Specs/capabilities of Rigol "Hardware 6 bits Frequency Counter" in DS2000A

[DIPLover]

Does anybody have specs/measurements of the capabilities of the hardware frequency counter in the DS2000/DS2000A oscilloscope?

What the hell is a 6 bits frequency counter?  It is the only information given in the user manual (also of DS6000).

What's the max frequency you can measure with it?
How many digits does it show?
Whats the accuracy?
Is it a traditional or reciprocal design?

Any info you can provide greatly appreciated!

[0xdeadbeef]

If I should guess, I would assume that a 6bit frequency counter means that 64 periods (1<<6==64) are measured to calculate the frequency.

[atx]

A 6-bit frequency counter would be pretty pathetic.  That must be one of those Chinese translation errors.  What it actually is, at least on my MSO2k, is a 6 digit counter that is not limited to 999999Hz but rather shows the first 6 digits of the respective frequency you feed into it.  Thus, a 1kHz signal is displayed as 1.00000 kHz and a 150Mz signal as 150.000 MHz.

Hope this helps

[Bud]

When i was unlocking my 2072a the counter worked to 400MHz. I did not test for sensitivity or accuracy.

[DIPLover]

Thanks for all replies,

a 6 digit counter makes much more sense indeed. Funny they have the same "chinglish" mistake in the DS6000 manual...

That they do not give specs at all in the documentation is rather pathetic also.

We are a long way from the Tek and HP manuals of the 80s! (remember "Theory of Operation" chapters WOAH!)

But hey I can AFFORD one now ;-)

[DIPLover]

A little bit of necromancy, but this thread is still the #1 result when searching "Rigol DS2000 frequency counter" on google.

I have played a bit with smallish signals and low frequency and have started to establish some minimum specs.

My signal source is an old but well performing HP 8903B Audio Analyzer.
When testing for calibration this unit has always been better than the manual specs by a healthy margin.

Tests have been made with the counter source set to channel 1.
20MHz bandwidth limit was active.
1x probes where used (BNC T connection between 8903B source and analyzer).
Scope Input impedence set to 1Mohm.


Accuracy
-----------
The Rigol counter is a 6 digit unit of unknown accuracy.
The HP counter is a 5 digit unit with a specified accuracy of .004% + 1 digit.

In testing with sine waves of 200mV (RMS) amplitude, The Rigol's first 5 digits match the HP from 20Hz to 100KHz, plus or minus one 5th digit, thus at least matching that spec, might be even better, but I have no way to confirm if the last digit is any good.


Sensitivity
-------------
The Rigol's sensitivity is unspecified
The HP is specified as 5.0mV (RMS) in AC level and signal-to-noise modes

Using a 400Hz sine wave, with the scope set to 100mV/div, the counter is stable for signals down to 25mV (RMS) amplitude. 24mV jumps by several Hz and 23mV jumps all over the place.

I consider the counter output stable when the first 4 digits are constant.

At 1KHz the rigol's needs 26mV (RMS) to stabilize.
 10KHz needs 27mV (RMS).
 50KHz needs 27mV (RMS).
100KHz needs 27mV (RMS).

If at any time I set the scope to 50mV/div and hear the relay click, the counter becomes jumpy and innacurate with signals smaller than 600mV, thus being rather useless for signals you might wish to observe at that setting.


Conclusion
---------------
After these experiments I think it is safe to describe the counter in the Rigol DS2000 series as a 6 digits counter with an accuracy of at least 0.004 percent and a sensitivity better than 30mV (RMS) for a sine-wave input when the scope's vertical amplifier is set to 100mV per division. The counter was accurate from 20 Hz to 100KHz and as been reported to measure at least up to 400MHz.

Chances are the Rigol DS2000A is the most accurate counter I own, though not the most sensitive.
If someone with access to a better specced lab counter could run a few tests of accuracy at higher frequencies, please post results.

[rs20]

1x probes where used (BNC T connection between 8903B source and analyzer).
Scope Input impedence set to 1Mohm.

Can you expand on exactly what you did here? Did you use a scope probe set to 1x, or did you plug a BNC cable directly into your scope? Why do you have a BNC T connection if you have only a single source and analyzer, and presumably only need a straight connection between them?

If someone with access to a better specced lab counter could run a few tests of accuracy at higher frequencies, please post results.

Just pedantically/FYI, there's little reason to believe that accuracy would falter at higher frequencies -- a frequency counter is just that, a counter. It's not like it'd just fail to count properly, that'd be a completely unacceptable failure, not a standard type of inaccuracy (below the rated frequency of counter, of course). This is why a frequency counter provides the specs of the crystal used, as this wholly determines the accuracy of the system (the rest of the system being purely digital and therefore "error free"). The sensitivity, though, is another matter of course.

[DIPLover]

1x probes where used (BNC T connection between 8903B source and analyzer).
Scope Input impedence set to 1Mohm.

Can you expand on exactly what you did here? Did you use a scope probe set to 1x, or did you plug a BNC cable directly into your scope? Why do you have a BNC T connection if you have only a single source and analyzer, and presumably only need a straight connection between them?

Indeed, no probes were involved. The scope was set to 1X.
A BNC T connector was used to join the 8903B Source to it's analyzer with the Rigol's channel 1 input in-between. This so I can read both counters at the same time with the same source.

If someone with access to a better specced lab counter could run a few tests of accuracy at higher frequencies, please post results.

Just pedantically/FYI, there's little reason to believe that accuracy would falter at higher frequencies -- a frequency counter is just that, a counter. It's not like it'd just fail to count properly, that'd be a completely unacceptable failure, not a standard type of inaccuracy (below the rated frequency of counter, of course). This is why a frequency counter provides the specs of the crystal used, as this wholly determines the accuracy of the system (the rest of the system being purely digital and therefore "error free"). The sensitivity, though, is another matter of course.

I would also be surprised if accuracy degraded significantly with frequency.
Having made no measurement past 100KHz however I don't feel I can make any claims for the whole range, which is, just like the specs of the crystal used and everything else, unspecified by the manufacturer.