Rigol DG1022 Function Gen Frequency Counter Function

[xrunner]

I was playing with the frequency counter function of my Rigol DG1022.

The manual states:

"Counter with high accuracy and wide frequency band can measure frequency, period, duty cycle, positive pulse width and negative pulse width of the signal with the frequency from 100mHz to 200MHz (single channel)."

When I feed it a 10 MHz signal using my own external GPSDO, which outputs a 10 MHz signal to less than 1 Hz accuracy, the counter on the DG 1022 has on it's display (using it's internal reference):

9.99995 MHz

I read that as 50 Hz low

If I switch the DG1022 to an Ext. 10 MHz reference, it shows on the display:

10.0004 MHz

which I read as 400 Hz high.

I obtained the calibration manual for the DG1022 and performed the calibration for the internal and external frequency generation. The generator outputs 10 MHz to within 1 Hz, but the frequency counter yet still reads the same way. There is no specific calibration step for the frequency counter. As far as frequency is concerned, there are only the steps for the Frequency (Ext) Adjustment and Frequency (Int) Adjustment.

I am not sure what is going on with the DG1022 frequency counter. It makes little sense to me. It seems to me it should be as accurate as the frequency generation, but it isn't.

Anyone got any ideas on how to make sense of this?

[xrunner]

This is interesting. I fed my DM3058E (which the manual states can measure up to 1 MHz) and my DG1022 (which the manual states can measure up to 200 MHz) both the same 1 MHz sine wave from a Marconi 2022. This 1 MHz signal is known to be accurate to better than 1 Hz using known accurate frequency counter locked to a GPSDO.

The DM3058E frequency counting function is closer than the DG1022's <-- a unit expressly dedicated to generating periodic waveforms! 

For some strange reason, the frequency counter in the DG1022 is not as good as the one in their DMM. It generates accurate frequencies, but cannot count them as well.

This is fascinating to me. I'm going to contact Rigol and see what they have to say about it. 

[jpb]

I think there has been another recent thread on this. I also have a vague feeling this came up in conjunction with the Rigol D4000 series some time ago. There the issue seemed to be the fact that the counter was a very simple binary affair which gave rather odd gates and some rounding issues but I can't remember the details.

I had a quick look at the manual which doesn't seem to state gate times and so on. It gives the accuracy at 6 digits/sec. I think the generator's frequency counter function is less accurate because it is simply an add-on, I don't think it was designed in. The accuracy I would guess is +-2counts at best. Though the DACs clock frequency is quite high I suspect that it is scaled up and that the counter clock is lower.

It will be interesting to hear what Rigol say.

[_Wim_]

I think there has been another recent thread on this.

Yes, see https://www.eevblog.com/forum/testgear/rigol-dg1022-frequency-counter-error-(to-a-lesser-extent-the-1054z-as-well)/msg868135/#msg868135

[xrunner]

First - thanks to both of you for entering the thread. I did go back and read the thread you referred to. Mine acts the very same way with the exact same results. I have opened a ticket this morning on the Rigol NA website - let's see what they have to say about it. It certainly doesn't appear to be a "Counter with high accuracy ..." like stated in the manual. 

But ... here's another weird thing I noticed too. When I lock my Marconi 2022 RF generator to my external 10 MHz GPSDO and set it to output 10 MHz and look at both waveforms on my scope - the output never drifts in relation to the 10 mHz reference.

This is not what happens on the DG1022. If you watch it closely over a few minutes - the 10 MHz output from the DG1022 will drift over time in relation to the reference it's supposed to be locked to. I know it's using the ext reference because if I disconnect it, it gives a message and the output changes to it being very obviously a frequency multiple-Hz different than the reference. I'd like someone to try that themselves. I do not understand that at all. 

[xrunner]

Update:

I called Rigol NA and spoke with a very pleasant gentleman - Jason Chonko. He was very well spoken and was interested in this issue. I spoke about the discussions of this on this forum (he was very familiar with the EEVBlog forum   ) and pointed him to this thread and the other one regarding the DG1022 frequency counter issues.

He also asked me to send him a screen shot off my scope of the drift occurring at the output of the DG1022 when it is generating a 10 MHz sine wave, using a GPSDO 10 MHz ext. reference. I simply put the scope in infinite persistence mode and recorded the drift for 60 seconds (although it never stops drifting). I have attached that to the post. We will see what happens now.

[mstoer]

I have a DG1022 and a DM3058E as well.   I set the DG1022 to 1 MHz and fed it to the DM3058E as well as it's own counter. 
DM3058E: 999,998 kHz
DG1022 Count (CH2): 999.998 KHz (last digit toggles between 8 and 9)

I do not have anything to test the external reference  with at the moment.  Well, I have an ancient Vectron 10kHz crystal oscillator module, but that is it.

Marcell

[xrunner]

I have a DG1022 and a DM3058E as well.   I set the DG1022 to 1 MHz and fed it to the DM3058E as well as it's own counter. 
DM3058E: 999,998 kHz
DG1022 Count (CH2): 999.998 KHz (last digit toggles between 8 and 9)

Thanks Marcell.

Can you post the version of firmware for your DG1022 please?

[mstoer]

The firmware is 00.03.00.09.00.02.11 on the DG1022.

After 30min + of uptime, the DM3058E now reads 1MHz exactly and the DG1022 counter remains the same.

Marcell

[xrunner]

The firmware is 00.03.00.09.00.02.11 on the DG1022.

Interesting, the firmware in mine is -

03.03.00.13.00.03.00

Very good info, I will forward this to Jason Chonko. 

[xrunner]

Well, the firmware situation seems a bit of a mess out there ...

Marcell has -

00.03.00.09.00.02.11

My version is -

03.03.00.13.00.03.00

And from this official Rigol website,

http://int.rigol.com/Support/SoftDownload/3

we have this for the DG1022 -

00.03.00.00.01

And yet from this official Rigol website for the DG1022 -

http://beyondmeasure.rigoltech.com/acton/form/1579/001e:d-0001/0/-/-/-/-/index.htm

We have this version listed as the official version -

00.02.12.00.00

I'm writing Jason right now to see what the actual real official latest version is supposed to be ... 

[xrunner]

*** Update ***

After several back and forth emails to Jason @ Rigol, they have decided to send me a DG1032Z demo so I can see if the same issue arises, but they said since it was a different design, it should operate correctly as far as the external clock and frequency counting accuracy. I'll have two weeks to test it (more than enough time). Then, it seems I may have the option to "trade in" my DG1022 for such a model.

They really stepped up to the plate. I'll post some pics/info when I get it. 

[xrunner]

The DG1032Z demo unit arrived from Rigol/Jason. Will be checking it out, specifically the external clock and frequency counter as compared to the DG1022.

[xrunner]

Well, it works the way it should, which is not the way the DG1022 did. First the frequency counter. When using an external 10 MHz ref, and measuring same, it shows the exact frequency. This is as expected, since it's a circular measurement, but the DG1022 could not do this.

Second, measuring an external 10 MHz signal, when NOT using it as a reference, gives a reasonable readout of:

9.999 989 9 MHz

which is off by ~ 10 Hz

Very good.

Next I'll see if the output drifts as compared to the ext. ref. it's locked to.

[xrunner]

Test #2. The generator output does not drift when using an external source. I'm using my HP 54501A project scope here set to infinite persistence. No drift noted.

Ch1 - Ext. Ref. 10 MHz
Ch2 - DG1032Z output 10 MHz

[xrunner]

Lastly, a check of the DG1032Z upper frequency counter limit - 200 MHz - looks good.

Have to see if I can keep it now. 

[rstofer]

Even though the 1032 is a couple of hundred bucks more than the 1022, it may be worth the extra money. 

I also like the part where Rigol participates in "Bring your child to work" days and allows the kids to design the front panel graphics.  Seriously, a multi-colored button?  I'll get over it...

Thanks for the thread, it has brought up some interesting points and if I decide to buy an AWG, I might just opt for the upgrade.

[xrunner]

I also like the part where Rigol participates in "Bring your child to work" days and allows the kids to design the front panel graphics.  Seriously, a multi-colored button?  I'll get over it...

Yea, it's a little too decorative for me, but hey, it's works better than the Dg1022. This model displays the "EXT" when it's using an ext. standard, whereas the DG1022 did not show anything on the display to let you know that, which I did not like.

But one thing that I still do not like is the way it acts when you select an external standard, and then for some reason it loses it.

What it does -

a. You select an external standard in the menu
b. If it finds one connected, it uses it, and displays "EXT" on the front
c. You turn the unit off, and back on again (with the external standard still connected) and it uses it again upon startup

(so far so good)

d. turn it off again
e. remove the ext. std. and turn it on
f. it displays a message that the ext. std. is gone and uses the internal std.

(still OK so far)

g. re-connect the ext. std., turn it off and on again
h. it does NOT attempt to look for the ext. std. this time, but uses the internal std.

(not good)

What should happen (IMHO) is that if you explicitly select an ext. std. in the menu, it should, from that point on, remember your preference, and always look for an ext. std each time upon power-on, and then default to the int. std. if it's not there. I brought this to Jason's attention on the phone regarding the DG1022 and he seemed to think I was probably right.

Anyway, I want to try and keep this, even though it's used. I'll see what kind of deal he can do for me. 

[xrunner]

Just got off the phone with Jason. They offered to let me return my DG1022 and receive a discount on a brand new DG1032Z. Very happy with their attention to this issue for me!

 

[BloodyCactus]

I have a 1032z, I remember being able to induce some really odd results from it depending on what precision/scale you pick. In the end I just use my racal dana 1992, which never lets me down.

[ormandj]

Hi,

As the OP of the other thread, you've had the same findings all of the rest of us did in that thread. Glad you got it sorted out, sounds like you went the same route as me - pick up a 1032z. It's a significantly better unit, they should just take the 1022 off the market or at least have a warning regarding the frequency counter being relatively useless.

[elex_enthusiast]

Does the hack fixes this counter issue? I've been eyeing to purchase DG1022z as my first gen. Given the budget constrains and scarce availability (at least here where I live, where most of the gears you need can only be bought online) it is my only option.