EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems

[Bud]

seems that if L1 and L2 are a bit wrong value-wise, you won't just get a bit of FMing, it'll be miles off, if it oscillates at all. ... When you're talking single digit nH, layout does of course make a difference

A quick calculation shows a 10mil via on a 1.6mm FR4 can add 1.46nH of inductance. Indeed L1/L2 should be of a smaller value and  may have contributed to the problem. It is better to solder directly to the ground plane than use vias (see the attached pic)

especially regarding parallel as opposed to orthoganol or linear placement and associated coupling 

On my board I placed them at 90 degrees. Also I solder L1/L2 cold ends directly to the ground plane.

[miguelvp]

Well, I didn't mind the jitter on my DS2000 but I'm glad that is gone. What is getting me is that they broke the USB reinsertion. So sometimes I can't get a screenshot without rebooting unless I get all my signals first.

I thought it wouldn't be a big deal but the way I usually work is I take the thumb drive after a set to captures to transfer then I insert it back and no go

I'll eventually adapt and remember to make sure I see the usb devices inserted if I plan to capture anything, but it's getting a bit annoying.

[Monkeh]

Well, I didn't mind the jitter on my DS2000 but I'm glad that is gone. What is getting me is that they broke the USB reinsertion. So sometimes I can't get a screenshot without rebooting unless I get all my signals first.

I thought it wouldn't be a big deal but the way I usually work is I take the thumb drive after a set to captures to transfer then I insert it back and no go

I'll eventually adapt and remember to make sure I see the usb devices inserted if I plan to capture anything, but it's getting a bit annoying.

I find it easier and faster (plus, you get timestamps) to take screen captures over the network.

[Howardlong]

Howard,

Can we see what the CP pin looks like with the Rigol loop filter values / old firmware combo?

Although unlocked, your eval doesn't have the serious FM happening that we've seen on some of the old firmware and beta units.  With the first sideband at -36dB, I'm getting a deviation of 3.1kHz, which translates to a displayed jitter of 20ps pk-pk @5us delay.

The first DS1054Z I looked at had a 50mVpp 100kHz sawtooth happening on the CP pin with the old firmware.  You obviously have the 100kHz modulation; I'm just wondering if examining the CP behavior will provide any clues.  (But your deviation may be too small to see on CP or Vtune).

Hello Mark.

Sorry, managed to blow up the 4360-7 on the board today when fitting a 25MHz canned oscillator instead of using the sig gen to see if it made any difference. Anyway, I've replaced it now, but debugging that delayed me a while.

The difference between the canned 25MHz reference oscillator and the signal generator in terms of phase noise was negligible as far as I could see, so the measurements I'm showing here are with the HP signal generator (8656B).

Not sure how you want me to measure this, on the scope the CP pin is in the noise at 10mV/div with both old and new register settings. Certainly not 100kHz 50mVpp.

Also, I need to make a correction, the PLL is showing locked on the old values too, it just wasn't being output on the pin due to the registry settings. Doh! Anyway, I guess that's as much to do with the way that the lock is configured and interpreted as much as anything else.


Signal generator phase noise (HP 8656B)


Canned oscillator phase noise (for information, not used in further testing).



4360-7 output, new firmware 200kHz span



4360-7 output, new firmware, 20kHz span


4360-7 output, old firmware, 200kHz span


4360-7 output, old firmware, 20kHz span

[photon]

My kudos to the debug work done on this topic.

I am having difficulty understanding the claim made here that the PLL does not lock. I would be interested in hearing an explanation of how it is possible for the PLL not to lock and yet the clkout it produces is so spectrally good that the first sideband is down some 30dB.

[Bud]

You must be kidding. 30dB spuri below the fundamental is an utter garbage in integrated PLL world.

[Howardlong]

You must be kidding. 30dB spuri below the fundamental is an utter garbage in integrated PLL world.

If you're referring to the pics I posted earlier, if you are transmitting or demodulating narrowband RF then yes, the adjacent channel interference and rejection would be rubbish. However, practically speaking in its current application in a scope, I'd say while it's not pristinely beautiful, it would have little if any noticable impact.

[MarkL]

Also, I need to make a correction, the PLL is showing locked on the old values too, it just wasn't being output on the pin due to the registry settings. Doh! Anyway, I guess that's as much to do with the way that the lock is configured and interpreted as much as anything else.

Are you saying this (and the others that had this modulation) from your post #1092 was actually locked??


Old firmware on Eval board with DS1000Z loop filter parts

[Howardlong]

Also, I need to make a correction, the PLL is showing locked on the old values too, it just wasn't being output on the pin due to the registry settings. Doh! Anyway, I guess that's as much to do with the way that the lock is configured and interpreted as much as anything else.

Are you saying this (and the others that had this modulation) from your post #1092 was actually locked??


Old firmware on Eval board with DS1000Z loop filter parts

When I switched the MUXOUT to Digital Lock Detect (Active High) it did indeed lock. Quick test, remove the ref oscillator, MUXOUT goes off.

Plus I think we need to be a little careful about the closer in measurements (<=200kHz span): there appears to be some ref oscillator residual FM with a +/-500Hz deviation on my 8656B that isn't there with the canned oscillator. This is definitely reflected in those close-in shots.

I ran the ref oscillators and 4360-7 outputs through one of my SDRs this morning: these can run RBWs down to fractions of Hz, but you need to be careful with them in that they are very sensitive, and pick up RF from anywhere: they are meant to be radios after all! The example pics below were taken without any antenna, just with the SDR sitting a couple of inches away, picking up the radiated RF. SDRs, being radios first and foremost, are also not calibrated like SAs are, and do not have the wide spans of SAs, but can be useful as bits of test gear nonetheless.


40kHz span (canned oscillator)


1.6kHz span (canned oscillator)


Coupling was wirelessly, mostly radiated from the 12" unterminated SMA cable to the left of the board.

[The Doktor]

Do any of you have a link to the latest firmware for my DS1054Z? I thought there was a link somewhere in this thread, but I have searched back about 15 pages and not found it.

I just received my scope today, and it is showing firmware version 00.04.02.sp3 board version 0.1.1. When they did a video about the gender fixed, he was showing firmware version 00.04.02.SP4, so it does not appear my scope has the very latest version. I would like to update it as soon as possible.


Thank you
Ed

[nanofrog]

Do any of you have a link to the latest firmware for my DS1054Z? I thought there was a link somewhere in this thread, but I have searched back about 15 pages and not found it.

I just received my scope today, and it is showing firmware version 00.04.02.sp3 board version 0.1.1. When they did a video about the gender fixed, he was showing firmware version 00.04.02.SP4, so it does not appear my scope has the very latest version. I would like to update it as soon as possible.


Thank you
Ed

00.04.02.SP4 (direct download)

[rolycat]

Do any of you have a link to the latest firmware for my DS1054Z? I thought there was a link somewhere in this thread, but I have searched back about 15 pages and not found it.

You didn't miss it by much; it's 17 pages back:

https://www.eevblog.com/forum/blog/eevblog-683-rigol-ds1000z-ds2000-oscilloscope-jitter-problems/msg578208/#msg578208

[dadler]

I knew about the jitter problem, but somehow the ac trigger coupling issue eluded me. When my new scope showed up today, newest batch from Tequipment (v00.04.02.SP3), I was quite freaked out when I turned on ac trigger coupling and saw bizarre out-of-phasey output. Updated to SP4 and all is well, ac trigger coupling acts as expected now.

[Wmacky]

Well this thread kinda died.

I was about pull the trigger on a 2072A when I spotted this issue /thread. Is this a bad time to buy?  I know the issue has been somewhat resolved, but still.  Would you wait for a possible hardware change? I assume the 2072A will have a lower turn over rate than the 1054Z, and new hardware would take longer to show up.  (If Ever) Is it worth wait a period of time to see? I guess if a hardware fix diesn't show up on the next board revision, then it probably never will.

What was the last board revision for the 2072A, known to have this issue. Has anyone seen a new revision board number for the 2079A since?  Anyone buy one in the last few weeks, that could check the board revison number on boot up?

[Teneyes]

Well this thread kinda died.

Yes this thread is dead, as the bug is fixed for DS2000/A with latest Firmware release.
As I was one of the owners that report this bug to Rigol 21 months ago, I am satisfied it is corrected!

I do Not know what hardware change you are looking for?
There may be other issues you have?

[Wmacky]

Well this thread kinda died.

Yes this thread is dead, as the bug is fixed for DS2000/A with latest Firmware release.
As I was one of the owners that report this bug to Rigol 21 months ago, I am satisfied it is corrected!

I do Not know what hardware change you are looking for?
There may be other issues you have?

Perhaps, but there seems to be some debate about that.  My understanding from reading all the threads is that the root of the issue was caused by the configuration / components used for the PLL loop. It was also suggested that Rigol could quietly make a hardware change to this area, on the next revision. I'd hate to make a purchase a week before a better hardware solution was offered.  Or, they could do nothing at all, and it doesn't really matter.

[rolycat]

My understanding from reading all the threads is that the root of the issue was caused by the configuration / components used for the PLL loop. It was also suggested that Rigol could quietly make a hardware change to this area, on the next revision. I'd hate to make a purchase a week before a better hardware solution was offered.  Or, they could do nothing at all, and it doesn't really matter.

There's no point worrying about what Rigol may or may not do in future revisions when the hardware available now is perfectly functional.

Test gear is much like consumer electronics: there always seems to be a newer version, a price drop or a special deal waiting around the corner until you commit to the current model.

On the other hand, there are suggestions that Rigol is planning a price increase for the DS1054Z and conceivably other scope models, and who is to say there won't be fresh bugs in any hypothetical new hardware revision?

Just buy the scope.

[nanofrog]

My understanding from reading all the threads is that the root of the issue was caused by the configuration / components used for the PLL loop. It was also suggested that Rigol could quietly make a hardware change to this area, on the next revision. I'd hate to make a purchase a week before a better hardware solution was offered.  Or, they could do nothing at all, and it doesn't really matter.

There's no point worrying about what Rigol may or may not do in future revisions when the hardware available now is perfectly functional.

Test gear is much like consumer electronics: there always seems to be a newer version, a price drop or a special deal waiting around the corner until you commit to the current model.

On the other hand, there are suggestions that Rigol is planning a price increase for the DS1054Z and conceivably other scope models, and who is to say there won't be fresh bugs in any hypothetical new hardware revision?

Just buy the scope.

Well said. 

As you say, they work with the latest firmware revision. And given the rumors of price increases are on the way, now really is the right time to buy.

[Wmacky]

I should have it by Friday.......

[Slasher]

I have had my DS1054 for some days now and used it for a little project I am doing. In that project I am using a small pulse signal superposed on a DC voltage. I change the DC voltage to drive a LED. Therefor there is a big change in amplitude on the pulse and the DC level and I want to change to DC level also. I was very frustrated when the scope didnt want to trigger using the AC feature. It was just like I was in DC trigger mode all the time. I knew that this must have been fixed but when I tried to upgrade with the firmware posted in this thread it said it already had that version (I had checked before so I knew that) but I said that it should update anyway. After that the AC trigger started to work better although it still has problem when the pulse has a small amplitude compared to the DC level.
It feels like the firmware version my scope came with wasnt really the same as the one posted here. Well, just my heads-up on that. Back to work.
 

[straycat]

Hi guys.Im looking into getting myself a DS1054z. I have seen DAVE,S REVIEW and have read most of the problems, seems to be a lot of issues with this scope. I have a ds1102e scope and never had any problems. What it boils down to is whether or not do i get one.Any good advice would be greatfull.

[tautech]

You would be wise to study this thread:
https://www.eevblog.com/forum/testgear/rigol-ds1000z-series-(ds1054z-ds1074z-ds1104z-and-s-models)-bugswish-list/

[straycat]

tautech. Thank you for your advice. I think il stick with old faithful for the time being and look around for maybe a better one. Thank you.

[marmad]

tautech. Thank you for your advice. I think il stick with old faithful for the time being and look around for maybe a better one. Thank you.

Regardless of the bug/wish list, the DS1054Z is an amazing bargain at the price, and I'm not sure there is anything else close to it in value per dollar. It really is a big jump technologically in many ways from the DS1102E.

And here's the thing: Rigol has a bad rep for communicating with their customers and informing them of updates, etc. (and much of the bad press is true) - and yet, they actually DO fix bugs and add new features - at a faster rate than any other Chinese DSO manufacturer that I've ever dealt with.

[straycat]

After reading about a lot of issues with the DS1000Z, i decided to go with a tektronix TDS2012C. I have a rigol ds1102e and never had any problems.So im not knocking them.