rigol DSO814

[epoch_time]

I purchased a new dso814 from amazon about 2 weeks ago.

i've been monitoring 60hz line frequency. The frequency counter shows accurate power line frequency. But the measurement frequency mode is not correct and does not agree with the frequency counter mode display.

The 60hz power accurate line (mains) frequencies are usually between +60.05hz and -60.05hz  with the .05hz(50mil-hz) being the extreme where grid operators expect failures.  Its normal to see frequencies excursions up to .05hz a couple times a day. 

With the DSO814 frequency measurement mode I always see frequency excursions exceeding about .10 hz within 4 or 5 seconds of resetting the statistics. At the same time the frequency from the frequency counter mode are only .001 hz off.  While in the measurement mode (frequency)  after maybe 100 to 1000 0r 10000 counts I will see always maximum an minimum frequencies off +- 100's of millihertz (OR +- .100hz) while the frequency counter mode shows maximums and minimums of about .02 hz (much more accurate).

I've tried every combination of triggering modes, coupling modes, sampling speeds and depths, bandwidths adjustments and every combination of settings there are but consistently measurement mode and counter mode maximum and minimums are way off. I believe this is a simple math bug in RIGOL's programming.

Could someone who own's a DHO814 connect to 120 volt hot mains side with 10x probe setting and 50volt per division and 5ms per division horizontal.
and let me know if counter mode agrees with frequency mode or if either one seem accurate?

Maybe I have a defective unit. Which I would like to send back to Amazon!
my dho814 Firmware 00.01.02
Hardware   12
Build                2023/11/09 17:35:40
Android Build    2023/11/16 11:08:27
Android version  7.1.2
launcher           00.01.06
WebControl       00.01.02

Thx Epoch_time@yahoo.com

[Manul]

I think you're a bit harsh on your scope. That's not such a huge difference. Frequency counter is just that - it does one thing and it does it good. Contrary, frequency measurement operates on sampled waveform data points which is powerful in a sense that it can measure all sorts of things, not just frequency, but it's not perfect. It would probably be quite precise if you would feed a very clean sine wave, but if it has noise and harmonics, it just can't get a very accurate result from just a short chunk of captured data. And that is why they include that frequency counter. That's what you are supposed to use if you want any real accuracy.

[thm_w]

This is normal behavior, so not an issue your scope. As Manul says the digital measurements just aren't that accurate.

There are definitely some software bugs in those measurements though. I tried to communicate with rigol to improve but no luck, maybe you can try.

You can also try turning on averaging or 20MHz filter, etc. to try to clean up the signal a bit.

[epoch_time]

After receiving confidence from Manul and thm_w that my post exhibits normal behaver for the DHO814 I continued my 60hz experiments.
My initial and exhaustive combinations of triggering, coupling, bandwidths. I was using horizontal setting of 5ms/div which displayed about 2.5 cycles of 60hz.
I then changed horizontal time base to 20ms/div which then displayed about 12 cycles of 60hz instead of only 2.5 cycles of 60hz this brought the measured mode of frequency  closer to the frequency counter reading.  Then after reading thm_w reply he recommended trying averaging  which I had already tried in my many configurations but I only used the default average setting of 2. I changed the average to 16 averages with success. Now the measure mode is about .005 hz difference from the frequency counter mode.  Making me very Happy with my rigol dso814...

Thx guys and eevblog.

[epoch_time]

 Hi thr_m.

I took your advice on using averaging which I was already using but only with the default setting of 2 averages.
I was using horizontal setting of 5ms/div which displayed about 2.5 cycles of 60hz.
I then changed horizontal time base to 20ms/div which then displayed about 12 cycles of 60hz instead of only 2.5 cycles of 60hz this brought the measured mode of frequency  closer to the frequency counter reading. Then changing the averaging from 2 to 16 averages. both measured mode and counter mode agreed by about .005 hz. 
Thanks