What are you using for an inexpensive frequency counter?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Member
Joined 2011
Paid Member
Oops, I neglected to point out that the surprise comes from the nonlinear, polynomial relationship between LC resonant frequency and inductance: a 2% change in inductance does not result in a 2% change in frequency. Luckily, very high precision frequency measurement is commonplace.
 
Administrator
Joined 2004
Paid Member
Hi Mike,
You can't just look at the number of digits in the display. You have to look at the specifications and work out an error budget. It is not uncommon for the meter to have a LSD that is meaningless as the error is far greater. I spent time calibrating meters for certification and what you learn would be a real eye opener.

-Chris
 
Administrator
Joined 2004
Paid Member
For measuring T/S parameters, for viewing waveform distortion with a scope, for playing around with filters, that sort of thing

Since this was the original question - for T/S i recommend investing in a Woofer Tester II or a WT3 (now called DATS). Quick, easy, repeatable and you can save or export the results for use in other software. For the other frequency measurements I just use my soundcard and software which works well for AF filters. A DVM with counter might do as well. I know these guys love their HP counters and all, but it does seem overkill for what you want to do.
 
Administrator
Joined 2004
Paid Member
Hi Pano,
I am on record for recommending the built in counter function in the HP 34401A first, and after that the HP 5385A with a view to using it for other tasks other than just this. It's a fast counter that can be had sometimes for $110. I just bought one at that price a couple months ago.

My only caution with using a sound card is the lack of input protection. If you have a jig that provides that protection, then by all means, use the sound card and application program.

-Chris
 
Another typo corrected.

No you are off by a factor of two pi.

Yes, that was the radian frequency, my slip.
It doesn't substantially alter my conclusion because my point was based on the fractional variation so it cancels out anyway.
The frequency moves down to around 12.5 MHz so the +_1 at the end matters more.
You could even contrive an example with the frequency of 10.01 so it's effectively 3 places.
But you said "five or sometimes four" so I assume 4 reliable places is a realistic worst case.

... the surprise comes from the nonlinear, polynomial relationship between LC resonant frequency and inductance:
For small variation it's practically identical to linear, for your example of 17.0 nH to 17.1 nH the variation is about 0.6%, the non-linearity is on the order of 0.6% of 0.6%.
I don't think a tiny 0.0036% non-linearity substantially alters my conclusion.

a 2% change in inductance does not result in a 2% change in frequency.
I already noted in my previous reply that's it's very close to half, so a 1% shift, that does push the requirement for precision up - but not very much.
That was more or less my point.

Best wishes
David
 
Last edited:
My scope is a TeK 2213A and does not have cursors.

I had wondered about those cheap crystal kits like Phil mentions. Chris says "They aren't accurate, and they aren't terrible either. But, as a frequency counter I would trust them. "

If I used Microsoft as an OS, I might just use a soundcard solution as several have recommended, but I use Linux mainly, and find that using the sound card is cumbersome and error prone (at least in my experience). I just have not had good luck playing with sound cards using Linux.

I do plan to use REW and have ordered a DATS that I will use with an ancient Windows 7 laptop that's not being used for anything, hoping that will work. But I really dislike using Windows software, and wouldn't enjoy or want to rely soley on a Windows software based solution.

A good handheld DMM (such as a Fluke) may be adequate for my needs, yet they are priced almost as much as the used HP benchtop counters, and I am mulling that over. If I could get a good working HP 5385A for $110 like Chris says, sure, that would be a better bet, but that seems a rare find. I do like the idea of an HP5385A or HP34401A, but the cost is a factor. (so might be bench space).

I _should have said initially_ that I am currently using an old Beckman FG2A for a signal source- very sloppy/inaccurate dial, which is why I need a counter. So I'm actually considering Gerhards suggestion of "a generator with a reliable frequency display. HP3325 A/B, for example". But for the cost... the cheap Beckman with an HP bench multimeter or 5385a would be a much more versatile toolset.

I'm inclined to go with something like an HP5385A, if I can talk myself into being happy with spending that much money on it.

Thanks for all the great discussion!
 
Administrator
Joined 2004
Paid Member
Hi Keith,
Just keep watching. The good HP 34401A tends to go from $450 to $700. I bought one for $600 for a friend. I'll tell you one thing, they are by far the best bench meter I have ever had the pleasure of working with. The Fluke handheld is a good meter and I use meters like that (HP 974A) along with the 34401A. That meter is extremely accurate and is packed with features you would actually use. I can't afford the current Keysight meters, but would really love one!

If you had to buy one instrument for your bench, I would highly recommend it to be the HP 34401A. You will get an amazing amount of use from it. I bought mine when they first came out, just introduced. It has never been out of tolerance during that time.

If you use a counter for higher frequencies beyond 300 KHz, you would get great use from the counter as well, but right now the 34401A answers all your current needs and needs you don't even know you have. I can't stress it enough how much you would come to love that meter.

-Chris
 
On calibration, I'd love to have a GPS disciplined oscillator, but even $200 is too much. If you can get WWV at 10 MHz, just tune it in on any cheap shortwave. Next, stick a couple inches of wire into the 10 MHz BNC on the back of the counter. Just enough to contaminate the WWV signal with the radio maybe 10' away. Now adjust the counter oscillator so it zero beats with the radio. You can usually get within a Hz or so with that method.
 
@Chris- Just looking over some eBay listings for the HP34401A I see this comment:

"Beware of some older models as the display processor is not obtainable and there is no replacement available. This is a fairly common issue. You basically have to buy a parts machine with a good display board to get it back up and running. The boards from newer units are not compatible with the older units even though they look the same."

How does one distinguish "newer" models? Is there a serial number guide? Feature difference?

@Conrad- yea, that's a good tip about using WWV.
 
Administrator
Joined 2004
Paid Member
Hi Keith,
Well, my original one is going to be the old style. The newer one I bought has an RS-232 port, so it is the newer type. The one I got for my friend will also be the newer type.

I have only ever heard of one with a bad display and hopefully it is on it's way here. I'll be able to see for myself. These meters have only the main PCB and a display PCB I gather. So its not a surprise to find out that adding another port would make the main boards incompatible with each other. The extraordinary long model life of the 34401A then explains why they may have changed components and boards (features) over the life of this unit. Whenever they were introduced is how old mine is. It's display is getting dim due to the high number of hours it has on it. I should really try to buy a replacement display if I still can. This meter may well run another 20 years or so!

To answer your question, the existence of an RS-232 port might indicate the newer model. Could be wrong here as there is ample time to make more than one major change in the design of this instrument.

I think the only other model with a longer model life would be the 3458A Lab, 8 1/2 digit DVM. They are accurate enough to use as a voltage asset in a calibration lab. We had one and I used it many times. I would love one just because instruments are my thing and I could use it to keep my other meters in calibration. It isn't suitable for a service type bench meter though, that crown goes to the 34401A (which is an industry standard multimeter). You would also find these in calibration labs, sometimes as the most accurate instrument they have.

The only meters better than the 34401A are the meters that replaced it. More features, faster and higher accuracy. Those kinds of things.

-Chris
 
I'm looking around for an inexpensive frequency counter to use for audio experiments, and hoping for some recommendations.

Keith, so if you are looking around for a simple thing to use you might try one of these.

I have the older version and for audio related stuff it is just fine. HP 3551A. The guy wants $29.95 LINK I got one for $20.00

Here is an overview of it from HP Journal: LINK

Either will work to get you started and it sends and receives signals too.I like old dial stuff it has a digital read out. It goes from 40Hz to 60kHz.

You can't beat it for the price, it is a real HP.

Do a search for a free manual to trouble shoot it.

I bought mine just for the knobs for my HP339a, but liked it. Heck these have battery power oscillator in it for low level signals too.

When you want something more save and buy it knowing you'll also have your HP3551A for back up and checking out your basic audio gear.

Cheers,
 
Hey Sync- Thanks for the links! Hah! From the name I would never have guessed it could be used as a sig gen. I went ahead and bought it just from your recommendation... Keysight actually has the manual on their website. I figure if I can get it to work it will handily replace the old Beckman I have, and if it doesn't I'll have an inexpensive toy to work on for several hours of entertainment while I look around for something more appropriate. I'll let you know how it turns out...
 
That was fast, I was almost thinking about buying it after I mentioned it to you,
as I can use the knobs though I waited to see if you'd pick it up.
You can use your it with your Beckman for a cross
check and vice versa.


This should appeal to any of us that are frugal, cheapo, tightwads, etc. These
just work and they do a lot of things any of us audio aficionados want to do.
Signal tracing made easy with built in speaker on battery power.

Find the noisy part of the circuit. See how well that crossover is working.


The HP 3551A will get you more than half way there with just about anything you want to do.
For the other half, there is the more expensive good stuff that everyone here recommends.
I posted about this unit before and for $30 being a "transmission test set"
I think most thought I was going to use it for my automotive transmission. :confused:

If one were so inclined I'm sure the weighting filters could be modified for different use etc.

:cheers:
 
Hey Sync and everyone- I bought that eBay HP3551A, $40 including shipping. When it arrived only the receive function worked, not the send. After some sleuthing, I discovered (with help) the problem was the freq vernier control wasn't working properly. It is a special dual stage pot which is used both for fine and coarse adjustment- unobtainable. Fortunately I was able to break it open with little damage in order to clean it up (de-oxit) and make certain the contacts were mating, and now it works like a charm! So now I can use it for accurately measuring and providing 40Hz to 60KHz signals for experimenting- much better than using that FG2A, and a lot less expensive than the nicer HP toys discussed above (which will wait until I can get more $$). As Sync sugggests, I will also be able to use it for signal tracing and other things ...
 
The HP3551A is rated for an output between 40Hz and 60KHz, I can measure it being stable down to just below 30Hz, and slightly above 65KHz.

The Beckman FG2A is rated for 0.2Hz to 2MHz, but I have not tested it to those limits.

Mine came with original probe sets in like new condition and a power cord. That one in your last link ($400) doesn't come with a power cord, "may need calibration", being sold "as is".

@Chris- I haven't measured distortion, and probably wouldn't know how to do it properly (yet), but on the scope the HP3551A signals look good all through its range.
 
Last edited:
Administrator
Joined 2004
Paid Member
Hi Keith,
Well, at some point, pick up a distortion analyser. I was curious if the HP unit had lower distortion than your function generator. Function generators typically have distortion measuring 1% to 0.5%, while other oscillators can be quite good. The HP probably has a little less distortion than that.

Looking at an oscilloscope will only show gross distortion well above 1%. For non-critical measurements and FM radio work, the old THD analysers from HP would do very well. Look at the 331A to 334A THD meters. The 331A and 332A are fully manual, the 333A and 334A have automatic nulling. You would use some sigal source with these meters. The matching oscillators were 652A and 654A, you would rather use the 652A. In case you see one of these cheap, HP made systems with both pieces in one (THD meter and audio oscillator together). These would be a 339A (which I use) and the 8903A or B (which I would like). The 8903A tend to be expensive, but they (339A and 8903A) are very, very good. There are other makes out there, but the Japanese brands like Leader and Kenwood aren't nearly as good as the HP branded stuff.

Best, Chris
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.