just below the Transmit Frequency Box, and an "R" (for Random) or "E" (for EME Doppler) is shown to indicate the reason for the receive frequency shift. If only RIT is involved, the receive frequency will be displayed, but no "R" or "E" will be shown since RIT has its own display line. Note that the Transmit Frequency Box never changes between transmit and receive. If two stations are trying to communicate, they should always have the same frequency displayed in that box. Also note that the only case where the Transmit Frequency Box changes with mode is CW. This is a quirk of making shifts between CW and USB painless. The way it is arranged, a station tuned in on USB will be properly tuned when going to CW, if the CW frequency is offset up in frequency by a proper offset, usually 600 or 800 Hz. This is the scheme used by many, but not all, commercial transceivers. Two stations using this scheme can receive CW or USB properly regardless of whether they are in either CW or USB modes. The one situation to avoid is setting the frequency of operation while in the CW mode and then changing to another mode. This will cause a sometimes-puzzling shift by CW Offset. If you do set the frequency in the CW mode, be aware that it is set high by the CW Offset. ********************************************************************* Now to the changes that have occurred between V1.6 and 2.0: MISCELLANEOUS CHANGES Repeater Offsets (v1.67) - There have always been 21 VFO's. They now have offsets, for FM only. It goes in the UHFA.CFG file with any text editor and is in the position called tvtr.control. The offset is in kHz and is an integer, meaning no decimal point. If you receive your repeater on 146.94 and transmit on 146.34, the offset is -600 kHz. They can be any multiple of 5 kHz that you can keep in the range of the transceiver. A tvtr line for your 34/94 repeater could be: tvtr 20 |34/94 Rptr| 0 -1 0.0 1 | | 0.000 -600 0 146940000.0 3 where the number 20 puts it in the position for CTRL-ALT-5. Be sure to have the 3 on the end for FM. The 21 VFO's are all available for any use. I set mine up for a whole bunch of UHF/Microwave transverters. This is arbitrary. If you need a bunch of repeaters, use more of the 21. Remember, these are VFO's, in that if you change the frequency and change to another VFO, or shut down, it will be where you left it when you come back. Memories don't do that. They come back to their preset frequency. The program does not have "memories," meaning frequencies that always come back to the same value when they are selected.. Screen Colors (v1.67) - Here is a tool to allow changing of the waterfall, and screen color; they are locked together in the restricted palette of VGA. It has always been possible to assign a palette of 16 colors from the available 64. This allows you to change the waterfall colors. But, now you can assign all the other display colors from this list of 16 to keep from having a very poor result. First pick your palette to make the waterfall to your liking. Then look in the .CFG file and there are 15 lines, like "c_text_bold 15" that assigns the color 15 to what is currently white text. These can all be changed in the .CFG file to suit your desires. Just be sure the number after the item is in the range 0 to 15. The default (no UHFA.CFG file) colors are left as they were to keep a base point. But, it is all there to be changed. Whatever you put in the .CFG file is what you get. By the way, you can delete a portion of the .CFG file, such as the palette, and get the default for that item only. The rest will remain customized, as they were. (v1.67) AGC does not cause a jump of 48 dB between on and off--the AFGAIN changes by 48 dB instead. (v2.0) The receive gain for SSB/CW, etc., has been lowered by 10 dB. Wearers of headphones were getting too much audio. If you still have too much volume in the headphones, it will probably be necessary to add a resistor in series with the 'phones. (v1.67) There is a separate AFGAIN in software for FM, as it never comes out right otherwise. This is all saved at shutdown as usual. (v1.67) CW speeds are faster at the high end. It now goes from 5 to 75 WPM in uneven steps: 5, 6, 7, ...24, 25, 27, 30, 35, 40, 50, 75. (v1.9) RIT can now be entered in 1 Hz steps: Use Ctrl-Alt R and Ctrl-Alt T (or lower case r and t). (v1.9) The 5 kHz Birdie is gone. Many thanks to Hannes, SM6PGP who identified the source of the 5 kHz spur as being the reference output (pin 3) of the MC145170 PLL. I had accidentally left it active, as a divide by 16, and the resulting 625 KHz signal was somehow mixing and producing a signal at 19.665 MHz (our I-F). The 625 signal is not used, and it has now been programmed off. The keys used for the spectral display dB/div and offsets are now Alt 1, Alt 2 and Alt 3, instead of the shifted versions of these keys. This was suggested by KD7TS when he found he was using the controls a lot. It seems to be a better arrangement. CW Offsets can now go up to 9990 Hz. A wider IF bandwidth is now available. It was 2.8 kHz before and caused the spectral display to be of no value from 2.8 to 4.8 kHz. The new one is 4 kHz and the spectral display is useful to almost 4.5 kHz. It causes a slightly different sound to the signals and noise. The drawback is that strong signals, in the 4.8 to 6 kHz range appear as alias signals from 4.8 down to 3.6 kHz. For weak signal work this is not usually a problem. But if it is found to be, it can be toggled in and out with ALT $. Receive frequency correction for EME Doppler is now available for CW and SSB (as well as the weak-signal modes LHL-7 PUA43, EME-2 and LTI that are detailed below). The correction can be toggled on and off by ALT L, or ALT l. It is in effect only when an EME path is being displayed on the bottom line. An 'E' below the frequency readout means the feature is enabled. If a receiver frequency is displayed to the left of the 'E", then the correction is being used. There are situations when the 'E' will show in beige, but the frequency is not displayed and corrections are not being made. An example of this is the FM mode. When the Doppler correction is being made, CW and SSB stations will be received as though no Doppler shift existed. For example, if the other station is transmitting USB on 144.085 000 MHz and the Doppler shift is 234 Hz, the Transmit Frequency box can be set to 144.085000 and the station will be tuned in. The receive frequency will show as "144.085 234 E." Note that the PC clock needs to be set closely to have the correct EME-Doppler correction. At 2-meters an error of 1 minute can create a 1 Hz error. At 1296 MHz, this will be 9 times greater. Software clock corrections can be applied to correct the clock as is explained below for the "PUA43 Mode." But, in general, it is best if the PC clock is maintained within a few seconds of UT/Local Time. The S-meter bars are now tied directly to the dBm number under them. S1 to S9 are exactly 6 dB each and the bars over S9 are 10 dB each. In addition the S-meter numbers and bar graph now work for signals that are off-screen on the upper spectral display. To make this happen, there is a change in the mode of operation at about -100 dBm input. Below that level the signal strength comes from the spectral data, indicating the peak on the upper spectral display that is marked with a red circle. At higher levels, an RMS (power) calculation in the signal band (from the DSP) is used. The transition is indicated by a 'P' showing to the right of the strength number, meaning that a power measurement is being used. Including the RF gain, this gives an input range from about -150 to -25 dBm that is displayed correctly. If one is using the Signal Level display to make comparative measurements, it is desirable to avoid switching between the two modes of operation. The time marks on the waterfall have been changed to show mm:ss instead of hhmm. I like this better, since the hour and day can almost always come from the lower left corner. This allows you to identify a time in the waterfall to the second. If you want to go back to hhmm, it is still available by changing the variable show_secs in the .CFG file to 0. Three alternate keyboard arrangements are now available from the .CFG file: kbd_alt1 swaps the 10 and 100 Hz keys kbd_alt2 moves Alt-W, Alt-I and Alt-Y to Scrl-W, Scrl-I and Scrl-Y kbd_alt3 moves Alt-F7 and Alt-F8 to Scrl-F7 and Scrl-F8 (RF gain) MODE_MASK - This is a .CFG file entered quantity that allows the operator to ignore transmission modes that are not of interest. It is the sum of the values for each mode desired. The values are listed below, and also in a comment in the .CFG file, just above "mode_mask x." As an example, if one wanted to only have CW=1 and USB=2 available, they would set mode_mask=3. In order to not allow a modeless radio, the CW mode is always available. CW mode_mask= 1 USB mode_mask= 2 LSB mode_mask= 4 FM mode_mask= 8 LHL7 mode_mask= 16 PUA43 mode_mask= 32 EME2 mode_mask= 64 LTI mode_mask=128 The 'Help' screens are not yet keyed into this change, but the Alt-Y screen is. If you ever wondered what key you hit, it now shows in the lower left, if the keystroke had any effect. It shows the function of the key, abbreviated to 6 characters. This lets you find out how things work by hitting keys experimentally! Out-of-Lock detection is now fixed. Run your frequency up and down and an error message should show the lock limits. Mine was 143.35 to 148.8. Fast PTT is implemented in the PC and DSP software. This requires: a-Change .CFG file item hardware_ptt to 1 (from 0) b-RS232 connector on DSP-10 (J201), remove wire at p7 and tape c-RS232 wire at p8 moves to p9 d-RS232 jumper p7 to p8 e-On the EZ-Kit PC board, cut the connection on U5 between pins 9 and 10 (and Exacto does it) f-In EZ-Kit box, run a wire from U5-pin 10 to C208 What this does is to take the PTT logic level at C208, run it through the RS232 level converter (EZ-Kit U5) to the Ring Indicator line on the RS232 plug (pin 9). If hardware_ptt is a logic "one," this line will indicate PTT to the PC program. The binaural audio Alt E delay is now only adjustable from the .CFG file. The default is 1024 set by a value of 9 in CFG. Values of 1 to 10 are allowable. Each delay unit is 1/9600 sec, or 104 microseconds, Several improvements will be noted in going to version 2.0. An example is the reduction of "pops" when changing modes, particularly in coming out of FM. Also the "pops" associated with T/R changes have been greatly reduced. All keys are now available in "lower-case" form. It used to be that the keys that involved the scroll lock also needed a shift. Version 2 still supports an audio processor, but only for reception of USB (interestingly shown as LSB on the screen). The reverse sideband can be produced by setting the "Offset Frequency" to a negative value. Transmit functions are no longer supported and are locked out. The frequency range for reception is 0 to about 21 kHz. ********************************************************************* KEYBOARD COMMANDS - Notes about the keyboard commands are intermingled in the descriptions of the new or changed functions. Here is a summary, in one place, of the changes, relative to earlier versions: Ctrl-Alt-R (or r) RIT down 1 Hertz. Ctrl-Alt-T (or t) RIT up 1 Hertz. Alt-$ (Alt-Shift-4) Toggle i-f bandwidth 2800/4000 Hz. Alt-J (or j) Sequence through the SpecAnl widths. Alt-L (or l) Toggle EME Doppler correction on/off. Alt-1 Change dB/div (was Alt-!) Alt-2 Spectral display offset down (was Alt-@). Alt-3 Spectral display offset up (was Alt-#). Alt-A Data box toggle. Alt-B Modal parameter box on (some modes). Scrl-F4 Modal parameter box for filter design. Alt-% Display Width - Not available, now automatic. '{' In CW or LTI holds the key down for 2 seconds. '}' In CW or LTI holds key up for 2 seconds Ctrl-W In long-term modes, clears old data Ctrl-E In long-term modes, straightens noise display. ********************************************************************* DSP COMMANDS - These have undergone major revision. A summary is forthcoming for those that want to use the existing DSP program with alternate control programs. ********************************************************************* ERROR/WARNING MESSAGES in the lower left corner of the screen have been renumbered and given short descriptions. The following items are currently implemented: E1-Serial Data Lost. Spectral data transmitted by the DSP arrived with an error in the check sum and is not being included in the spectral display updates. This is not unusual during mode changes when a data set may have been truncated. E2-Invalid Status, Receive. About 20 status bytes are sent to the PC from the DSP during reception. Validity checks show if these have not been received correctly. This can again occur wit mode changes without indicating a real problem. Status bytes are discarded if they are not valid. E3-Illegal DSP Command. If the DSP receives a command that is out of the vocabulary, this is reported back to the PC as a status bit. These commands are ignored by DSP. E4-Invalid Status Transmit. Status is sent to the PC during transmit, No spectral data is sent. At the time of issue for V2.0 a bug remained that causes some invalid E4 messages. It is apparently associated with the presence of an Escape character in the status data. E10-Excessive Auto-Display Correction. Over 10 dB of correction is needed to keep the display centered. This indicates a major change in the RF gain of the system. If this gain shift is intentional, the Auto-Display (Alt-C) should be turned off and back on to reinitialize the operation. E11-Flattner Curve Fit. Indicates numerical error in the polynomial curve fit used for the "yellow" trace of EME-2 or LTI. Yellow trace will be invalid. Please report any occurance of this error to W7PUA. W12-No Clock Set. It is a warning that the software clock has not been set. For LHL7 and PUA43 modes it would be unusual to not need a clock set. W13-Illegal during Audio Input. A command that was not proper for the Audio Processor has occurred. W14-No Tvtr. To use a transverter setting (Ctrl-Alt-?), the UHFA.CFG configuration file must have the full set of information entered. Transverter settings without this data will not occur. E15-No UT. Doppler corrections require a knowledge of Universal Time (UT). To make this work, there must be a DOS environmental variable TZ, that is the offset from the local time zone to UTC (GMT). See the V1.6 read-me file for more information. E16-3rd LO Out of Range. In order to not have transient problems from re-locking the 5 kHz synthesizer when in transmit, only the 3rd, DSP LO can be changed during transmit. If this is shifted too far, it will fall outside the D/A and Crystal filter limits. The solution is to go into receive before making major shifts in the transmit frequency. E20-Illegal number of bins in LHL-7. This is locked out normally, but can occur at startup. Use Alt-F3 and Alt-F4 to select a valid combination. E21-PUA43 out of memory. Dynamic memory allocation has failed for the PUA43 mode and it will not be available. Check the users of DOS 640K memory as there should normally be adequate amounts of memory. E22-PUA43 Illegal Time Tag. PUA43 uses a software triggered FFT to gather the spectral data. Along with this trigger goes a time tag that is in turn returned to the PC with the spectral data. This tag is checked to align the data with the proper symbol in reception. Data is discarded if a time tag error should occur. E23-UT needed for LTI randomization. See E15. E24-UT needed for PUA43 randomization. See E15. W25-Millisecond clock not available. Weak signal modes, LHL7,PUA43, EME2, LTI, all require a software millisecond clock. The Borland routines are not finding proper BIOS support in the PC when W25 occurs. E30-PLL Un-Lock. There are 2 PLL's in the DSP-10 for the 126 and 19.68 MHz synthesizers. One of these is not locking properly, often because the tuning coil/capacitor is not set properly, or the DSP-10 has been tuned past the range of the PLL. E40-Graphics Error. Make sure that the system has VGA 640x480 16 color as an option. Check the details of the graphic system that can be seen at startup time by setting "startup_detail 1" in the UHFA.CFG file. E41-Memory error, Curve straightener for EME2 and LTI. See E21. E42-CFG File Error. This covers a broad class of file errors. Check the availability of the drive/directory that held UHFA.EXE and will be the destination for UHFA.CFG. If there is a serious disk error, there may also be an text error message over the upper left corner of the screen. The absence of UHFA.CFG at startup is not an error and is the way to return to all the default configuration values. E43-LTD File Error. This file saves long-term integration data. See E42. E44-Other File Error. See E42. W45-Ctrl C Trap. Control C is the default command for aborting program operation. This is a bit harsh for a program such as the DSP-10 and so it has been trapped and only generates this warning. Use Scrl-Alt-F4 to stop the program in an orderly way. E46-Math Error. A floating point error has occurred. Every attempt is made inside the PC program to test values that can be a problem that generates this error. If one was missed, it will show up here. Please report these along with the conditions that seemed to create it to W7PUA. E47-DAT File Error. This is the file that collects large amounts of detailed spectral data. See E42. E48-Dynamic Memory Allocation for Filter Design. See E21. E49-Math Error in Filter Design. Most of the common errors in filter parameter specification are trapped at the "Alt-B" Box and will produce messages there. If E49 is encountered, check that the parameters are reasonable, as it would appear that the filter was not possible to design but was not caught at the B-Box time. ********************************************************************* SPECTRAL DISPLAY There is a new "windowing function" called Tukey25. I have wanted to find something between "None" and "Hamming". This is it. Not as wide a noise bandwidth as Hamming, but not nearly as prone to spreading on strong signals as None. Play with it on a strong signal and tune around while using None and you'll see the problem. Tune in 1 Hz steps (CTRL-ALT-F9 and F10) to really see it. In the summary that follows, the bandwidth corresponds to a SpecAnl width of 4800 Hz and they should be divided by 2 or 4 for the 2400 or 1200 Hz widths. Window fcn Bandwidth Loss of S/N Sidelobe falloff ---------- --------- ----------- ----------------- None 9.4 Hz 0 dB -13 dB and slow Tukey25 10.3 0.4 dB -14 dB and fast Hamming 12.8 1.3 dB -43 dB and slow BH-92 18.8 3.0 dB -92 dB or lower Along with the 4800 Hz spectral display width that has always been available 1200 and 2400 Hz have been added along with a control for selecting the width (Alt-J or Alt-j). These are all the same 1024 point length and the difference is the sample rate. This gives spectral-bin bandwidths that are 1/2 or 1/4 of the previous v1.67 or 1.9 software. The "bin bandwidth" of the spectral display is in the upper right corner, below the dB offset. It depends on the spectral width above and on the windowing function being used. It is a measure of the coherent sensitivity of the spectral display. As the bandwidth of the spectral bins is varied, the noise bandwidth changes and, if the signal levels are adjusted to remain constant, the noise will drop as the bandwidth decreases. This is the way to operate the display if the primary interest is the signal. But if the interest is the noise, the gain should go up as the bandwidth decreases. To allow either arrangement, and any in-between, there is a .CFG variable called "knoise." If knoise is set to 0.0 the noise power will be kept constant. If knoise is set to 1.0 the signal power will be constant. Because of the noise statistics it is not always best to keep the noise power exactly constant and for this knoise can even take slightly negative values. ********************************************************************* FILTERS AUDIO FILTERS and FILTER_MASK - There are now 8 audio filters, set by F4 and toggled on and off by F3. The eight audio filters are maskable just like the mode_mask. The default filters are: #1 MTCH12 W8MQY matched 12WPM filter_mask= 1 #2 6CW200 200 Hz at 600 Hz filter_mask= 2 #3 6CW300 300 Hz at 600 Hz filter_mask= 4 #4 6CW450 450 Hz at 600 Hz filter_mask= 8 #5 7CW600 600 Hz at 700 Hz filter_mask= 16 #6 SSB-N 250 to 2500 Hz filter_mask= 32 #7 SSB-M 200 to 2900 Hz filter_mask= 64 #8 8CW300 300 Hz at 800 Hz filter_mask=128 The W8MQY matched filter is kind of fun. It is a sin(f)/f shape which is the ideal shape to maximize the S/N on a CW station. Filter #1 is set for dots at 12WPM. You must use the spectral display to tune the station exactly to 600 Hz. You can't tune by ear because it has responses on the side and the response falls off slowly. But it must be peaked, because the nose of the response is sharp and the sides must hit the side responses quite closely. When a weak signal is tuned in, it hops right out of the noise! If it is way off in code speed it won't work right either. I-F FILTERING - The architecture of the DSP-10 applies DSP I-F filtering before either the FFT spectral processing or the Hilbert-transform SSB summation. This filter was originally chosen to be very conservative in removing alias signals. The results of this could be seen on the spectral display where the frequency response rolled off rapidly above 2.8 KHz. In version 2 there is the option of a wider filter that holds the response up to above 4 KHz. This is toggled in and out by Alt-$ (or Alt-Shift-4, if you like). The drawback is that a strong signal in the spectral display frequency range of 4.8 to 5.4 will be seen as tuning the wrong direction from 4.8 to 4.2 kHz. This is normally not a problem and the wide filter should generally be used. The I-F filter width is displayed as a 'N' or a 'W' following "Filter" in the left-hand column. As an aside, for the technically curious, the I-F filtering is applied as a pair of low-pass filters on the I and Q outputs, following the DSP-implemented third conversion. This is totally equivalent to a band-pass filter applied before the third conversion. Audio filtering is applied after the Hilbert-transform SSB summation. Within the limitations of the Hilbert-transform accuracy, this is again equivalent to I-F filtering. The audio filtering has the advantage of needing only a single filter. DESIGN-A-FILTER - There are 8 audio filters, 1 to 8. The first 7 are fixed (for now) but #8 can be customized. Use Scrl-F4 to bring up the design box. Pick your parameters and see the results. Close the box and the filter is downloaded to the DSP. If you are in filter 8 you get to hear it being changed! The only parameter in the Filter Design Box that is not obvious is the "1st Sidelobe, dB" line. Out of the filter pass-band the response tends to be a series of hills, or side-lobes, that generally diminish as the frequency gets farther from the passband. The first of these side-lobes outside of the pass-band can be raised or lowered in the design. The advantage of not pushing this side-lobe unnecessarily low is a more rapid transition from the pass-band to the stop-band. Generally the parameter should be between 40 and 70 dB, but value between 0 and 99 dB are allowed. In addition, this newly designed filter uses the 13 point graphic equalizer that is near the bottom of the .CFG file under the item called "aeqrdb." The frequencies are printed as a comment in the file just below the values All values are relative dB and can be anything up to +/-100 dB. Thus you can use the equalizer to add a notch band, or whatever. The plot in the Alt-B box shows the effect of both the filter design and the equalizer. Play with it! The design parameters are saved in the .CFG file, but it is NOT automatically loaded when you start up. As with all modal dialog boxes, the Filter Design Box is closed with an "Enter." As with other portions of the UHFA.EXE program, the filter design will operate without a math co-processor. However, you will find the slowness to be very annoying! kfilt GAIN VARIABLE - As you change filters there is quite a bit going on, some of which is the gain of the filter. Narrow-band filters can have a lot of gain to sine waves at the center freq. The matched filter has a gain of 64 but the wide SSB-N filter has a gain of 1. If the filter gain is all removed, the noise can become so weak that it is lost in the least-significant bits of the DSP. When this happens, the noise sounds distorted and weak signals have harmonic distortion, all of which is unpleasant to listen to. If you have outside gain (preamp, converter) raising the receiver noise level, removing the filter gain is not a problem and the noise/weak signals sound fine. But, if you don't remove the gain, strong signals change level as you change filters, which can be annoying, and tend to cause overload problems. So, to make this work, there is a configuration file variable called kfilt that can take values from 0.0 to 1.0. If it is 1.0 the gain to sine waves is constant. If it is 0.0 the gain varies according to the FIR filter. Starters: if you have no distortion on noise and weak signals use kfilt=1.0. If there is distortion start reducing to kfilt=0.5. It is unlikely that you would need 0.0, but... This is only available in the .CFG file. HOOKS - It doesn't affect the operation of the radio much, but several important "hooks" have been added to allow the changing of the audio and i-f filters, without recompiling the PC computer program. This always involves changing the coefficients of a FIR filter somewhere in the DSP. There are now tables, and pointers to these tables, that are made available from the .EXE file for the DSP and from the .CFG file for the PC program. This will allow changing of the built-in filters as an off-line activity. This can also make "graphic" equalizers available for any filter as well as for the transmitter audio. There are 7 "fildat" entries that you will see in the .CFG file. These are a series of 9 parameters that specify filters 1 to 7. Changing these parameters will not change the filter! These are informational inputs to the PC program so that one can change the FIR filters in the DSP and then get the display to agree with the the FIR. As of now, there are only 2 parameters used, the |name| that appears just after the filter number and gain which is the last parameter. The gain is a power of two that is the inherent voltage gain of the FIR filter. Eventually, the filter response will be shown on the spectral display as a tuning aid. This provides the information to make such things work. ********************************************************************* LMS PROCESSING The Denoiser was reworked based on a simulation I did a while back. I think it works quite a bit better, but it still needs different settings based on the signal levels. Both the Denoise and Autonotch should be used with the AGC on to both minimize this problem and to get the gain up high enough for best operation. The settings for the Denoise settings will be quite different from what you may have used before. I have found it to be desirable to follow the LMS De-Noise by a bandpass filter (F3, F4) to reduce the high frequency singing. Try SSB-N or SSB-M along with the De-Noise. LMS De-Noise is still activated with Shift F3. You set the adaptation gain with Ctrl F3/F4. The latter is now scaled 0 to 100 and if you get lost, 40 is a good starting point. Too low and it can't find a signal, too high and it oscillates. A second adjustment is the Decay rate, set in the .CFG file This controls how long the denoise waits for a signal to return after it has been found. Too high and it can built in strange ways on noise. Too low and it loses good answers after they are found, resulting in poor sound and low output. Around 90 to 95 is a good start. The LMS De-Noise delay-line length is adjustable from the .CFG file only. The default length is 61, called 1 in the .CFG file.The other option is 31 (called 0) that may under some circumstances be superior for voice signals. There is now an LMS auto-notch. This is pretty much the one from KC7WW's article in Sept 96 QEX. The adaptation gain is adjustable by bringing up the autonotch (Shift F3 now sequences through LMS Off, Denoise and Autonotch) and using the same adjustments as for the Denoiser (CTRL F3 and CTRL F4). Higher numbers are more aggressive autonotching. At 0 it does nothing and by 30 to 50 it should get most any collection of steady, clear tones. It can't find coherence in rough or noisy tones and doesn't do well on those (this is why a voice can almost completely make it's way through). Be careful about accidentally leaving the notch on while on CW. By the way, LMS stands for Least Mean Squares and is the name of the algorithm used, a part of 'Adaptive Filters.' I rearranged the processing times to allow the LMS denoise/notch and the audio filter to run in cascade without over-running "real time." Play with combinations if you want. ********************************************************************* NEW WEAK-SIGNAL MODES EME-2 MODE - The EME-2 mode is an automatic Earth-Moon-Earth echo system utilizing Long-Term Integration to enhance the apparent signal-to-noise ratio. The transmitter sends out a 2 second pulse, after which the receiver is enabled. After a delay of 2.6 seconds from the start of the pulse (to allow for the EME delay) the spectral analysis is started. This analysis continues for 2 seconds to process the entire received pulse. The spectral data is sent from the DSP to the PC for further processing. The 2 second pulses are repeated every 5 seconds for as long as one wishes. The spectral data consists of the average power received over the 2-second period for each spectral (FFT) "bin." These bins are narrow-band filters spread across the receiver pass-band and are the source of the standard display that is used in the DSP-10. The only difference from, say, USB is that the EME-2 measurement is triggered by the software. In USB it is free running and reoccurs as rapidly as possible. A "Long-Term trace" is produced on the DSP-10 screen by simply averaging the total of the spectral signal powers that have been received. With the default DSP-10 colors, this produces a Yellow trace, and this is the common nickname for the upper display in this mode. The yellow trace does not occupy the full width of the screen since the only portions of the display of interest are near the Doppler-shifted operating frequency. This is marked by a tall red-line (when using the default colors). Detailed data is taken for 10 bins above the center frequency and 10 bins below. In addition, numerical data is available in a "data box" on the returns received for the 21 bins of special interest (activated by Alt-A). Doppler correction is automatically applied to the receiver frequency so that the return will always line up with the center marker. It takes anywhere from 1 to 10,000 or more pulses to receive an unambiguous echo, depending on the station capability. The yellow trace will start as noise and, in time, this will average towards a smooth line. At the center frequency there is a sum of signal and noise power and the trace will show a spike over the noise curve. This continues to be more obvious as time progresses. Because of variation in the receiver response across the frequency band, the noise trace will not be straight as one might wish. A flattener function fits the noise curve with a fourth-order polynomial which can be subtracted from the yellow and white traces to make them nominally straight and horizontal. EME-2 is capable of major enhancements in sensitivity. Early experiments have shown the ability to work with signals 25 or more dB below the audible levels. For instance, W7LHL receives reliable returns on 1296 MHz using 50 Watts and a 44 element KLM yagi. W7SZ has had similar success using 5 Watts and a 13 foot dish. At 2-meters, this author (W7PUA) is receiving well defined echoes at 5 Watts with 4x12 element Yagis. The following material describes the controls and display for the EME-2 mode. Following this is a discussion of the steps that are ordinarily followed to obtain EME echoes in this mode, and then some Q&A on this mode. ALT-B BOX FOR EME-2 - There are not many options for the EME-2 mode. Most of these come from the Alt-B modal dialog box with a heading EME2 Echo Mode. This dialog is opened by the Alt-B (or Alt-b) command when in the EME-2 mode (Alt-M or m). The rules for all the dialog boxes are the same: The up/down cursors move between items which become highlighted. Radio buttons or check-boxes are changed by highlighting the item and hitting the space bar. Text entry is at the position of the cursor in the text box. Backspace over existing text and type in new items. Close the dialog box by hitting Enter. Item-by-item in the Alt-B Box: Random Frequency Spread, Hz, - The maximum total range that the transmitter and receiver frequencies are shifted to minimize problems due to "birdies" that can corrupt the EME data. Too little spread allows birdies to cluster near the center and to be a source of interference. Too much spread makes it difficult to find an operating frequency that does not include birdies. The selection of this parameter depends on the operating environment. A value of 200 Hz is often adequate. Record Data to Disk - Provides for a disk file that saves the important spectral data for later analysis. This feature is not normally used, but it gives a way to reanalyze the received data after Moon echoes have been received. The data format for the resulting files is in the appendix. LTerm Trace max x dB/ - This series of radio buttons prevent the hiding of details in the yellow trace by a strong, off-frequency signal. The yellow trace is self scaling and sometimes it is desirable to allow portions of the trace to run off the screen. Normally this is not a problem and the max can be left at the first button, 10 dB/. Ident String - Up to 9 characters can be entered in the text box. Every few minutes this string is sent in CW to identify the station sending the 2 second pulses. Either capital or lower-case letters will work along with any of the CW character set. Noise Blank, dB - Wide band noise can disrupt the EME-2 results by adding a large amount of power to all bins. This tends to cause the signal to sink into the noise. The average of the 18 noise-measurement bins is compared with a running average of these same bins. If the current average exceeds the running average by more than the "Noise Blank, dB," all the data from that 5 second period is discarded. A value of 0 dB will exclude roughly half the data. A value of 99 dB will never exclude any data. Values aroud 1 dB are good starting points. When the data is discarded, the yellow trace is not shown and a note "*NB*" appear on the left side of the upper spectral box. TRANSCEIVER SETTINGS FOR EME-2 - Many of the settings for EME-2 are automatic. The data is always 16 bit and the spectral display is always the lower half of the spectrum analyzer width. The SpecAve value (the effective number of FFT's being averaged per screen update) is always the value corresponding to 2 seconds worth of data, i.e., it is 4.5, 9 or 18 depending on whether the SpecAnl (Spectrum Analysis) is set to 1200, 2400 or 4800 Hz width. In EME-2 the audio portions of the receiver are fully operational, but have no interaction with the display. Controls such as Filter, LMS and AF Gain can be changed at will. The same is true of the settings that affect the upper and lower spectral displays, such as Contrast, Brightness, Trace Normalization, AutoDisplay, db/div and dB Offset. The value of the SpecAve, which is the number of power spectrums averaged before a display update takes place, is set automatically in EME-2 to update every 2 seconds. The value of SpecAve still changes with SpecAnl width since the amount of time for each FFT changes with this setting. The transmit base frequency is displayed in the Transmit Freq Box and should be chosen to find a "birdie" free frequency. Xmit Pwr - this mode has a 40% duty-cycle and the transmit power should be kept consistent with the cooling capabilities of any following amplifiers. Clock Set - No software Clock Setting is required for EME-2. However, to make the Doppler calculation correct, the PC clock should be within a few seconds. This becomes more important as the frequency of operation goes up. Twenty seconds is probably adequate at 2-meters but about 2 seconds is needed at 1296 MHz, and so forth. The windowing function (Alt-W, Alt-w or Scrl-W with kbd_alt2=1 in .CFG file) is programmable for None, Tukey-25 dB, Hamming or Blackman-harris 92 dB. These windowing functions allow trading off the selectivity of the FFT relative to the off-frequency rejection. No windowing function gives the best sensitivity and can often be used. If there are problems with strong signals or birdies, one can use a windowing function such as Tukey-25 dB, or possibly Hamming. The width of the spectral display is set with Alt-J. All FFT's are 1024 point, but the sample rate is 2400, 4800 or 9600 Hz. This produces bin resolution bandwidths, without windowing functions, of about 2.3, 4.6 and 9.2 Hz. The display in the left-hand column is "SpecAnl Window Width" where Window is the windowing functions and Width is the 1200, 2400 or 9600 Hz. Experience to date shows that the 1200 Width can be used at frequencies up to 1296 MHz. It gives better sensitivity because of the reduced noise bandwidth. As an aside, when one cuts the bin bandwidth in half, the noise bandwidth is half and the S/N of a single FFT measurement improves by 3 dB. But, each FFT measurement takes twice as long and so in any given time there is only half as much non-coherent integration. This decreases the advantage of the narrower bin bandwidth to 1.5 dB. The center FSK frequency varies with the SpecAnl width in order to center the display. For a 1200 value, the center frequency is 323 Hz, for 2400 it is 600 Hz and for 4800 width it is 900 Hz. This is shown by the 5 division tall line. The horizontal line below the tall center line shows the region for the 21 bins discussed above that get careful study. To allow the Doppler correction, it is necessary to have a "Moon:" line at the bottom of the screen (Scrl-F3). This designates the latitude and longitude of the station or stations being used. The coordinates are set in the .CFG file as is explained there, and in the previous software notes. No command is needed to activate the Doppler correction, nor can it be turned off. The receive frequency is shifted for Doppler correction. The receive frequency is shown just below the Transmit-Frequency Box and reflects the affects of both randomization and EME Doppler. To indicate that these corrections are being made, there is an "R' on the left and an "E" on the right. The transmit frequency is the "base frequency" and the FSK modulation that modifies the transmit frequency is shown just above the Transmit-Frequency Box. This system allows one to tune the frequency being used without ambiguity. ALT-A BOX for EME-2 - A non-modal information box is available for the EME-2 display. This occupies the right edge of the spectral waterfall. As the display scrolls the data up, it gets covered by this box, but no data is actually lost. The 21 bin area is far enough to the left to never be covered by the Alt-A box. This box is both opened and closed by the Alt-A keyboard command. The box is updated every 5 seconds. The top line shows the number of data points that have been averaged together and a note "Rcd=N" or "Rcd=Y" to indicate the status of data recording. The equivalent amount of effective Signal-to-Noise improvement is shown in the second line as "NonCohInt=xx.xdB." The third and fourth lines indicate the amount of signal found at the center frequency. The display shows "fffHz S+N/N=y.yyydB" and the next line is "Cntr Sig= -zzz dBm." The signal + noise/noise values are determined from the running average of the 18 noise bins. Calculations for each of the 21 center bins is made to determine the presence of a signal. The S+N/N value will be both + and - across the set of bins. Negative values are noise generated errors and correspond to signals that have negative power. Positive values allow the calculation of a signal level as shown in the "Cntr Sig=" line. An assumption needs to be made about the noise-power density and this comes from the eme2_te entry in the .CFG file. The default value for this noise temperature is 290 degrees K and corresponds to -174 dBm/Hz. One can fill in a different value for the effective noise temperature, if it is known. Only the S+N/N values are shown for the 10 bins above and the 10 bins below the center frequency. These are shown to the left and right of N1, N2, ..., N10. Finally at the bottom of the Alt-A box are the Randomization value in Hz and in the bottom line the Running Average and the Current Noise power values, in relative dB. This last line allows one to observe the operation of the Noise Blanker. The difference between the values is compared with the "Noise Blank, dB" setting from the Alt-B Box to see if noise blanking will occur. OPERATION OF EME-2 - This mode is capable of considerable sensitivity when operated for sufficient periods of time. It produces an estimate of the actual returned signal strength and can be used as a system test device to insure that all elements of an EME system are operating properly. In addition, it allows stations that cannot hear their own echoes to see how far they are below audibility and to evaluate how well the station is working. Several requirements are placed on the station hardware. The transmitter is on for about 40 percent of the time and cooling of any power amplifiers must support this level of activity. The frequency reference being used by the station need not be precise, since the return frequencies are all relative to the transmitted frequency. However, the frequency reference must not change between transmit and receive. This can be a problem when the station uses a transverter to translate microwave frequencies. The oscillator in the transverter may shift enough to prevent the returns from hitting in the 21 center bins. The use of RIT to work around this problem is described below. Sequencing for the Transmit/Receive changeover is available in the DSP-10 software. These delays are adjusted in the .CFG file and include a special delay for EME-2 that is called dly_emet2r. This delay needs to be adjusted to center up the timing of the EME-2 return. It depends on the other delays and the formula is dly_emet2r= 575 - dly_ant2amp - dly_amp2xmit - dly_xmit2amp - dly_amp2ant where all times are in milliseconds. Nominal values are: dly_ant2amp = 50 dly_amp2xmit = 75 dly_xmit2amp = 25 dly_amp2ant = 25 dly_emet2r = 400 To use the EME-2 mode, first select the EME-2 mode (Alt-M) and open the Alt-B Box. Select parameters as described above. Close this box (Enter) and open the Alt-A box to show the result data. After the antenna is positioned on the Moon, using the Az-El data at the bottom of the screen, activate the transmissions by the Home key. If transmissions are at the proper power level and you are ready for the EME-2 measurement, clear the long-term data with a Ctrl-W. The yellow trace will first look like the white trace and probably start at 5 dB per division. After a few transmissions the yellow trace will begin to smooth into a better defined curve. As it does this, the scale of dB/division will reduce automatically and the noise will be reducing in magnitude. The signal will appear as a narrow spike at the tall red line. The S+N/N numbers and the center estimated signal levels will be displayed in the Alt-A box. Let the process run until the signal becomes well defined from the noise. The amount of time required for this varies greatly depending on the station capabilities. If the station is capable of hearing their own echoes, this requires 5 to 30 seconds for an excellent display. Stations with 25 dB less capability can receive their echoes and see a high confidence display by waiting for about 30 minutes to an hour. QUESTIONS AND ANSWERS ABOUT EME-2 Q-What should you do if the display is not centered on the tall red line? A-Several things can cause this. The Doppler calculation has errors up to about 15 Hz per GHz. Until the causes of this error are identified and removed this will require a correction. Shifts in transmit/receive frequency can also cause this error. The correction is applied with RIT. To determine the amount of RIT that should be applied, some calculation is needed (this is planned to be done automatically in a future version of the program). First the bin spacing is needed. This is based on SpecAnl value: SpecAnl=1200 Bin Spacing B=2.34 Hz SpecAnl=2400 Bin Spacing B=4.69 Hz SpecAnl=4800 Bin Spacing B=9.37 Hz The RIT shift needed is roughly that to move the largest N-value to the center. N-values on the left are referred to as negative N-values as they correspond to frequencies below the expected center. Suppose, for example, that the maximum response was at N2. The RIT value needed is thus 2xB. If B was 2.34, the needed RIT would be 2 x 2.34 = 4.68 Hz and the closest available value is 5 Hz. This should be applied. If the maximum N-value was a negative N-value, the RIT would set to a negative value. In all cases, after the RIT has been changed, one should clear all the long-term data with the Ctrl-W command. Next it is possible to apply fine tuning of the RIT to not only make the center frequency return the largest value(this is the one shown at the top of the Alt-A Box as "xxxHz S+N/N+y.ydB"). The RIT can experimentally be moved a small amount to make the -N1 and +N1 values about the same. Depending on the signal strength this may require some patience. Q-How do I make the yellow noise trace straight? A-The Ctrl-E line straightener can be applied at any time. It processes the long-term averaged data (the entire yellow trace) to fit a fourth-order polynomial to the noise data. This is used as a normalization curve for both the yellow and white traces. To prevent the bending of the curve towards a strong signal, there is a weighting function applied. This decreases the importance of points that are far from the smooth curve. Q-Can I interrupt the transmission? A-Yes. All the long-term data is saved and when you hit the Home key again, it will pick up where you left off. There are two ways to clear the long-term data: the Ctrl-W key and changing modes. Q-How do I select the best SpecAnl width? This is somewhat complicated! To a point, the narrower widths have better signal-to-noise ratios since the bandwidth of the FFT bins is narrower. This does not apply if the signal being received is broader than the bin bandwidth. This can be caused by the libration of the Moon, especially for frequencies above 432 MHz. In addition, W7LHL and W7SZ have been able to observe dispersion due to rain scatter at 1296 MHz that is wider than the 2.34 Hz bin bandwidth of the narrowest width. In general, the phase noise of the DSP-10 does not seem to be a limiting factor for the available bin bandwidths. So, the answer is that one should use the narrowest width for which most of the energy is concentrated in a single bin. As the operating frequency is increased this can require going to SpecAnl widths of 2400 or 4800 Hz. Q-Do the contrast and brightness controls affect the EME-2 measurement? A-No, these are for the waterfall only. The same is true with all controls that affect the audio signal. It is easy to get a good yellow trace when no trace is observable on the waterfall. Q-How does the DSP-10 know what frequency to use for the Doppler correction? A-The frequency used is that displayed in the Transmit Frequency Box, just above the S-meter. If a transverter is being used for a frequency other than 2-meters, it must be selected (Ctrl-Alt-? where ? is the band desired). This is all done properly if the displayed frequency is the operating frequency. Q-Why is it that when I listen to the transmitted signal in a receiver it is not staying on frequency? A-This was selected in the Alt-B Box as "Random Spread, Hz" and moves both the transmit and receive frequencies in synchronism. The actual random shift for each 5 seconds is displayed in the Alt-A box. Q-How do I set f_eme in the configuration file? A-This is not used here. It applies only to the audio processor. ******************************************************************* PUA43 MODE - The PUA43 mode is designed for extremely weak communication, using very low data rates. This modes takes advantage of narrow band filtering along with long-term integration and introduces a concept called message estimation. The message lengths are a fixed number of symbols but the transmission time for the message is variable from 1 minute to as long as one has patience. In theory there is no limit to the minimum signal level. In practice the improvement over human ear/CW seems to be 20 to 25 dB, with potentially more gain for the very patient! The following is a quick summary of the code characteristics. This is followed by information on the controls available and then the operation involved in using the PUA43 code. The 43 symbols defined for the PUA43 code are The letters A-Z (or a-z converted to A-Z) The numbers 0-9 Space Period Comma (available for redefinition) Forward Slash '/' Pound Sign '#' defined as "Message Received" (redefineable) Question Mark Dollar sign '$' to shift the meaning of the following character. Each message can be selected as either 14 or 28 symbols. If the shorter length can be used, a 1.5 dB S/N advantage is gained. Starting from the top of each minute, the symbols are transmitted in message order, one every 2 seconds. If the message is only 14 symbols, it is repeated starting at 28 seconds into the minute. The 4 seconds at the end of the minute is reserved for a CW identifier that is automatically inserted. The frequencies of transmission are arranged according to the locations of the spectral (FFT) bin centers. Every fourth bin frequency is used for a tone. The bins adjacent to the 43 tone frequencies are guard bins to allow for spectral broadening of the windowing functions and for any small frequency error. The bin half way between two tone frequencies is for noise estimation. Each tone estimates the spectrally-local noise by using the power in the two second-adjacent bins. The total band of FSK offset frequencies for the 43 tones depends on the SpecAnl width as follows: SpecAnl = 1200 Hz width FSK Band = 450 to 843.7 Hz SpecAnl = 2400 Hz width FSK Band = 450 to 1237.5 Hz SpecAnl = 2400 Hz width FSK Band = 450 to 2025 Hz The CW ID is always sent at FSK = 400 Hz. In order to minimize the interference from local birdie signals, a randomization is applied to the frequency of each symbol. This is coordinated between the transmitting and receiving stations by means of a random number file. Each minute the randomization "stir" is changed that shifts the tone frequencies corresponding to a particular symbol. Provision is made for using this mode with EME work, for which it is quite suited. Doppler corrections can be automatically applied on receive and a delay of about 2.6 seconds is added to the receiver timing. Message Estimation is used to control the display. Basically each of the 14 or 28 symbol positions has two characters printed, The large-type symbols are the "most likely" and the smaller printing is the "next most likely" set of symbols. Three colors are used to represent the likelihood that the symbol is valid. The estimated message is updated every two seconds as a new symbol is received. Time and frequency accuracy are both important for this mode. Read the "Clock Set" section to see how the software clock for the DSP-10 can be set to within 10 or 20 milliseconds of UT. Frequency accuracy should be within 0.5, 1 or 2 Hz depending on the SpecAnl width. This requires an external reference for the DSP-10 of high quality and this same reference needs to be applied to any transverters used with the radio. ALT-B BOX FOR PUA43 - The options for the PUA43 mode mostly come from the Alt-B modal dialog box with a heading PUA43 SETUP. This dialog is opened by the Alt-B (or Alt-b) command when in the PUA43 mode (Alt-M or m). The rules for all the dialog boxes are the same: The up/down cursors move between items which become highlighted. Radio buttons or check-boxes are changed by highlighting the item and hitting the space bar. Text entry is at the position of the cursor in the text box. Backspace over existing text and type in new items. Close the dialog box by hitting Enter. Item-by-item in the Alt-B Box: Quality Ratio - This quantity decides the color division of the message estimation. It is a decision threshold applied to the S/N power ratio of each symbol. This value divides the low confidence color (black in the default colors) from the medium confidence (beige). This quantity squared is the threshold between the medium and high confidence (white) colors. A nominal value of 1.2 works well. The value can generally be 0.0 to 3.0. If the value is low, the bright colors are easily obtained in printing the PUA43 message. A value of 3 would make the bright white letters very difficult to obtain. Noise Decay - Long noise averages are used to estimate S/N of each of the 43 symbols. These noise estimates are decayed to discard very old data. The Noise Decay is a number very close to 1. For message times of a few minutes a value of 0.99 is reasonable. For message times in the hours a value of 0.9999 or 0.99999 can be used. 14/28 Characters/Msg - This is the number of symbols that will be transmitted or received. Ident String - Up to 9 characters can be entered in this text box. Each minute this string is sent in CW to identify the station. Either capital or lower-case letters will work along with any member of the CW character set. Transmit Lines - 14 symbols are allowed on each line. The second line is not used if only 14 symbols are being sent. If any character is in a Transmit Line that is not one of the 43, a '?' will be substituted, but no warning is generated. TRANSCEIVER SETTINGS FOR PUA43 - In PUA43 the audio portions of the receiver are fully operational, but have no interaction with the PUA43 decoding. Controls such as Filter, LMS and AF Gain can be changed at will. The same is true of the settings that affect the upper and lower spectral displays, such as Contrast, Brightness, Trace Normalization, AutoDisplay, db/div and dB Offset. The value of the SpecAve, which is the number of power spectrums averaged into each display update, is set automatically in PUA43 to the value corresponding to 2 seconds. The value of SpecAve still changes with SpecAnl width since the amount of time for each FFT changes with this setting. Therefore, the settings that need to be made are: Transmit base frequency - this is displayed in the Transmit Freq Box and needs to be the same at both the transmitting and receiving end of a contact. Xmit Pwr - this mode is virtually continuous duty-cycle and the transmit power should be kept consistent with the cooling capabilities of any following amplifiers. SpecAnl - The spectrum analysis settings are variable for Width and Windowing function: The width of the spectral display is set with Alt-J. All FFT's are 1024 point, but the sample rate is 2400, 4800 or 9600 Hz. This produces bin resolution bandwidths, without windowing functions, of about 2.3, 4.6 and 9.2 Hz. The display in the left-hand column is "SpecAnl Window Width" where Window is the windowing functions and Width is the 1200, 2400 or 9600 Hz. Experience to date shows that the 1200 Width can be used at frequencies up to 1296 MHz. It gives better sensitivity because of the reduced noise bandwidth. As an aside, when one cuts the bin bandwidth in half, the noise bandwidth is half and the S/N of a single FFT measurement improves by 3 dB. But, each FFT measurement takes twice as long and so in any given time there is only half as much non-coherent integration. This decreases the advantage of the half-size bin bandwidth to 1.5 dB. The windowing function (Alt-W, Alt-w or Scrl-W with kbd_alt2=1 in .CFG file) is programmable for None, Tukey-25 dB, Hamming or Blackman-harris 92 dB. These windowing functions allow trading off the selectivity of the FFT relative to the off-frequency rejection. No windowing function gives the best sensitivity and can often be used. If there are problems with strong signals or birdies, one can use a windowing function such as Tukey-25 dB, or possibly Hamming. SpecAve - In the middle of this line is a number of either 8 or 16. This is the number of bits in the spectral data word being sent from the DSP to the PC. This MUST be 8 for PUA43 to operate (use Scrl-F2). Clock Set - The software clock must be set for PUA43. See the separate section on Clock Setting below. EME Corrections - This is toggled on and off by the Alt-L or Alt-l key and enables the Doppler correction and the 2.6 receive timing offset. The path must be selected on the bottom line as "MOON:Local Stn, Other Stn..." by use of the Srcl-F3, '<', and '>' commands. This requires that the latitude and longitude for both stations has been entered into the .CFG file, HARDWARE REQUIREMENTS FOR PUA43 - PUA43 requires the RF portions of the DSP-10. In addition, an external reference of high quality (0.5 Hz or better at the operating frequency. Current testing of this mode is being done with high quality crystal oscillators interfaced to the Brooks Shera GPS controller (July 1998 QST). The processor load for the PC is quite a bit more intensive than for the basic modes such as USB or CW. A 486 computer operating at 50 MHz will generally support this mode, or at least that has been the experience so far. The transmit function is continuous and the cooling needs to support this. Use the Xmit Pwr function (Scrl-O, P) to reduce the transmitter output if cooling is a problem. OPERATION OF PUA43 - Both the transmit and receive ends of a PUA43 contact must be coordinated in * Base transmit Frequency * SpecAnl Width * 14 or 28 Characters per Message * EME or terrestrial path (Alt-L, l). In addition, both stations must have their software clocks set to within 30 milliseconds of UT. To make this possible the DOS environment must have their time zone defined with "TZ PST8" or the appropriate terminology for the local PC clock time (see...). If the path is EME, be sure that the Doppler correction is correct. The best place to check this is in the EME-2 mode. At 2-meters it is probably good enough, but as the frequency goes up the error gets greater. The transmitting station hits the Home key to initiate transmit. At roughly the same time, the receiving station hits Ctrl-W to clear all the long-term data. Both stations should have their Alt-A boxes open that shows the status of the communication. The process is allowed to continue without any operator intervention required. The transmit end will continue to repeat the message in FSK until the Home key toggles transmit back to receive. At the receive end the estimated message will be displayed in the top text area. The most-likely message is displayed in large print with a cursor showing where the processing is currently taking place. Above each large letter, in small print is a second most-likely set of symbols. The confidence level is reflected in the color of the printing, with black, beige and white being low, medium and high confidence. With time, and an adequate signal level, the correct message will be displayed in white print and the line above it will be random symbols in black print. For weak signals this can take quite a bit of time, running into the hours. QUESTIONS AND ANSWERS ABOUT PUA43 Q-What do you do if nothing is being received? A-Check the timing against WWV. In transmit for all paths this can be done by watching the "T:tt" time indicator in the lower-left corner of the screen. This is indicating the current seconds. It should transition at the same time as the tick on WWV. In receive, this can be observed as well for terrestrial paths. The lower-left corner shows R:tt and should be exactly the same as the T:tt value. For EME paths, the receiving station needs to compare with WWV by bringing up the clock-set routine (Alt-K). The correct minute can be confirmed by comparing the Alt-A Box values for Random Stir. If these don't agree there is an error in the time or date at one end or the other. If timing is correct, check the frequencies and the other parameters. Q-The right characters are received, but they are in the wrong position in the message. What is wrong? A-This is the sign of a timing error. If the stations disagree by one second there can even be a duplication of some letters in adjacent positions. Q-What is wrong when a message seems to be coming through but the characters are wrong, like W7ABC is X8BCD. A-This is the sign of a frequency error. If the path is EME and the frequency band is above 2-meters, the problem may be a Doppler calculation error. The best way to find the RIT value, if needed is for the receive station to go back to EME-2 mode, set the bottom line for self-echoes, and find the RIT for their own echoes. Then apply this same RIT value at the receive station. Experience to date suggest that this is a good procedure to follow at the beginning of a UHF/Microwave EME contact using PUA43. Q-Why do I get OK returns on EME-2 but nothing in PUA43. Assuming that all the parameters are set correctly, it is important to note that PUA43 requires more than 5xlog10(14)=5.7 dB (for 14 symbol message) extra signal. This translates to an increase in time of at least 14 times. This is the price for only working on 14 different symbols! In addition, there may be additional increases in time due to the statistics of getting good confidence on the entire message. Q-Copy was good on a terrestrial path and the message went to poor copy. What happened? A-This sounds like a Boeing 747! Airplane reflections can be much stronger than the tropospheric-scatter signal. Often the Doppler shifted signal will disrupt copy. It may be necessary to clear the long-term buffers (Ctrl-W) and start again. Q-What is the effect of rain scatter on PUA43 copy? A-It seems to depend a lot on frequency of operation and the nature of the rain. Below 1296 MHz, this is probably not a factor. At 10 GHz, rain can enhance the signal. But the Doppler shift is sometimes more to one side of the center frequency and can disrupt the copy. If the Doppler spread is close to uniform about the center tone frequencies, copy will generally be acceptable. In general, wider SpecAnl widths are better with rain. ********************************************************************* LHL-7 MODE When one looks at a waterfall spectral display it is practical to copy very slow Morse code by looking where the trace is lighter. W7LHL has come up with a code that is based on this principle, but is faster for a given level of performance and lends itself to easier manual and automatic decoding. This code is of a fixed length and therefore does not adapt to the signal strengths as can be done with PUA43. But due to the arrangement of the frequencies it is much more robust to frequency errors. It likewise is less subject to disruption by airplane and rain scattering. LHL-7 uses 7 FSK tones arranged as End of Character 600 Hz Dot 750 Hz Dash 900 Hz Times 2 1050 Hz Times 3 1200 Hz Times 4 1350 Hz Times 5 1500 Hz For instance, the letter 'E', a single dot, would be two tones, 750 and 600 Hz, one after the other. The number '7' dash-dash-dot-dot-dot would be five tone bursts, 900, 1050, 750, 1200, 600 Hz. The lengths of the tones are preset by agreement between the sending and receiving ends and can be 2, 4, 10, 20 or 60 seconds per tone. The top of the minute is always a break between tones. Thus if the tones were 20 seconds each, they would start at 0, 20 and 40 seconds after the start of the minute. To gain the maximum advantage, the clocks at each end should be set to a small part of a second. The demodulator is flexible in the frequency tolerance it will accept. The best performance comes from a frequency accuracy of a Hz or better, but automatic demodulation/decoding can be set to accept up to about 40 Hz of frequency error. Manual decoding can be more tolerant by watching the wavy lines! Provision is made for adding an identification in Morse code. This is entered as a ampersand '&' and the preset ID phrase is sent at 500 Hz offset. A Test Pattern consisting of each of the seven tones, in order of low to high, is sent if the plus sign '+' is entered. This is very useful for manual decoding. DSP-10 SETTINGS FOR LHL-7 - Both ends of a path must agree on the base frequency, displayed in the Transmit Frequency box. The only SpecAnl width available is 4800 Hz, but the screen only displays half of this as experience has shown this works best for manual copy. The SpecAve values are adjustable to all values that are consistent with the length of the tones. The power should be adjusted to not overheat any following power amplifiers with the nearly 100% duty cycle. The windowing function should be set to None (Non) or Tukey25 (Tuk) if a single center bin is being used for the signal. This implies that very good frequency accuracy is available on both ends. If multiple bins are assigned to the signal, the windowing function should be set to Hamming (Ham). Clock Set - The software clock must be set for LHL-7. See the separate section on Clock Setting below. EME Corrections - This is toggled on and off by the Alt-L or Alt-l key and enables the Doppler correction and the 2.6 receive timing offset. The path must be selected on the bottom line as "MOON:Local Stn, Other Stn..." by use of the Srcl-F3, '<', and '>' commands. This requires that the latitude and longitude for both stations has been entered into the .CFG file, ALT-B BOX FOR LHL-7 - This box is used to select options for the LHL-7 code. The details of the options are explained below, but the Alt-B Box entries are as follows: 1, 3, 5, 7 or 9 bins to be included Peak or Average Levels 2, 4, 10, 20 or 60 seconds per tone The ID string, such as "DE W7XYZ" Single or 2 Line decoder output The last item refers to the very top of the upper spectral display that is used for input and output of symbols for the LHL-7 transmission and reception. The selection is either a single line of twice high characters, or two lines of standard characters. Either way, when the line() get full they will scroll to make more room. ALT-A DATA BOX FOR LHL-7 - After you get to LHL-7 mode (Alt M) use Alt A to bring up the informational box on the right. This can be left open. It shows the operating parameters and the measured S+N/N. The following items are shown in the box: "Tag" is the seconds after the top of the minute that was attached to the last measurement. It is not the current time, as it takes about 3 seconds to process the data, and if the seconds per tone is greater than 2, the tone length must be added in. "bigi" is the strongest of the 7 tones for the last measurement period. They are numbered 0 for 600 Hz, 1 for750 Hz, ... "mcbyte" is the sequence of tones that have been decoded. This is compared with the Morse code table. Any time it is found not possible to decode a character in the future, the code is reset. A reset code is a single 1, as this is a prefix character to know where the number begins. For instance, all characters must start with a dot or a dash, and we will not accept a x3 at the beginning. This process continues as the mcbyte is built up. If you watch it, the incorrect byte will show up and it will then disappear and be followed by either "1", "10", or "11". Due to the changing of tones, over the months, and my desire to not redo the character table, the mcbyte numbers are not quite the same as the bigi values. For mcbyte, 0=dot, 1=dash, 2=x2, 3=x3, 4=x4, 5=x5 and the EOC does not show up in mcbyte. "Bins per tone" refers to the 9.375 Hz spaced filter bins. 1 means that only the bin at the center frequency will be examined. 3 means that the center and one on either side of it will be examined. This continues up to 9. If these are averaged, it is fully equivalent to having a wider FFT bin. There are guard bins on either side of the tone bins, since the windowing causes widening of the signal. Outside of the guard bins are four noise measurement bins that are averaged together to estimate the noise level. Thus the frequency bins for "3 bins per tone", for the 600 Hz tone would be: Bin 60 562.5 Hz Noise Bin 61 571.9 Hz Noise Bin 62 581.2 Hz Guard Bin 63 590.6 Hz Tone signal Bin 64 600.0 Hz Tone Signal Bin 65 609.4 Hz Tone Signal Bin 66 618.7 Hz Guard Bin 67 628.1 Hz Noise Bin 68 637.5 Hz Noise Bins 69 to 75 are not used and the 750 Hz related bins start at 76. Higher numbers of "Bins per tone" use more of the in-between bins until they are almost all used. "Peak" and "Ave" refer to the way that the Tone signal bins are treated. Either the average of all of them is compared with the noise, or the largest power is. For wide band signals, like rain scatter, average would be best. For drifting signals, peak might be---these are areas that need further exploration. OPERATION WITH LHL-7 CODE - This operation is quite similar to CW. Symbols to be sent are typed in and show on the very top line. Transmission is initiated by the Home key. Reception is monitored in the Alt-A box and can sometimes be seen on the waterfall. Only those tone combinations that create a valid symbol will produce an output on the display. EME paths will require the Alt-L (or Alt-l) Doppler correction that is explained above. This only required at the receiving end of the path. ********************************************************************* LTI LONG-TERM INTEGRATION MODE - This mode is a study tool for the use of Long-Term Integration to enhance the apparent signal-to-noise ratio of the received signal. It is capable of working with signals in the -180 to -190 dBm range, which is 30 to 40 dB weaker than can be copied by ear in CW. One may devise ways to use this mode to perform extremely slow communications, but the primary purpose is to explore the nature of difficult propagation paths and to allow determination of the potential for using other codes such as PUA43, LHL7 and CW. In transmit, this mode is very similar to CW, except that provision is made for randomizing the frequency. In receive, the processing of the data and its presentation is very much like EME-2. The duty cycle in LTI can be raise to almost 100% which allows better sensitivity than can be achieved with EME-2, which is set at 40%. The following material should be helpful in using the LTI mode. The emphasis is put on the controls and the functioning of the mode, rather than a specific "how to use it." This is primarily a measurement tool and the applications for the tool are mostly left to the creativity of the user! Reception takes place with un-triggered FFT spectral processing. Two displays are presented in the upper spectral display. The conventional "white" trace is at the bottom. Above this is a "yellow" trace with the LTI results. A "Long-Term trace" is produced on the DSP-10 screen by simply averaging the total of the spectral signal powers that have been received. With the default DSP-10 colors, this produces a Yellow trace, and this is the common nickname for the upper display in this mode. The yellow trace can be set to not occupy the full width of the screen to reduce the amount of processing to update the display. The center operating frequency is marked by a tall red-line (when using the default colors). Detailed data is taken for 10 bins above the center frequency and 10 bins below. In addition, numerical data is available in a "data box" on the returns received for the 21 bins of special interest (activated by Alt-A). Doppler correction for the EME path is optionally applied to the receiver frequency so that the return will always line up with the center marker. It may take many thousands of FFT processes to produce an unambiguous signal trace. The yellow trace will start as noise and, in time, this will average towards a smooth line. At any frequency where a signal is present, there is a sum of signal and noise power and the trace will show a spike over the noise curve. This continues to be more obvious as time progresses. Because of variation in the receiver response across the frequency band, the noise trace will not be straight as one might wish. A flattener function fits the noise curve with a fourth-order polynomial which can be subtracted from the yellow and white traces to make them nominally straight and horizontal. ALT-B BOX FOR LTI - Most of the options for the LTI mode come from the Alt-B modal dialog box with a heading LTI Long Term Mode. This dialog is opened by the Alt-B (or Alt-b) command when in the LTI mode (Alt-M or m). The rules for all the dialog boxes are the same: The up/down cursors move between items which become highlighted. Radio buttons or check-boxes are changed by highlighting the item and hitting the space bar. Text entry is at the position of the cursor in the text box. Backspace over existing text and type in new items. Close the dialog box by hitting Enter. Item-by-item in the Alt-B Box: Display Fr Low, Hz - This is the lowest frequency that is displayed in both the white and yellow traces. It must be consistent with the SpecAnl width or a warning message will be generated and the value will be adjusted. Center Freq, Hz - This is the frequency in the display that is over the tall red line. Detailed data is shown for this frequency along with 10 spectral bins above and below (see Alt-A box below). In transmit, this is the FSK offset above the Base Transmit Frequency before modification by a random frequency. It must be consistent with the SpecAnl width or a warning message will be generated and the value will be adjusted. Display Fr High, Hz - This is the highest frequency that is displayed in both the white and yellow traces. It must be consistent with the SpecAnl width or a warning message will be generated and the value will be adjusted. Random Frequency Spread, Hz, - The maximum total range that the transmitter and receiver frequencies are shifted to minimize problems due to "birdies" that can corrupt the data. The spreading changes each minute and is based on the same file of random numbers that is used in PUA43 mode. Too little spread allows birdies to cluster near the center and to be a source of interference. Too much spread makes it difficult to find an operating frequency that does not include birdies. The selection of this parameter depends on the operationg environment. A value of 200 Hz is often adequate. To disable the random feature set this value to zero. Record Data to Disk - Provides for a disk file that saves the important spectral data for later analysis. This feature is not normally used, but it gives a way to reanalyze the received data after Moon echoes have been received. The data format for the resulting files is in the appendix. LTerm Trace max x dB/ - This series of radio buttons prevent the hiding of details in the yellow trace by a strong, off-frequency signal. The yellow trace is self scaling and sometimes it is desirable to allow portions of the trace to run off the screen. Normally this is not a problem and the max can be left at the first button, 10 dB/. Noise Blank, dB - Wide band noise can disrupt the LTI results by adding a large amount of power to all bins. This tends to cause the signal to sink into the noise. The average of the 18 noise-measurement bins is compared with a running average of these same bins. If the current average exceeds the running average by more than the "Noise Blank, dB," all the data from that 5 second period is discarded. A value of 0 dB will exclude roughly half the data. A value of 99 dB will never exclude any data. Values around 1 dB are good starting points. When the data is discarded, the yellow trace is not shown and a note "*NB*" appear on the left side of the upper spectral box. TRANSCEIVER SETTINGS FOR LTI - In LTI mode the audio portions of the receiver are fully operational, but have no interaction with the display. Controls such as Filter, LMS and AF Gain can be changed at will. The same is true of the settings that affect the upper and lower spectral displays, such as Contrast, Brightness, Trace Normalization, AutoDisplay, dB/div, dB Offset and SpecAve. The transmit base frequency is displayed in the Transmit Freq Box and should be chosen to find a "birdie" free frequency. Xmit Pwr - the duty cycle in the LTI mode can be adjusted by the message being sent. In turn, this adjusts the duty cycle duty-cycle being transmitted. The transmit power should be kept consistent with the cooling capabilities of any following amplifiers and the selected duty cycle. Clock Set - The software clock is needed if Randomization is applied to the transmit/receive frequencies. The accuracy requirements are relaxed from that required for some of the other modes and 0.1 seconds is quite sufficient, since the randomization only changes once per minute. See the Clock Setting section below for further information. The windowing function (Alt-W, Alt-w or Scrl-W with kbd_alt2=1 in .CFG file) is programmable for None, Tukey-25 dB, Hamming or Blackman-harris 92 dB. These windowing functions allow trading off the selectivity of the FFT relative to the off-frequency rejection. No windowing function gives the best sensitivity and can often be used. If there are problems with strong signals or birdies, one can use a windowing function such as Tukey-25 dB, or possibly Hamming. The width of the spectral display is set with Alt-J. All FFT's are 1024 point, but the sample rate is 2400, 4800 or 9600 Hz. This produces bin resolution bandwidths, without windowing functions, of about 2.3, 4.6 and 9.2 Hz. The display in the left-hand column is "SpecAnl Window Width" where Window is the windowing functions and Width is the 1200, 2400 or 9600 Hz. Experience to date shows that the 1200 Width can be used at frequencies up to 1296 MHz. It gives better sensitivity because of the reduced noise bandwidth. As an aside, when one cuts the bin bandwidth in half, the noise bandwidth is half and the S/N of a single FFT measurement improves by 3 dB. But, each FFT measurement takes twice as long and so in any given time there is only half as much non-coherent integration. This decreases the advantage of the narrower bin bandwidth to 1.5 dB. EME Doppler correction (Alt-L or At-l) is available in the LTI mode. To allow the Doppler correction, it is necessary to have a "Moon:" line at the bottom of the screen (Scrl-F3). This designates the latitude and longitude of the station or stations being used. The coordinates are set in the .CFG file as is explained there, and in the previous software notes. The receive frequency is shifted for Doppler correction and is shown just below the Transmit-Frequency Box, reflecting the affects of both randomization and EME Doppler. To indicate that these corrections are being made, there is an 'R' on the left (if the randomization frequency spread is non-zero) and an 'E' on the right (if EME Doppler is enabled). The transmit frequency is the "base frequency" and the FSK modulation that modifies the transmit frequency is shown just above the Transmit-Frequency Box. This system allows one to tune the frequency being used without ambiguity. ALT-A BOX for LTI - A non-modal information box is available for the LTI display. This occupies the right edge of the spectral waterfall. As the display scrolls the data up, it gets covered by this box, but no data is actually lost. The 21 bin area is far enough to the left to never be covered by the Alt-A box. This box is both opened and closed by the Alt-A keyboard command. The box is updated along with every screen update. The top line shows the number of data points that have been averaged together and a note "Rcd=N" or "Rcd=Y" to indicate the status of data recording. The equivalent amount of effective Signal-to-Noise improvement is shown on the right "xx.xdB." The third and fourth lines indicate the amount of signal found at the center frequency. The display shows "fffHz S+N/N=y.yyydB" and the next line is "Cntr Sig= -zzz dBm." The signal + noise/noise values are determined from the running average of the 18 noise bins. Calculations for each of the 21 center bins is made to determine the presence of a signal. The S+N/N value will be both + and - across the set of bins. Negative values are noise generated errors and correspond to signals that have negative power. Positive values allow the calculation of a signal level as shown in the "Cntr Sig=" line. An assumption needs to be made about the noise-power density and this comes from the eme2_te entry in the .CFG file. The default value for this noise temperature is 290 degrees K and corresponds to -174 dBm/Hz. One can fill in a different value for the effective noise temperature, if it is known. This signal level takes into account the noise-bandwidth of the FFT and windowing processes as well as the estimated noise power density. Only the S+N/N values are shown for the 10 bins above and the 10 bins below the center frequency. These are shown to the left and right of N1, N2, ..., N10. Finally at the bottom of the Alt-A box are the Randomization frequency shift in Hz and in the bottom line the Running Average and the Current Noise power values, in relative dB. This last line allows one to observe the operation of the Noise Blanker. The difference between the values is compared with the "Noise Blank, dB" setting from the Alt-B Box to see if noise blanking will occur. OPERATION OF LTI - This mode is capable of considerable sensitivity when operated for sufficient periods of time. It produces an estimate of the actual returned signal strength. It is very useful as a test device between two suitably equipped stations to insure that all elements are operating properly. This can be applied to terrestrial or EME paths. The single biggest hardware requirement for the LTI mode is the frequency reference. The spectral bins are narrow and the frequency control for both transmitting and reception must be tight relative to the bin width. As is the case for the PUA43 mode the accuracy, at the final operating frequency should be 0.5 Hz or better. This can be relaxed by a factor of 2 or 4 if the SpecAnl width is 2400 or 4800 Hz, but the requirements are still tight. To use the LTI mode, first select it (Alt-M) and open the Alt-B Box. Select parameters as described above. Close this box (Enter) and open the Alt-A box to show the result data. Reception begins immediately. When the transmitting station starts to send, you should clear the long-term data with a Ctrl-W. The yellow trace will first look like the white trace and probably start at 5 dB per division. After a few transmissions the yellow trace will begin to smooth into a better defined curve. As it does this, the scale of dB/division will reduce automatically and the noise will be reducing in magnitude. The signal will appear as a narrow spike at the tall red line. The S+N/N numbers and the center estimated signal levels will be displayed in the Alt-A box. For the transmitting station, the Home key puts the transmitter into operation. This then behaves very much like CW. Hit a key for a CW character and it is shown on one of 3 lines at the top of the screen. This is sent in CW as soon as it is typed. To make a more useful LTI signal, the loop feature ('#' pound signs) should be used, along with the '{' character that will hold the key down for 2 seconds and '}' that will leave the key up for 2-seconds. So to continuously transmit 20 seconds of key down followed by a CW identification, one should enter "#{{{{{{{{{{ DE W7XYZ #". This can be entered while in receive, but the entered data will not show since it is being overwritten with spectral data. Let the process run until the signal becomes well defined from the noise. The amount of time required for this varies greatly depending on the station capabilities. Depending on the time available and the system noise temperature, it generally is possible to see good traces on signals in the -170 to -195 dBm range (audible copy for CW is about -145 to -155 dBm). ********************************************************************* SOFTWARE CLOCK SETTING - The clock that is built into the PC is not generally accurate enough to hold the timing required for the PUA43 and LHL7 modes. In addition, a clock setting system is required to be able to move the clock time into exact synchronism with UT (old GMT). Keep in mind that there are two problems to be dealt with, the rate that the clock runs and the setting of the clock time. The following description of the method is based on an excellent write-up by Mike Reed, KD7TS: The easiest method for me has been using WWV time signals listening to the TICK, that is easiest to hear between announcements. Initially I checked the rate of the PC clock by letting the computer run for 48 hours with the clock rate set at 1.0000000. This gave results with enough significant digits that being accurate to the nearest second resulted in a good number to crank in for Clock Rate. I used WWV for all my references. To generate the Clock Rate number I used the following formula: For a PC clock that runs too fast: One minus the error in seconds, ES, divided by interval in seconds, IS: ES Clock Rate = 1 - ----- IS For a PC clock that runs too slow: One plus the error in seconds, ES, divided by interval in seconds, IS: ES Clock Rate = 1 + ----- IS The resulting number should be slightly larger or smaller than 1. Then I brought up the clock screen (Alt-K) and used the < or > to scroll the Clock Rate number up or down until I got the correct digits. The correction required for my computer was 1.0000940. After I entered the corrected Clock Rate I used the WWV ticks to find the Offset necessary to get the software clock within a few milliseconds. As shown in the Clock Box, the rules are: ( and ) Retard or advance the "Software Clock" time by 1 second { and } Retard or advance the "Software Clock" time by 10 millisec [ and ] Retard or advance the "Software Clock" time by 1 millisec First I used '{' and '}' until the TICKS occurred at the same time as the clock change in the Clock Box. When the Clock Box is open there is a TICK coming from the PC internal speaker that can be aligned audibly with WWV TICK, as well. Next I listened to the time announcement to get the whole seconds correct which is controlled by parenthesis '(' and ')'. With a little practice the process becomes easier. WWV TICK =x PC TICK = + | x + T | x + i | x + m | x + e | x + | x + running the TICK time difference to | x + zero with '{' and '}' | x | +x | + x | + x |____________________ difference If you shut down the PC program the Offset in time will be lost but the correction factor (Clock Rate) stays in UHFA.CFG and doesn't have to be done again. It only takes a minute or two to get the offset set correctly again. The Clock Box is closed by the asterisk, '*' and the program operation is held up by the modal box operation until the box is closed. Keep in mind there are three timepieces we have to deal with: 1. WWV (our standard) 2. PC clock (the wanderer, nothing can be done about this) 3. Software Clock (where we apply RATE and OFFSET to track WWV) We multiply the PC clock by the Clock Rate to arrive at WWV's time. We use the offset to match the last small difference because the CMOS clock setting can't be set very accurately. In addition to Mike's write-up, note that the Clock Rate setting in the Clock Box follows the rule: < and > Retard or advance the "Software Clock" speed by 1x10^-7 The nominal software clock speed, relative to the PC clock is 1.0. A single step of the '>' key makes Clock Rate 1.0000001, relative to the PC clock, or 1 part in 10 to the seventh faster. The next '>' key makes it 1.0000002, and so forth. If one used the '<' key Clock Rate would go to 0.9999999 of the PC clock. The Clock Rate is available in the .CFG file under the name clock_speed and can be modified there, as well as in the Clock Box. Not used in Mike's procedure, but available for use if it simplifies someone's operation, is a software clock sync pulse. This comes from the parallel port, bit 0, as a 20-microsecond, positive-going pulse at the start of the software clocks second. This appears on pin 2 of the 25-pin connector with ground on pins 18 through 25. Activation of this feature is done by changing the value of lpt_port in the .CFG file to 888 for LPT1 or 956 for LPT2, assuming that the port assignment is conventional. If the value of lpt_port is the default 0 the pulse will not be sent. This pulse can be used to trigger a 'scope that is looking at the audio coming from WWV or it can look at the 1 second pulse coming from some GPS receivers, such as the Motorola UT+. If one is into precision, it is possible to correct for the propagation time from WWV, but this is not necessary for the codes being used. For those out-of- range of WWV it would be good to get a listing of other time references that could be listed here. Eventually, it is planned that the clock setting can be automated using the GPS receivers.