EEVblog #703 - Rigol DS1054Z Oscilloscope Review Summary

[Fungus]

Shall we make a list of all the buttons you never push on the front panel?

I'll start with the 'MODE' button in the trigger area and the big 'Clear' button at the top. Does anybody ever push those?

I'm sure they could have been used for something useful.

[EEVblog]

Shall we make a list of all the buttons you never push on the front panel?
I'll start with the 'MODE' button in the trigger area and the big 'Clear' button at the top. Does anybody ever push those?

I thought I showed the CLEAR button in my extended review? I think it's useful.

[Fungus]

Shall we make a list of all the buttons you never push on the front panel?
I'll start with the 'MODE' button in the trigger area and the big 'Clear' button at the top. Does anybody ever push those?

I thought I showed the CLEAR button in my extended review? I think it's useful.

I'm just watching that video now. It wasn't published here so I missed it.

Ok....I see where you use it. You'd think that the "huge-button-at-top-left" would do a really common function though.

eg. Use it to confirm a menu selection instead of pushing the stupid multi-function knob (swap the functions - use the multi-function knob to clear the display).

[aveekbh]

Incidentally, the 9.9E37 you saw on the software actually represents +infinity. I see this quite often with all sorts of instruments at work.

 9.9E+37 = INFinity (Infinity)
-9.9E+37 = NINF (Negative Infinity)
9.91E+37 = NAN (Not A Number)
Source: http://rfmw.em.keysight.com/rfcomms/refdocs/wcdma/wcdma_prog_invalid_results.html

But, yeah, that software is simply unusable.

[MarkL]

Here are some tests on the RP2200.

Below is a screen shot of an RP2200 looking at the swept output of a spectrum analyzer tracking generator from 1MHz to 201MHz.

I don't have an external power sensor, but plugging the TG back into the spectrum analyzer and also looking at it with an Agilent N2894A (passive 700MHz probe) both show that it's reasonably flat through the test sweep range.    The scope is an Agilent MSO3104A which has a brick wall BW of 1GHz.

The RP2200 is in yellow and the N2894A is included as a reference in blue.  Ground is even with the bottom, so you're seeing half the waveform.  The full horizontal sweep is shown, so each division is 20MHz.  Both probes were compensated before testing.

Also included is a shot of the both probes looking at a 50ohm terminated output of a ADCMP580 eval board.  This chip has a fall/rise time of 37ps typ. (20% to 80%), so it's effectively zero as far as the probes and scope are concerned.  There's also a shot of the pulse going directly into the scope.  There's a little bit of mismatch somewhere.

The RP2200 has more variation over frequency than I'd want to see, and some long overshoot on the pulse input.  But to be fair probes are part of the end-to-end acquisition path and it's only a true test when used with the intended scope.

EDIT: I should also point out that the amplitude on these sweeps is *linear*.  Most probe response curves you normally see are in dB.

[phenol]

the 'brickwall' bw rolloff of that keysight scope could explain the ringing on the pulse edges. Other competitive brands, such as R&S RTO scopes, have a gentle roll-off and, provided there are no transmission line mismatches, the intrinsic ringing is rather low...

[Lightages]

Well.... with Dave's response and this latest round of tests on the probes, I am doubting my results. I will retest and see if I screwed up somewhere, but I never screw up!

[MarkL]

the 'brickwall' bw rolloff of that keysight scope could explain the ringing on the pulse edges. Other competitive brands, such as R&S RTO scopes, have a gentle roll-off and, provided there are no transmission line mismatches, the intrinsic ringing is rather low...

Yes, I forgot about that.  You could be right.  It does have that look.

I double checked on a 500MHz TDS3054 which I believe has a gaussian roll off.   The ringing is the same (940ps), but a little less in magnitude as would be expected.

Because of the close agreement I'm not exactly sure in this case.  A different high speed edge source would settle it.

[Lightages]

I just retested all of my four RP2200s and my 2 RP3300s. I tested on my DS1054Z upgraded to 100MHz and on my DS1052E upgraded to 100MHz and on all channels. The results are the same. The RP3300 probes show the same as the yellow trace in image I posted, and the RP2200 probes all show the same as the blue trace. There are some very small differences between the probes but the RP3300 probes always show higher frequency content. This is the way that I am interpreting it.

Am I interpreting this incorrectly?

If not what are the chances that all of my four RP2200s are below spec and both of my RP3300s are in spec.?

Unfortunately I have no convenient test signal that I can use to actually quantify the difference. Any ideas what signal I might have one hand I can use to test this?

[Dave Turner]

Fungus made a point about the 'Mode' button in the trigger menu.

Would someone please  care to explain the the difference between the dedicated buttons - 'Auto', 'Run/Stop' & 'Single' and the Mode button which cycles between 'Auto', 'Normal' & 'Single'.

Ok I know it should be obvious - let's just say I have brain fade.

[MarkL]

I just retested all of my four RP2200s and my 2 RP3300s. I tested on my DS1054Z upgraded to 100MHz and on my DS1052E upgraded to 100MHz and on all channels. The results are the same. The RP3300 probes show the same as the yellow trace in image I posted, and the RP2200 probes all show the same as the blue trace. There are some very small differences between the probes but the RP3300 probes always show higher frequency content. This is the way that I am interpreting it.

Am I interpreting this incorrectly?

If not what are the chances that all of my four RP2200s are below spec and both of my RP3300s are in spec.?

Unfortunately I have no convenient test signal that I can use to actually quantify the difference. Any ideas what signal I might have one hand I can use to test this?

Are you using the alligator clip or the ground spring?  I assume you're referring to your post #32.

The tests I did were with the ground spring on both probes.  You need to make the ground as short as possible to reduce parasitic inductance.  The longer the ground, the more ringing you're going to see.

Some probes come with BNC adapters and those are even better, but there was no BNC adapter in the RP2200 package that I tested.

[Lightages]

Hmm yes. Actually now that I remember the question about the ground leads, the ground leads on the RP3300 are a bit longer. I will try again with the ground leads exchanged and with the short ground sprung points. I should have paid more attention to the ground lead lengths but I can be a bit of a stubborn ahole sometimes.

[Howardlong]

Fungus made a point about the 'Mode' button in the trigger menu.

Would someone please  care to explain the the difference between the dedicated buttons - 'Auto', 'Run/Stop' & 'Single' and the Mode button which cycles between 'Auto', 'Normal' & 'Single'.

Ok I know it should be obvious - let's just say I have brain fade.

The Run/Stop does exactly that, if it's either armed or triggering, it stops and displays the current acquisition, if there is one. Press again, and it restarts.

The only way to re-arm a Single shot is to press the Single button. Either that or you go to Auto or Normal, then Single, but that's not quite the same as the scope may well have already triggered as a result of going to Auto or Normal.

Force is not the same as Single.

If you press Run/Stop in Single mode, it goes to Auto.

However there is not a way to select Normal with these buttons. That must be done either from the Mode button or within the Trigger Menu.

So yes, all the Mode button functionality is also in the Trigger menu, but it's a more direct way of getting there.

I use the Run/stop all the time, and the Single button too. Hardly ever the Auto button, that has nothing directly to do with triggering, it's the auto setup button. Occasionally Clear, usually to clear the nag screen about expiring options on switch on. Or not, as the case may be.

[Lightages]

I just returned from doing the retest with the spring clips for the ground and also with the ground leads switched. With the ground leads switched there is no discernible difference in the appearance of the waveform. With the ground clips attached and probing the end of the BNC instead, both models of probes show the ringing on the signal even more, but they still show basically the same difference between appearance as before.

So that you know my setup for the waveform generation: I am using an analog function generator and using the TTL/CMOS output. The output is connected to a T with one side connected to a 50 ohm terminator and the other end a coax with alligator clips on the end. This is used for the probes with the probe grabber tips installed and the ground lead. For the spring clip test the alligator cable was switched to a normal BNC bale and the end measured with the bare probe tip and spring clip ground. The setup was chosen because it provides the fastest rise time on the aquare wave of any other method I have right now. The ringing is in the source cables used, not the probes.

[Howardlong]

You need to terminate the other end of the coax too.

The TTL/CMOS output probably won't be 50 ohms either, but it should be low impedance.

[Lightages]

I understand how the termination changes the waveform. That is not the point. I am observing the ringing wave form on purpose. Are you saying that the different probes are loading the source  differently and causing the waveform to change and I am seeing this change rather than the response of the probes? Therefore I need to terminate the source so that the change in the probe's loading will be negligible in comparison to the termination load?

Sorry we are getting so far off topic here......
Time to move to move to a separate thread. Hopefully we can determine the value of the probes that are supplied.

[phenol]

why don't you probe (using the ground spring) the T output directly and then remove the termination and probe the unloaded cmos gen output...

[Lightages]

I don't want to hijack this thread anymore than it has been sorry. Refer to this thread:
https://www.eevblog.com/forum/testgear/scope-probes-as-supplied-by-rigol-for-the-low-end-scopes-a-discussion/

[max_torque]

Dave, can you try some tests with the scope triggering?

On my 74Z, with two channels selected, the system triggers well on signals that are short and relatively infrequent, say a 100uS 5v pulse once every 100ms, but as soon as i turn on a 3rd or 4th channel, the triggering is lost completely (even without any input to those extra channels)?

  Now, it's quite possible i'm doing something stupid here ( ) but it would be nice to see if other people find the same thing?

[Dave Turner]

HowardL - thanks for your explanation. I do use the scope in the way you describe but I'd missed the subtle differences between the run/stop and single buttons and the mode control options.

[MarkL]

Thought I'd share a quick followup on some previous tests I posted on the RP2200:

  https://www.eevblog.com/forum/blog/eevblog-703-rigol-ds1054z-oscilloscope-review-summary/msg591033/#msg591033

I said:

But to be fair probes are part of the end-to-end acquisition path and it's only a true test when used with the intended scope.

So here's the same test with the spectrum analyzer tracking generator, but this time with a DS1054Z I was able to borrow for a day.  This unit has had the DSER hack installed.

Below are shots of the tracking generator output again, with and without the RP2200.  The sweep was 1MHz to 201MHz as it was before, and was made to span 10 divisions.  Direct input to the scope has a BW of almost 150MHz (3dB point, 70.7%).  With the RP2200 it's about 130MHz.  (Sorry about the small vertical size - I couldn't figure out how to get the cursor display out of my way to see where I needed on the trace.)  The blue trace is the trigger output from the tracking generator (high when sweeping).

You can see that the unevenness in the curve with the RP2200 is the same type of response as was seen with the RP2200 on a different scope.  It's definitely the probe.  Not great, but I've seen worse.

The displayed direct pulse input from the ADCMP580 eval board has a rise/fall of about 2.8ns, and with the RP2200 maybe a tiny bit less.  If you want to assume a gaussian filter (I don't know what's in there but it looks like it) this test puts the BW at around 125MHz direct and 130MHz with RP2200.

I've also also included a couple of shots of what the pulse looks like using the supplied 5.5" ground alligator clip.  The first is with the same pulse as the other tests, and the second with a lower frequency where you can see the ringing more clearly.

From what I'm measuring, the scope and scope+probe do significantly better than the expected 100MHz.

EDIT: Fixed link ref.

[Lightages]

Thank you very much MarkL.

[TerminalJack505]

Looks like a nice scope for the money.  Unfortunately you can't actually buy it for $399 now thanks to the eevblog effect.   

Ebay sellers are charging $550.  The seller on Amazon (TEquipment) is charging $459.  Although, if you go to their site they still have it listed for $399 but none in stock.

Probably just as well since that saves me from compulsively buying one.

[CrashO]

Unfortunately you can't actually buy it for $399 now thanks to the eevblog effect.   

Nah.
The scope has been hard to get (from stock) since it was released back in September.  Altough the waiting period has been steadily increasing since then.

I spotted one in stock last Sunday and immediately ordered one for €365 ( about $412 ) including VAT and received it on Thursday. 

[coppice]

The Tao Bao page of one of Rigol's resellers in China says Rigol has sold 20k of the DS1000Z series worldwide. I wonder how accurate that is, and how old this claim is?

http://detail.tmall.com/item.htm?spm=a230r.1.14.35.f3cxEj&id=39640772399&abbucket=19#