EMRFD Message Archive 5926

5926 2011-03-17 03:32:13 Tim Ham band freqs from microprocessor crystals Message Date From Subject EMRFD usefully points out some crystal-controlled or VXO-type rigs that use digital-logic-divide followed by simple harmonic multipliers. This is particularly appealing because the divider outputs are so rich in harmonic content (especially odd harmonics), and it affords a lot of isolation. In the spirit of this, I took the list of 1-50MHz crystal frequencies (largely "microprocessor type") from the Mouser catalog and looked for simple m/n ratios that end up in what I consider desirable parts of the HF CW bands. My taste in HF CW frequencies won't match everyone's - I seem to be excluding some of the former novice bands - and I seem to have included some swaths that are largely PSK31 or other digital modes lately. I find no shortage of possibilities (note that not all multiply/divide numbers are as easy as others... an even divider followed by an odd multiplier is perhaps most desirable but I list the others because they seem not too much harder). The leftmost column is the Mouser-catalog-crystal frequency, followed by the multiply/divide ratio, and then I give the CW-ish ham band frequency that results. It is sorted by ham band frequency: 9.341 * 3 / 8 = 3.502875 4.096 * 6 / 7 = 3.51085714285714 4.9152 * 5 / 7 = 3.51085714285714 6.144 * 4 / 7 = 3.51085714285714 8.192 * 3 / 7 = 3.51085714285714 12.288 * 2 / 7 = 3.51085714285714 24.576 * 1 / 7 = 3.51085714285714 12.28888 * 2 / 7 = 3.51110857142857 9.385 * 3 / 8 = 3.519375 35.2512 * 1 / 10 = 3.52512 4.032 * 7 / 8 = 3.528 28.224 * 1 / 8 = 3.528 9.409 * 3 / 8 = 3.528375 6.176 * 4 / 7 = 3.52914285714286 12.352 * 2 / 7 = 3.52914285714286 35.35 * 1 / 10 = 3.535 3.93216 * 9 / 10 = 3.538944 4.433618 * 4 / 5 = 3.5468944 17.734475 * 1 / 5 = 3.546895 4.433619 * 4 / 5 = 3.5468952 5.0688 * 7 / 10 = 3.54816 17.7456 * 1 / 5 = 3.54912 4 * 8 / 9 = 3.55555555555556 6.4 * 5 / 9 = 3.55555555555556 8 * 4 / 9 = 3.55555555555556 16 * 2 / 9 = 3.55555555555556 32 * 1 / 9 = 3.55555555555556 16.000312 * 2 / 9 = 3.55562488888889 5 * 5 / 7 = 3.57142857142857 6.25 * 4 / 7 = 3.57142857142857 25 * 1 / 7 = 3.57142857142857 28.63 * 1 / 8 = 3.57875 3.579 * 1 / 1 = 3.579 9.545 * 3 / 8 = 3.579375 21.4772 * 1 / 6 = 3.57953333333333 28.6363 * 1 / 8 = 3.5795375 3.579545 * 1 / 1 = 3.579545 10.738635 * 1 / 3 = 3.579545 14.31818 * 1 / 4 = 3.579545 28.63636 * 1 / 8 = 3.579545 10.73864 * 1 / 3 = 3.57954666666667 3.57955 * 1 / 1 = 3.57955 2.048 * 7 / 4 = 3.584 4.032 * 8 / 9 = 3.584 4.096 * 7 / 8 = 3.584 16.128 * 2 / 9 = 3.584 3.1872 * 9 / 8 = 3.5856 5.9904 * 3 / 5 = 3.59424 11.98135 * 3 / 10 = 3.594405 4.1943 * 6 / 7 = 3.59511428571429 4.194304 * 6 / 7 = 3.59511771428571 3 * 6 / 5 = 3.6 6 * 3 / 5 = 3.6 12 * 3 / 10 = 3.6 9.341 * 3 / 4 = 7.00575 11.228 * 5 / 8 = 7.0175 16.38 * 3 / 7 = 7.02 6.144 * 8 / 7 = 7.02171428571429 8.192 * 6 / 7 = 7.02171428571429 9.8304 * 5 / 7 = 7.02171428571429 12.288 * 4 / 7 = 7.02171428571429 16.384 * 3 / 7 = 7.02171428571429 24.576 * 2 / 7 = 7.02171428571429 49.152 * 1 / 7 = 7.02171428571429 12.28888 * 4 / 7 = 7.02221714285714 8.432 * 5 / 6 = 7.02666666666667 6.25 * 9 / 8 = 7.03125 9.385 * 3 / 4 = 7.03875 35.2512 * 1 / 5 = 7.05024 4.032 * 7 / 4 = 7.056 11.2896 * 5 / 8 = 7.056 28.224 * 1 / 4 = 7.056 9.409 * 3 / 4 = 7.05675 6.176 * 8 / 7 = 7.05828571428571 12.352 * 4 / 7 = 7.05828571428571 35.35 * 1 / 5 = 7.07 5.5 * 9 / 7 = 7.07142857142857 3.93216 * 9 / 5 = 7.077888 8.5 * 5 / 6 = 7.08333333333333 4.433618 * 8 / 5 = 7.0937888 17.734475 * 2 / 5 = 7.09379 4.433619 * 8 / 5 = 7.0937904 5.0688 * 7 / 5 = 7.09632 17.7456 * 2 / 5 = 7.09824 11.228 * 9 / 10 = 10.1052 16.849 * 3 / 5 = 10.1094 8.432 * 6 / 5 = 10.1184 6.7458 * 3 / 2 = 10.1187 9 * 9 / 8 = 10.125 13.5 * 3 / 4 = 10.125 27 * 3 / 8 = 10.125 17.7456 * 4 / 7 = 10.1403428571429 16.000312 * 7 / 8 = 14.000273 9.341 * 3 / 2 = 14.0115 11.228 * 5 / 4 = 14.035 16.38 * 6 / 7 = 14.04 16.849 * 5 / 6 = 14.0408333333333 12.288 * 8 / 7 = 14.0434285714286 16.384 * 6 / 7 = 14.0434285714286 19.6608 * 5 / 7 = 14.0434285714286 24.576 * 4 / 7 = 14.0434285714286 32.768 * 3 / 7 = 14.0434285714286 49.152 * 2 / 7 = 14.0434285714286 12.28888 * 8 / 7 = 14.0444342857143 8.432 * 5 / 3 = 14.0533333333333 6.25 * 9 / 4 = 14.0625 9.385 * 3 / 2 = 14.0775 14.4756 * 5 / 4 = 18.0945 24.00014 * 7 / 8 = 21.0001225 9.341 * 9 / 4 = 21.01725 16.38 * 9 / 7 = 21.06 16.849 * 5 / 4 = 21.06125 16.384 * 9 / 7 = 21.0651428571429 18.432 * 8 / 7 = 21.0651428571429 24.576 * 6 / 7 = 21.0651428571429 29.4912 * 5 / 7 = 21.0651428571429 49.152 * 3 / 7 = 21.0651428571429 8.432 * 5 / 2 = 21.08 9.341 * 8 / 3 = 24.9093333333333 24.00014 * 7 / 6 = 28.0001633333333 16.000312 * 7 / 4 = 28.000546 9.341 * 3 / 1 = 28.023 11.228 * 5 / 2 = 28.07 16.849 * 5 / 3 = 28.0816666666667 24.576 * 8 / 7 = 28.0868571428571 32.768 * 6 / 7 = 28.0868571428571 49.152 * 4 / 7 = 28.0868571428571 6.25 * 9 / 2 = 28.125 25 * 9 / 8 = 28.125 Hi Tim, Many thanks for your hard work on identifying possibilities for Microprocessor crystals that result in frequencies on the ham bands. In the spirit of adding my two cents here are a couple of additional inputs regarding the application of the Microprocessor crystals with a VXO using a slightly different approach that puts you right in the ham bands. Mouser sells a 12.228 MHz Microprocessor crystal. If you use several in a Super VXO configuration the frequency swing is in excess of 30 KHz. As luck would have it if you were to use an IF or Mixer Frequency of 5.185 MHz (another Microprocessor crystal) the result is that you have a signal source that tunes from 7.040 to 7.010 MHz. This approach has been successfully implemented in hardware. A simple mixer is the venerable NE602/SA602/SA612 Gilbert Cell Mixer. Three Microprocessor Crystals, at 12.96 MHz were used in a Super VXO configuration and heterodyned with a 6.176 MHz Microprocessor Crystal and this resulted in a 60 kHz frequency spread at 19.1 MHz for use as an LO in a 20M SSB transceiver using a 4.9152 MHz homebrew crystal filter crystal. I went one step further to make the VXO crystal switched so that the VXO range remained fixed but the heterodyne crystal is switched. The 6.25 MHz frequency you listed would put the resultant transceiver in 2 ranges with the second covering 14.290 down to about 14.230. As of about a week ago Mouser advised that the manufacturer Abracon was having trouble making that frequency and they were cancelling my order. The circuit to accomplish this consists of a SA602 that is used both for the VXO (pins 6 & 7) and the mixer (pins 4 & 5 for the output using a tuned network at 19.1 MHz). A single 2N3904 is used as the oscillator and fed into pin 1. This approach was successfully implemented into a 2 watt SSB transceiver that measures 2' x 3" x 5". With the 6.176 MHz crystal the tuning range is 14.215 MHz to 14.155 MHz. Oh should mention that the VXO is on a small board that is about 1.25 inches by 2 inches. An ultra-miniature board mounted relay is used to switch between the heterodyne crystals. A panel mounted switch controls the relay which gives two ranges. If two relays and two switches were used then there is a possibility for three ranges. The VXO gives you the ability of frequency agility with the added bonus of frequency stability! You can see the 20M Shirt Pocket Sized QRP SSB Transceiver in action here http://www.youtube.com/watch?v=qYxF_2Pk6go To assure that posts contain information pertinent to EMRFD the 20M XCVR uses two bilateral stages that are a direct lift from Figure 6.110 and use of this circuit really contributed to shrinking down the size of the transceiver. 73's Pete N6QW PS I believe when I was running my initial frequency study there is a 6.0 MHz Microprocessor Crystal and if used in the crystal switched Heterodyne VXO that would place the signal (same IF frequency) down on the low end of 20M CW --so that would give rise to possibilities for a Shirt Pocket Sized CW & SSB transceiver using a VXO. [Non-text portions of this message have been removed] Hi Tim: I've factored a number of crystal frequencies and put them on a web page linked from my crystals page at: http://www.prc68.com/I/Xtal.shtml#Freq http://www.prc68.com/I/pdf/Crystal_Freq.pdf In addition to the factors there's also info on the nominal application. If you know of applications for other frequencies please let me know. Have Fun, Brooke Clarke http://www.PRC68.com > 1a. Ham band freqs from microprocessor crystals > Posted by: "Tim" timshoppa@yahoo.com timshoppa > Date: Thu Mar 17, 2011 3:32 am ((PDT)) > > EMRFD usefully points out some crystal-controlled or VXO-type rigs that use digital-logic-divide followed by simple harmonic multipliers. This is particularly appealing because the divider outputs are so rich in harmonic content (especially odd harmonics), and it affords a lot of isolation. > > In the spirit of this, I took the list of 1-50MHz crystal frequencies (largely "microprocessor type") from the Mouser catalog and looked for simple m/n ratios that end up in what I consider desirable parts of the HF CW bands. My taste in HF CW frequencies won't match everyone's - I seem to be excluding some of the former novice bands - and I seem to have included some swaths that are largely PSK31 or other digital modes lately. I find no shortage of possibilities (note that not all multiply/divide numbers are as easy as others... an even divider followed by an odd multiplier is perhaps most desirable but I list the others because they seem not too much harder). > > The leftmost column is the Mouser-catalog-crystal frequency, followed by the multiply/divide ratio, and then I give the CW-ish ham band frequency that results. It is sorted by ham band frequency: > 77503 Hz is for exact 77500 Hz operation in a TRF receiver. Such circuit needs other load specs (not the 6 to 13 pF for such frequency and form) than typical for XTAL manufacturers interest. In the end this is seen as a different nominal frequency. It is not for CMOS inverter oscillator circuits made. - Henry -- ehydra.dyndns.info Brooke Clarke schrieb: > Hi Tim: > > I've factored a number of crystal frequencies and put them on a web page > linked from my crystals page at: > http://www.prc68.com/I/Xtal.shtml#Freq > http://www.prc68.com/I/pdf/Crystal_Freq.pdf > In addition to the factors there's also info on the nominal application. > If you know of applications for other frequencies please let me know. Over the years there have been various lists of what can you get from microprocessor clock rocks. Other rocks that are common are old CB rocks (some are third overtone other are half frequency and various IF offsets with .455 being most common. Generally there are a huge assortment of standard frequencies that can be summed and differenced to hit a ham band often in a good spot. There are also many that if divided by two or three and summed with another will hit a desired spot. You can also divide by two, four or 5 and multiply by 3 or 7 to get some choice as well as mix the result for yet other possible cases. Using small integer divides are easiest as that is usually one package The 74HCTxxx series (390 has two /2 and two /5 and the 393 is two divide by 2/4/6/16). They are cheap and available in TTL and CMOS technologies. Add to that all of the crystals can be used in a VXO to pull them about. For example 5mhz will usually go +3 -5 from nominal without effort and can go further. A 22.118 will easily go +5 -15khz. I built a VXO 6M SSB rig based on 9.6mhz crystals for the filter and carrier osc and VXO at 13.5175mhz times 3 to hit and cover the most active part of 6M (50.123- 50.170). The 13.517 is a comm