Author Topic: Rigol DG1022z frequency counter - is it REALLY this bad for everyone?  (Read 861 times)

0 Members and 1 Guest are viewing this topic.

Offline julianhigginson

  • Frequent Contributor
  • **
  • Posts: 691
  • Country: au
so... anyone who owns this signal generator tried using the frequency counter section?

I've had my 1022Z for just over a year, and while the signal generator functionality has been great for my purposes, I only just started needing to use the counter this week.

I was measuring a nominal 1Hz 0-3V square wave signal coming from a micrcontroller pin, to look at ppm accuracy of a crystal oscillator.  The waveform was showing perfectly fine as a square wave on my 200MHz scope...

But when plugged to the counter, even with careful setting of level and sensitivity and coupling, the counter display would jump all over the place. it was *more often than not* getting an accurate expected value, but not consistently. It'd jump up even to MHz values sometimes, and even when it had the frequency right, it'd have the pulse width measurement way off from what I could see on my scope, which was pretty off-putting.  And the HF supression filter actually made detection stability worse.

Today the thing has stopped measuring completely... lucky my trusty old keithley 2700 has a frequency counter that is measuring to the same ppm value that the Rigol was yesterday, so I can keep working. And I have the option to send the Rigol off for warranty repair...

So what I was wondering, was if anyone else has one of these things that they used as a frequency counter, have they've seen similar silliness as i was yesterday, or if what I saw was completely different to their experience with the counter?

I'm wondering if that behaviour was the indicator of some kind of fault pre-complete-failure, or if it's just like that. Because if they all behave like this normally, its probably not worth shipping off for a warranty repair, and my 1022z can just stay as a function generator for the rest of its life...
« Last Edit: July 04, 2019, 05:29:37 pm by julianhigginson »
 

Offline julianhigginson

  • Frequent Contributor
  • **
  • Posts: 691
  • Country: au
I figure the DG1022z  is a relatively popular bit of gear on here for price/performance as a function generator -  but nobody is using a DG1000z as a frequency counter?
 

Offline WhichEnt2

  • Regular Contributor
  • *
  • Posts: 91
  • Country: ru
I used it several times on audio band and up to 400 kHz, no serious stuff. Sometime it is rather fiddly to blindly adjust sensitivity and level.
If I remember correctly "sensitivity" softbutton have a help page on the instrument that have some tips about adjusting it regarding to freq and level of signal. User manual does not have that info.
Short pieces, high value, small period, huge amount, long delay.
 
The following users thanked this post: julianhigginson

Offline das_strobel

  • Contributor
  • Posts: 8
  • Country: de
I had the same issues when I started to try use the counter. I figured it out eventually. (I think...)

Looking at this, I believe I know what you did.
I was measuring a nominal 1Hz 0-3V square wave signal coming from a micrcontroller pin, to look at ppm accuracy of a crystal oscillator.

I tried something similar. On my DSO I would set the trigger to DC and the level to about 1.5V in this case. This doesn't seem to work on the DG1022z. It worked perfectly after I had set the trigger to AC and the trigger level to 0V, which makes sense somehow.

The so-called sensitivity is the hysteresis between two trigger points. The higher you set the sensitivity the less the signal level needs to drop below the trigger level before another crossing of this level is generating a trigger event.

You can find this detail in the user guide only not in the detailed section about the counter but in the spec section about the counter. Search for "hysteresis" in the pdf and you will find it.

 
The following users thanked this post: julianhigginson

Offline julianhigginson

  • Frequent Contributor
  • **
  • Posts: 691
  • Country: au
thanks, yeah I'm aware of the standard level being the point where the counter input decision gets made, and sensitivity as hysteresis control to filter out noise, from standalone frequency counters in the past.

Also AC and DC coupling is a pretty straightforward concept. I wanted DC as I'm dealing with 1Hz (someone else set the crystal division... I probably would have set something faster, but oh well) and I wasn't sure exactly where the cutoff would be for AC coupling. That said I had tried using both AC and DC coupling. Setting the threshold and sensitivity for each, and well, they're about as bad as each other.

the thing that threw me most was the noise filtering. I would expect some kind of LPF function to be associated with something like that - absolutely do not understand how the noise filter being on would cause worse triggering for a 1Hz square wave that i was getting without it.

The behaviour I'm seeing makes it really hard to visualise  what's happening to my signal inside this unit.. like.. how does it POSSIBLY see 33MHz and 40MHz on the counter input no matter where I run the sensitivity to, when I've also clamped a scope probe to the same signal at the same time and can see what is for all intents and purposes a perfect square wave with no ringing, nothing, easily going straight up and down between 0V and 3V3... even with AC coupling on, if it was introducing sag on the squarewave from inappropriate filter cutoff.. this kind of thing shouldn't cause that kind of result. It just doesn't make sense.

Part of me thinks it's just got to be faulty behaviour, because this level of usability is really not OK even for a supplementary feature... then I think, well, maybe it really just does suck this hard. which is a shame because I have absolutely no complaint at all with the function/AWB generator. it's been fantastic and worked exactly as expected for the last year.

I was really hoping that someone would pop up and tell me about just how easy it was to use for them, like a normal standalone frequency counter.  but not so far...
« Last Edit: July 09, 2019, 05:05:05 pm by julianhigginson »
 

Offline das_strobel

  • Contributor
  • Posts: 8
  • Country: de
Sorry for implying you didn't know all that stuff. Regarding the usability of the counter in the Rigol, I came to the same conclusion in general as you did. While I'm happy with the function gen and AWG part I consider the counter more or less useless.

I played around a bit with it, starting with connecting it to he ch2_out of my Tek 475A (which is especially there for this use case I guess) and looking at the 1kHz calibration signal. After trying to make it work for way too long (don't remember if I actually succeeded) I connected a few known signals to the counter and was able to measure them correctly in the end. But only after analyzing the signal and fiddling around with the controls of the counter. I didn't try very low freq signals though, always used AC coupling.

But of what use is this if you need to connect a DSO to analyze the signal first? When a DSO is connected you can already measure the freq with the built in counter or the cursors in most cases. These days a counter is probably only used for accurately measuring very low freq signals where the Rigol fails. Maybe one rainy Sunday afternoon I will give it another try and play with it. It is not an important use case for me anyway.
 
The following users thanked this post: julianhigginson

Offline julianhigginson

  • Frequent Contributor
  • **
  • Posts: 691
  • Country: au
hey it's all good, I'm not offended - was more trying to explain the level of experience that my confusion with this thing is coming from.

I'm still undecided about getting it repaired.. probably will, when I think I can live without the function generator for a while.. and at least it has a 3 year warranty they seem serious about supporting.

I think that no matter what, I've got a standalone frequency counter on my "to get eventually" list now.

the problem with scope measure functions to measure frequency (at least mine, and any I've ever used) is they don't measure down to ppm. they seem to only really work on the display memory. mine seems able to give me about 3 significant figures. I agree for non-crystal measuring general frequency capturing that's all I ever use.

In the meantime, the keithley 2700 is pretty good, as long as I can test above 3Hz...  It turned out to have an issue with 1Hz counting too... it actually seems to work around that area, but as you approach very close to 1hz (within 15 or so ppm!) it stops working. I looked at the manual and it only claims to work down to 3hz (despite having a mHz display range that is implemented and often working) I guess that spec was a bit of creative problem solving when they discovered their frequency counter feature didn't like 1Hz for some reason...

 

Offline das_strobel

  • Contributor
  • Posts: 8
  • Country: de
Sometimes one can't lay a topic to sleep yet even if there are more important things to do. ;) This is such a case for me.

During a break today I switched my DG1022z on again and connected its CH1 to the counter with 90cm BNC cable. Can your DG1022z do this?


I noticed 2 things:
  • Gate time set to Auto is not a good idea for this scenario. The device is trying to figure out the gate time and the human in front of it to set the right trigger sensitivity. I had manually set it to 10s (1sec worked as well but gave much less accurate results.)
  • The trigger sensitivity setting 50% did not work. 100% did not work either (gave me a measurement of ~136nsec). 75% did the trick. Hysteresis for 0% is 140mV and for 100% is 2mV. I consider 50% to be hysteresis of 70mV and 75% of around 35mV. Still I don't understand why 50%/70mV didn't work either for a 250mVpp signal.

In the end it gave me a bit peace of mind.
« Last Edit: July 09, 2019, 07:25:16 pm by das_strobel »
 
The following users thanked this post: julianhigginson

Offline julianhigginson

  • Frequent Contributor
  • **
  • Posts: 691
  • Country: au
yeah I had to use the >10 gate setting to get anything remotely sensible out of it at 1Hz input. Agree that the auto button is useless (at least for 1Hz digital squarewave from a micro)
 

Online bd139

  • Super Contributor
  • ***
  • Posts: 10219
  • Country: gb
Just seen this. The counter on the DG1022Z is DIRE. It is the worst piece of shit I've ever used. Just getting the damn thing to trigger is hard half the time and mine barely even works when AC coupled even on large signals. I'll quote a comment I made independently about the counter on July 5th:

"If it’s anything like the “counter” that graces the Rigol units then you might as well hire a monkey with an abacus and shove the BNC up his clacker."

I ended up buying a 42 year old Racal counter that was better. That blew up so I bought an HP 5334B in the end. That thing is marvellous. Have retrofitted the 1.3GHz option and use it with an external OCXO from aliexpress.

I'd be annoyed at Rigol if I expected it to be a decent counter but this is an AWG really and it does a pretty good job of that side of things.
« Last Edit: July 10, 2019, 06:51:19 pm by bd139 »
 
The following users thanked this post: julianhigginson

Offline julianhigginson

  • Frequent Contributor
  • **
  • Posts: 691
  • Country: au
Re: Rigol DG1022z frequency counter - is it REALLY this bad for everyone?
« Reply #10 on: July 10, 2019, 08:56:40 pm »
excellent! well that's settled then.. I have no need to go out of my way to get that counter input looked at or repaired.

Maybe in a few years when the 3 year warranty is out, I'll open it up and look in there myself.. but from what you're both describing, your experiences lined up with my experience before it completely stopped working, so I probably saw it working as best as it ever would, which isn't well enough at all.

oh well. Lucky I still really like the function generator channels. I'm glad they made them better than the counter.
 

Online bd139

  • Super Contributor
  • ***
  • Posts: 10219
  • Country: gb
Re: Rigol DG1022z frequency counter - is it REALLY this bad for everyone?
« Reply #11 on: July 10, 2019, 09:02:03 pm »
Yes I'm convinced mine went from working to useless at some point. I might take it to bits even if it's in warranty and have a look.

Incidentally you can crack the DG1022Z and run it up to 60MHz so I'm rather happy even if the counter doesn't work. Details here: https://www.eevblog.com/forum/testgear/need-help-hacking-dp832-for-multicolour-option/msg2471817/#msg2471817 (read up and down the thread - that's just my success reporting)

TCXO in the DG1022Z is pretty good as well. It's always within 1Hz of my GPS standard
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf