I currently have a signal generator which I am pretty sure is accurate. I would like to get a frequency counter to insure its exact accuracy during alignment though.

Here is my question:

Do frequency counters have to be calibrated as well? If they do, wouldn't I just be chasing my own tail in terms of accuracy and calibration by buying anything other than new or used but calibrated?

EDITED

Yup, they have to be calibrated as well.....

Ideally, the higher end lab grade ones usually have a 5 or 10Mhz input that is used as a time reference...

The trick is to get an accurate reference... Some sort of GPS 10Mhz DO (disciplined oscillator) is usually best as a home lab reference...

http://ve2zaz.net/GPS_Std/GPS_Std.htm

Is the one I'm thinking of putting together.... It outputs an (allegedly) accurate 1/5/10Mhz signal.

David

For calibration work, I use a previous generation of this Frequency Reference.
Very impressive specs at a modest price make it a bargain.

http://www.aade.com/newfreq/new%20freq.htm

Easy. You will need a receiver for WWV/H on SSB and a free programme called "Spectran" - your counter oscillator will have to be on one of the WWV frequencies. Set up Spectran like this....



...except I like the 'waterfall' to be vertical.

Tune in the WWV so that you get a tone about 1kHz - if you have a digital receiver, even a little Degen 1103, then tune to 14.999kHz - then bring your counter close to the radio so that you hear the counter oscillator as well.

Now connect the audio from the radio into "Spectran" and you will see two audio signals, one is the WWV tone and the other is your oscillator. Adjust the counter oscillator tweaker so that the two lines coincide. Watch for a while so that you can assess how the counter drifts and adjust to minimise.

The accuracy of the receiver or the sound card in your computer has no effect as both signals are going through the same path from the antenna to the display so ideally your counter oscillator will now be on the same frequency as WWV.

What make/model of frequency counter do you have?

What does the manual say about the accuracy of its time base (AKA crystal oscillator)?

A frequency counter is inherently tens, if not hundreds of times more accurate than the dial markings on service grade generators, and if its oscillator is off by more than, say, ten cycles at 1 MHz it would be in desperate need of calibration.

If you have a receiver that can receive WWV signals, its an easy proposition for you to calibrate your own counter to an accuracy of significantly less than one cycle per second.

Older lab type counters are quite accurate, having temperature controlled ovens, and can be had for not much money.

I haven't bought a frequency counter yet but I see Ramsey CT-50 for sale. The Signal generator I currently have is a Leader/Starkit LSG-10

Some counters have a tweak across the crystal. I used to set mine with a wire to pickup the horizontal
flyback pulse from a color tv that was tuned to a network color show.

If it read 15734 it was good enough for me.

Those counters are no longer supported by ramsey, and if the postings on other sites are any indication, documentation is apparently difficult to come by as well. I strongly advise against buying something like that unless its only ten or fifteen bucks.

Do not calibrate a frequency counter against WWV. The propagation variances make the frequency too "iffy" for any decent counter, other than for ballpark checks. For most counters, tweaking the oscillator will rarely result in a change in the display as the oscillator is more accurate than the display can resolve. Now, the cheapie imports .... well, they're as iffy as WWV.

Consider that the dials of even the best non-digital receivers won't resolve better than 1KHz and family-duty radios are shots in the dark on the dial. The counter is so much more accurate than these receivers, even uncalibrated for several years, that you'd never see the difference.

Now, if you have a WWVB receiver, that's another story altogether.

It depends on what make/brand/type/model counter you have but each of my 3 counters use a crystal that can be zero-beat with WWV and have a calibration output jack with which to do this. I usually do this at WWV's 10 mHz frequency. I agree that propagation can affect extreme accuracy but not enough for our purposes here.

Dave

An HP117 WWVB VLF receiver/phase comparator, driving an HP105 crystal oscillator, modified, with a newer HP10844 TCXO. I mated the two by disabling the 117's original phase locked oscillator, and substituting the 105's new TCXO.

Adding the newer time base required some level shifting in the 105 to handle the phase lock error signal from the 117, and added a divide by 2 to bring the TCXO's 10Mc output down to 5Mc to drive the original phase locked dividers in the 105. I saw no valid reason to replace them. The only operational difference for the 105 is that the original 5 Mhz output BNCs are now 10Mhz.

If you are very careful about it, you can get a 10 MHz. timebase within at least a couple Hz. of WWV. That will not suffice for lab grade but for hobby grade work, especially at the lower frequencies, it is way plenty enough accuracy. You do not state what you are gonna use the signal generator for, but for working on consumer stuff like AM/FM radios, that is plenty enough. I have used that technique with a good receiver in the past, and the results when later checked against a standard, bear out that the results are correct for hobby purposes. The problem is, that some counters do not have a temp. compensated standard, or an oven controlled standard. Therefore you can set it today, and next week it might drift off a few Hertz. Temperature is at work there, largely. As well as power regulation to the oscillator. Still, if you are aligning an AM radio, being that the error decreases as freq. does, having your signal generator off by a couple Hz. is of no concern whatsoever. You must have a good strong signal on WWV of 10MHz, preferably with little or no fading or interference, and you have to get the oscillator output of the 10 MHz. oscillator so that it does not overwhelm the radio signal, but is strong enuogh to make a beat note you can hear, or see on an S meter, the S meter will fluctuate back and forth within a few Hz. of WWV and when you get it right on, it should only change slowly. You can see the effect as you adjust the osc. trimmer across WWV on either side. Or you can do like I did, spring for a Rubidium standard on ebay. Or a GPS freq. standard.

The notion that an antique radio collector who simply wants to check the dial accuracy of a Stark LSG-10 kit signal generator needs a rubidium, cesium, GPS, or WWVB frequency standard for his frequency counter is ludicrous. Like suggesting that somebody who wants to measure a window for curtains needs a laser interferometer to check on the exact length of his Stanley tape measure.

The short answer is, yes, frequency counters are calibrated devices. In fact, one of the main differences between a high end counter and the cheaper models is the quality of the reference oscillator. The more you pay, the more stable it is, both short term and in the long run. If you buy a new counter, it should be calibrated "out of the box," and that calibration should hold for a long time. Older, used counters can vary, both in terms of the quality of their references, and how they have been handled. A counter that came from a temperature and humidity controlled lab environment will almost always be more accurate than one of the same make and model that spent its life bouncing around in the back of a two-way radio service van.

But for your application, there is no reason why you should need to worry about the calibration--or recalibration--of a frequency counter unless you buy an exceptionally shoddy, abused, or ancient unit.

Yea, but I would still want to check it once in awhile, even if it is only just seeing how closely it beats against WWV. After a substantial warmup, of course.

I suppose I must clarify my answer a little, perhaps "best" was a poor choice of words..... maybe "most easily available, with high accuracy" would of been better? (I don't like playing the ebay game. (Unfortunately) patience isn't always one of my virtues.)

Chris is correct, a GPSDO is highly overkill for a frequency standard.

But the question was, do you need to calibrate frequency counters, and the answer is yes, you do. To do that you need some sort of frequency standard.

Didn't like the idea of trying to beat against WWV, something just didn't appeal. I guess it didn't seem accurate enough.....

My rational for the one I linked is I've been looking around (for quite some time) for some sort of frequency standard to use at my home... That one looks like one of the easiest, and cheapest to build ones I've seen in a very long time. The OCXO I'm going to use with it cost 25$. The pulse-per-second GPS receiver 20$. And <admittedly this is a swag> it looks to be about 50$ in remaining parts.

One of the things I like best about it is, it has no adjustments. You have to modify the circuit slightly to set the voltage range based on the tuning range of the OCXO, but that's it. All the parts are through hole, nothing terribly esoteric is used that's no longer available. Easily built, IMO. Matter of fact, I'll be building one here "real soon now", since I just received the GPS receiver today.

I suppose I could of bought the AADE frequency standard, but where's the fun in that?

At least that's my story and I'm sticking to it......

Cheers.

David

Just for fun, how many frequency counters do you think were in radio and TV shops in the 50's and 60's? How accurate do you think the signal generators they used were?

RRM

How many radio shops were working on anything besides consumer junk?
You see a lot of Collins gear in those old shops? Just out of curiousity...

Quality, surplus test gear is cheap these days. Sixty years is a long time.

Pete

Now that you mention it, what kind of equipment did the manufacturers of high end gear use in their labs in those days? They made some really good stuff without a digit in sight. It would be interesting to find the specs on some of the equipment they used for cal and alignment.

I remember seeing some of the old military gear used for frequency measurement but never saw any data on the accuracy or how it was calibrated.

RRM

BC221 perhaps?

I don't know too much about the radio shops of the 1950s or the '60s, as that was a little before my time, but I do know that by the late 1970s and early '80s, frequency counters were starting to become common in the TV repair business. They were needed for fixing certain VCR and laser disc player ailments, and they also came in handy for troubleshooting chroma demodulator circuits in color TVs.

It is interesting to note that the NTSC color subcarrier frequency, 3.579545 MHz, had to be maintained +/- 1 Hz at the transmitter, per FCC spec. This works out to about 0.3 PPM. We found it best to use counters that were accurate to 1 PPM for TV service work. If the color subcarrier frequency is off by more than about 3 Hz either way, it could cause noticeable shifts in picture hue.