Hello Bingo, thanks for the question!
Yes, time-nuts has been totally invaluable resource; I am just scratching the surface of my learning there (which I joined initially to learn about GPSDO which would be used to calibrate this ^*#&$*&!! frequency counter.)
I actually came back here to post an obsevation about the counter, now that I have put it in service.
It appears the counter cannot resolve the least significant digit as expected (tested when counting up to 100 MHz signal) any closer than +/-6 counts. For example, if I seek to measure 99.000000 MHz (from a synthesizer), as I adjust the synthesizer up slowly (Hz by Hz), the counter first shows 99 000 000, then 99 000 006, then 99 000 013, then 999 000 020, etc. This is very undesirable behavior, because it says my "8 digit" counter is really 7 digits useful.
As for the wrap problem (which is only seen in "totalizer" mode I use with the circuit described, for setting the reference oscillator), the reference to a 7 second gate is with purely external circuitry I built for the purpose of calibrating the thing. I use an old Motorola ONCORE UT+ GPS module which has 1 PPS output to drive a divide-by-7 circuit (74HC163 configured appropriately), then that output into a 74HC74 (half of) to get 7 seconds high, 7 seconds low. I then routed the 13 MHz reference oscillator (the 26 MHz OCXO module I installed, with 74HC74 dividing by 2 at the output to get 13 MHz) from the output of the counter back panel reference oscillator output jack, through an AND gate (74HC08) gated by the 7-second-hgh GPS-derived signal, into the counter's B channel (lower frequency) input. When the gate opens, the counter counts from 0 up to (ideally) 91 000 000 counts (13 MHz * 7 seconds), then stops counting for 7 seconds, then the cycle repeats. One of the counter's front panel buttons lets me reset the totalizer back to 0 (which I do during the 7 seconds between count-up periods). I adjusted the bias voltage on the 26 MHz OCXO module (divided by 2 inside the counter before going in where the original piece-of-junk reference oscillator existed) until I get 91 000 000 counts. This adjustment does appear to have 1 count resolution (unlike the frequency counting mode) so if the counter part worked as I expected, I'd have reasonable accuracy; as it stands the accuracy is ~6x worse than I expected.
Where the "wrap" problem happens is this - I originally had set the gate period (of my external circuit that connects counter's refernce output to counter's totalizer input) to be longer - 10 seconds - in which case I expected to see 130 000 000 (10 seconds at 13 MHz), with the first digit absent (so I'd really see 30 000 000 if properly adjusted). I was very puzzled at first seeing a random variation of a few thousand after each cycle, which I initially attributed to using the wrong (trailing) edge of the GPS module's 1 PPS output (which I was). Once I fixed that, and continued to see the random variations, I eventually discovered that as the counter crosses from 99 999 999 to 00 000 000 (and does "overflow") in totalize mode, the microcontroller is evidently busy doing other things - perhaps lighting the overflow indicator - and misses several thousand counts. Given the large number of simple ICs (counters and otherwise) inside - although no schematic - I had expected (wrongly) that the microprocessor was just doing configuration and display management; as it turns out it is involved in the counting itself in some way. I have no plans to try to re-engineer it.
Of course, this is a very low cost benchtop counter (I think I paid about $45 US for it, plus about $5 I have in the OCXO module bought a few years prior), and I can calibrate it (such as I have tried) with $13 invested in a GPS module and a few glue parts, so it's adequate, but not the home run I had hoped for in the first place. Given the newly discovered problem with least significant digit, I'd now dissuade anyone else thinking about buying this counter (which is still available on ebay).
I certainly didn't take the most cost-effective route to a counter with use of GPS to adjust ref oscillator, but I've learned about GPS modules which I had no prior experience with, and in the end that makes the experience worthwhile.
Home
Help
Search
Login
Register
