Author Topic: Our New (possibly faulty) DS1054Z (and it's channel amplitude discrepancies)  (Read 6560 times)

0 Members and 1 Guest are viewing this topic.

Offline adamboon

  • Contributor
  • Posts: 26
  • Country: au
Hi,
So i read around this forum, interested in purchasing a new DSO to replace our ageing and unreliable 100MHz HP Scope.  Settled on the Rigol DS1054z.
We work mostly on Audio circuits, with tubes, opamps, FET's, BJT's and Tape Machines.  Now I understand that most of these budget DSO's are 8 bit, and are far less accurate in the vertical than a good meter _BUT_ I'm pretty disappointed with the discrepancy in measurement between ch 1 and all the other channels on our new DS1054z.

Measuring it’s own test point - the RMS values are giving me are 2.33 - 2.38 on channels 2, 3 & 4 (I’m totally fine with the 0.05 v variance) but ch 1 reports 2.00v (I’m not okay with a 0.38v, or just shy of 20%, variance).  I have swapped the probes around, and the difference stays with the channel rather than the probe.
 
I have "Self Calibrated” the unit twice (the first time only after about 15 mins of (warmup) on time, the second after about an hour).
 
I also connected it to a 9V battery - the DC side seems to match up quite nicely.

Quote from: My reply from the Australian Distributer
The Rigol DS-1054Z is a general purpose 8 bit scope and accuracy on the unit is around 3% for the signal level you are measuring BUT this is 3% of the full scale or screen so on 2V/Div it’s 3% of 16V which is +/- 0.48 so is within spec.
 
To get a better accuracy for your measurements you could decrease the V/div to 1V or 500mV/div that will give a better accuracy on a 2V signal.
 
Hi-res is an effective mode of increasing the resolution of the signal for lower speed signals by doing a real-time average of the signal using the long record length but you still come up against the basic DC gain accuracy of the scope.

so at 500mV / div i'm getting 1.46v on ch1 vs 1.75v on ch2 - this should be within 4v * 3% (4 x 0.03) = 0.12v? but is closer to three times that.

I would be making serious measurements with an RMS meter - but I did expect the scope to be a little better than this.  or at least let me apply a compensation to the channels to make the readings useful.

Are my expectations too high?  or have i got a dud? - I wouldn't even mind as much if the discrepancy was on ch's 3 or 4...
« Last Edit: June 16, 2016, 10:25:08 PM by adamboon »
 
The following users thanked this post: markonj

Offline saturation

  • Super Contributor
  • ***
  • Posts: 4750
  • Country: us
  • Doveryai, no proveryai
    • NIST
I recall  older post on this issue.  A more recent post suggests variance is normal  :-//.

http://www.eevblog.com/forum/testgear/ds1054z-voltage-differences-between-channel/msg907571/#msg907571

Search the forum as  "1054Z voltage" and choose a discussion relevant to you.

A properly made scope will have all channels nearly identical and to spec, say 3% both for AC and DC , and particularly if you are exacting with the self calibration.  Even if they are off calibration, all channels should drift close together <=3%.  In actual measurement this often far less, < 1%.

I have both a 1052e and a newer GDS 1054B with all channels identical in spec, both for measurement readings and function; I need them that way and I paid for that performance.
« Last Edit: June 17, 2016, 12:48:25 AM by saturation »
Best Wishes,

 Saturation
 

Online wraper

  • Supporter
  • ****
  • Posts: 5686
  • Country: lv
Does not sound OK at all, if it was 5%, then maybe fine. But 20% is not fine for any oscilloscope. I had similar problem with Rigol Germany when my DM3068. They tried to prove that it's was performing OK, while it was certainly not. 6 months of brainfuck trying to deal with them.
« Last Edit: June 17, 2016, 01:00:35 AM by wraper »
 

Offline alsetalokin4017

  • Super Contributor
  • ***
  • Posts: 1562
  • Country: us
There is the trace display, and then there are the scope's calculated RMS measurements. As we know, the calculated RMS measurements can be quite bad, with the scope even able to report large RMS voltage values for channels that are _grounded_ and receiving no signal at all.

I'm not sure if this "feature" came along with the latest firmware or not, as we didn't notice it until the SP2 firmware was in widespread use, and it's not possible to roll back to earlier firmware to check if the RMS errors are still there.

One thing that is important to realize is that the scope uses the portion of the signal that is displayed onscreen for the calculated measurements. So for certain measurements, it makes a difference how many cycles or partial cycles are displayed. RMS can be one of those measurements. But still, one would like any errors due to partial cycles displayed, to be consistent across channels.

It would be interesting to see the actual traces on adamboon's screen when the measurements disagree so much. Do the traces themselves also disagree by the same amount, or is this "just" the RMS measurement bug doing its thing?

In any case, I suggest documenting the heck out of this problem, and if the scope is consistently shown to be out-of-spec, you can and should insist on a replacement under either the vendor's warranty or Rigol's factory warranty.

« Last Edit: June 17, 2016, 01:08:06 AM by alsetalokin4017 »
The easiest person to fool is yourself. -- Richard Feynman
 

Online wraper

  • Supporter
  • ****
  • Posts: 5686
  • Country: lv
I also connected it to a 9V battery - the DC side seems to match up quite nicely.
What is the actual signal you are measuring, frequency? Have you properly adjusted probe compensation on particular channel?  :palm:  it's there, should read more carefully.
« Last Edit: June 17, 2016, 01:43:02 AM by wraper »
 

Online wraper

  • Supporter
  • ****
  • Posts: 5686
  • Country: lv
I'm not sure if this "feature" came along with the latest firmware or not, as we didn't notice it until the SP2 firmware was in widespread use, and it's not possible to roll back to earlier firmware to check if the RMS errors are still there.
This seems like a very obvious firmware bug, nothing else. Waveform certainly does not show anything like this being even remotely possible.
 

Online wraper

  • Supporter
  • ****
  • Posts: 5686
  • Country: lv
BTW what about Vpp measurment, actual signal amplitude shown on the dispay? Maybe you should make a screenshot and post it here.
 

Offline adamboon

  • Contributor
  • Posts: 26
  • Country: au
Thanks for your helpful reply.  One question on the below comment:

I have both a 1052e and a newer GDS 1054B with all channels identical in spec, both for measurement readings and function; I need them that way and I paid for that performance.

Is the (older, 2ch) Rigol 1052e a better, more accurate or more reliable piece than the 1054z?

I guess the same question for the Instek.  I went with the 1054z because it seemed like it was getting lot of love on this forum.  Also I liked that I got it through an Australian Distributer.  I was also looking at Hantek, Siglent and USB Picoscope models.  I don't need 4 channels, or fast speeds.
 

Offline adamboon

  • Contributor
  • Posts: 26
  • Country: au
Attached is the original photo i snapped (Rigol's "Auto" settings, i tried it with the probes connected one way, then plugged them in backwards to check the problem wasn't a probe).

Second pic is manually changing the V/div as suggested by the distributer.  I _MAY_ have fiddled the fine vertical amplitude adjust (I was playing with it to see if it affected the Vrms value - it doesn't).

I will check the Vpp and Vmax this morning and report back.

Thanks for all your helpful and patient replies.
 

Online wraper

  • Supporter
  • ****
  • Posts: 5686
  • Country: lv
I would say it's a firmware bug, not actual hardware issue because waveform amplitudes look exactly the same. What happens if you try to disable all other channels? Try to update the software and/or reset it to the factory settings. Try this reset method too:
Quote
On the left side of the front panel, there are six dark grey menu select buttons. And the magic one is the fifth from the top. To reset the oscilloscope, just power off and then on the scope, repeatedly pressing the fifth button and the oscilloscope resets to factory settings. Yeah, the menu changes to chinese, but that’s easily fixed by just selecting English in the language settings on the menu to the right of the screen.
 

Offline adamboon

  • Contributor
  • Posts: 26
  • Country: au
Wraper, I think your onto somthing.  Vpp is within .02v

The unit reports Version
00.04.03.SP2

And board version
0.1.4

I'll try the reset when I'm back in front of the scope.  Thanks all for all your help!
 

Offline Meka77 wd

  • Contributor
  • Posts: 20
  • Country: tr
Hi everyone... This is for adamboon. I am waiting a solution about 6 months for this and other concerning problems. And our local rigol dealer does not answer my phone calls anymore :-- . Before 4.03.02.03 firmware everything working fine and fast for me anyways. I think main issue is devices ability to flash former firmware, because accidents may happen in factory when developing or debugging on new firmware. Then user choose return to former working firmware, like cellphohe or pc etc. I actually beg them for return to sp1 firmware because i am using this device for work. But nothing so far and everyone seems happy with this and not complainig about it to rigol. I have 4 rigol scopes at my home right now, my first ds5022m still works as expected, ds1052e at 100mhz works as expected, ds2000a at 300mhz works nicely and you can flash whatever firmware ver. you want. Last this 1054z sitting over couch about 5 months and still have a  useless warranty. Sorry for my bad english and thank you for your time. http://www.eevblog.com/forum/testgear/ds1054z-issue/msg919023/#msg919023
 
The following users thanked this post: saturation

Online wraper

  • Supporter
  • ****
  • Posts: 5686
  • Country: lv
I have an idea. I guess you always trigger on CH1, what happens if you trigger on some another channel, or using other channels one by one with all other disabled? BTW, it seems that actually CH1 is showing correct Vrms but all other channels are lying.
Edit: as Meka77 wd said, this seems to be the same firmware bug, read that tread. Digital crosstalk  :palm:, do Rigol verify their firmware at all?
« Last Edit: June 17, 2016, 08:27:40 PM by wraper »
 

Offline saturation

  • Super Contributor
  • ***
  • Posts: 4750
  • Country: us
  • Doveryai, no proveryai
    • NIST
Hi I don't own a Z so I can't compare. 

The 1052e performs per its data sheet and what bugs it had or remains are not critical, they don't impact measuring capabilities; e.g. my unit may lock up if I hot swap a USB drive when triggering, the rotary encoders sometimes get too sensitive, type issues. 

The c2008-2009 1000e has had time to mature.  But even on initial release there were no critical bugs and brought Rigol's fame for its price performance. A variant of the 1000e series was rebadged for Agilent with the Rigol version having slightly more features, such as memory. 

Instek as a company makes reliable devices [I've used them off/on for 30+ years] but they cost more and have less features for price, if you compare spec sheets.  But it lives up to its spec sheet and its quality is consistent. 

The 1.5 year old GDS 2000E series had bugs on release but none critical, you can read of the genesis of issues in eevblog forum reviews.  The lower end GDS 1000B series is ~ 1 yr old and had no critical bugs, actually no bugs with my unit OOB and with current firmware.  Because of the new technology used in 2000E and 1000B models, they are lower priced too to come close to the price performance of Rigol equivalent models, but with Instek reputation.

I suggest you get a scope from a local dealer, so you can test and exchange whichever until one works for you.  What we see in eevblog's international forum is take home prices for models as well as availability vary with country.  If you buy mail order international you could save but if things go wrong and its warranty not honored locally, it can cost a lot to ship back internationally. 

Last cautious words: over the years of reading eevblog forum chatter, the quality control of Rigol has taken a step down in entry level devices of its product line: the Z, the 1000e series [ newer production] and the DP 832 PSU, to name a few.  So test it thoroughly after receipt within the window you can return it to the dealership for an exchange.


Thanks for your helpful reply.  One question on the below comment:

I have both a 1052e and a newer GDS 1054B with all channels identical in spec, both for measurement readings and function; I need them that way and I paid for that performance.

Is the (older, 2ch) Rigol 1052e a better, more accurate or more reliable piece than the 1054z?

I guess the same question for the Instek.  I went with the 1054z because it seemed like it was getting lot of love on this forum.  Also I liked that I got it through an Australian Distributer.  I was also looking at Hantek, Siglent and USB Picoscope models.  I don't need 4 channels, or fast speeds.
« Last Edit: June 17, 2016, 11:26:28 PM by saturation »
Best Wishes,

 Saturation
 

Online wraper

  • Supporter
  • ****
  • Posts: 5686
  • Country: lv
I guess the same question for the Instek.  I went with the 1054z because it seemed like it was getting lot of love on this forum.  Also I liked that I got it through an Australian Distributer.  I was also looking at Hantek, Siglent and USB Picoscope models.  I don't need 4 channels, or fast speeds.
I have multiple pieces of Rigol equipment and certainly regret buying their DM3068 multimeter. Kind of fine with DG4062 signal gen but as I don't use it too much, have not stumbled into the bugs which are there according to what people write. DS2072 scope didn't make me feel very happy about the purchase either, mostly because of hanging and painful controls (as example pushable intensity/menu selection encoder  |O, pain in the ass. Never can select what you want). But at the time there wasn't anything comparable from Instek to replace my old trusty Instek GDS-1152A (similar to Rigol DS1000E). Recently I bought Instek GDS-2204E and it is like day and night compared to Rigol. If you are fine without hacking the scope and don't need getting serial decoders  for free by doing that (which I'm sure most people don't need but say it as 100% argument why you must buy it and then hack), GDS-1054B is much better deal, and the most of all it can be trusted, cost even a bit lower than DS1054Z.
 

Offline adamboon

  • Contributor
  • Posts: 26
  • Country: au
Oh I wish I would have purchased the instek...  I can't see them in Australia though - so i'd need to ship it from overseas.  I'm torn between sticking with the 1054z and hoping Rigol get their act together.  But I might ask if I can send it back and look into the Instek.  It will probably depend on if the RMS issues os resolved by the reset - haven't had a chance to try it yet.
 

Offline saturation

  • Super Contributor
  • ***
  • Posts: 4750
  • Country: us
  • Doveryai, no proveryai
    • NIST
Perusing google I see several GWInstek dealers in Oz but the model is not listed on the website, can't hurt to ask them.  As always, local dealers can provide support; a demo was sent to Dave back for the 11/2015 review so maybe there is a preferred dealer.

There is one side issue: if your workplace requires "NRTL equivalent in Australia recognized" EMC and safety certification, this device is only CE. 


Oh I wish I would have purchased the instek...  I can't see them in Australia though - so i'd need to ship it from overseas.  I'm torn between sticking with the 1054z and hoping Rigol get their act together.  But I might ask if I can send it back and look into the Instek.  It will probably depend on if the RMS issues os resolved by the reset - haven't had a chance to try it yet.
Best Wishes,

 Saturation
 

Offline adamboon

  • Contributor
  • Posts: 26
  • Country: au
So I've done a bit ok poking around, and everything except the Vrms seems to okay.  It is the rms issue as outlined in the other thread.  there is another function Vrmsp which is also called Pers. Vrms

I'm taking a wild guess that this is Persistent RMS - But I don't know what that means.  It does seems to math it out correctly though.

Are there any other "Issues" I should be aware of with the 1054z?  The distributer is going to call me back soon and I'm going to make some noise about the RMS issue and also request that I send it back and get a GDS-1054B if possible.  Mostly because of how many other bugs and issues may be in the machine that I am yet to discover - if I wanted an unreliable 'scope I could just use my old analog HP :P
« Last Edit: June 22, 2016, 10:56:14 AM by adamboon »
 

Offline saturation

  • Super Contributor
  • ***
  • Posts: 4750
  • Country: us
  • Doveryai, no proveryai
    • NIST
Here's an updated list of bugs, it would be a great service to all if you can confirm it:

http://www.eevblog.com/forum/testgear/new-rigol-ds1054z-oscilloscope/msg966871/#msg966871


So I've done a bit ok poking around, and everything except the Vrms seems to okay.  It is the rms issue as outlined in the other thread.  there is another function Vrmsp which is also called Pers. Vrms

I'm taking a wild guess that this is Persistent RMS - But I don't know what that means.  It does seems to math it out correctly though.

Are there any  other "Issues" I should be aware of with the 1054z ?  The distributer is going to call me back soon and I'm going to make some noise about the RMS issue and also request that I send it back and get a GDS-1054B if possible.  Mostly because of how many other bugs and issues may be in the machine that I am yet to discover - if I wanted an unreliable 'scope I could just use my old analog HP :P
Best Wishes,

 Saturation
 

Offline alsetalokin4017

  • Super Contributor
  • ***
  • Posts: 1562
  • Country: us
So I've done a bit ok poking around, and everything except the Vrms seems to okay.  It is the rms issue as outlined in the other thread.  there is another function Vrmsp which is also called Pers. Vrms

I'm taking a wild guess that this is Persistent RMS - But I don't know what that means.  It does seems to math it out correctly though.

If you are talking about the Rigol DS1054z, the measurement is actually "Per.Vrms" and it stands for "Period RMS". If you push the Help button at the top right of the control panel, then press the "Per.Vrms" hotkey in the Measurements area on the left side of the screen, you will see what the scope means by this function. It is the RMS value within a single period. The normal "Vrms" measurement is the RMS value across the whole screen display.
 
The scope's "Help" function is a nice feature which will give more-or-less thorough explanations of just about any hotkey or control panel knob or button.

Quote

Are there any other "Issues" I should be aware of with the 1054z?  The distributer is going to call me back soon and I'm going to make some noise about the RMS issue and also request that I send it back and get a GDS-1054B if possible.  Mostly because of how many other bugs and issues may be in the machine that I am yet to discover - if I wanted an unreliable 'scope I could just use my old analog HP :P

The various other issues with the Rigol DS1054z have been discussed elsewhere in the forum. The one that bothers me the most at present is the Measurements Fail bug that causes all Measurements to simply stop working after a random time period, when 3 or 4 channels are turned on and a Math trace is displayed, requiring a power-cycle reboot to reestablish Measurements working.
The easiest person to fool is yourself. -- Richard Feynman
 

Offline adamboon

  • Contributor
  • Posts: 26
  • Country: au
I spoke with John South from emona, where I purchased the scope.  He has verified the (Vrms) fault and is working on it with Rigol.

He has been very good to deal with.  I think I am happy again to have paid up and waited to get local support.

I'll keep y'all posted with the details.
 

Offline alsetalokin4017

  • Super Contributor
  • ***
  • Posts: 1562
  • Country: us
I spoke with John South from emona, where I purchased the scope.  He has verified the (Vrms) fault and is working on it with Rigol.

He has been very good to deal with.  I think I am happy again to have paid up and waited to get local support.

I'll keep y'all posted with the details.

Thanks for keeping the pressure on.

I wonder if you or John South have encountered this Measurements Fail bug as shown in the video below yet. I did report this one to Rigol USA six or seven months ago, they said they were indeed able to reproduce it on their lab scopes, but that's the last "official" word I've heard of it.
The easiest person to fool is yourself. -- Richard Feynman
 

Offline bitseeker

  • Super Contributor
  • ***
  • Posts: 3248
  • Country: us
  • Lots of engineer-tweakable parts inside!
I spoke with John South from emona, where I purchased the scope.  He has verified the (Vrms) fault and is working on it with Rigol.

That's great news! Hopefully, a fix for that and the lockup during Math functions (already reproduced by Rigol) will be forthcoming. And that "Pluses" spelling error (talk about an easy fix). Anyway, I'll try not to get my hopes up, but it sure would be nice.

You can do it, Rigol. ::)
Got GAS? Get TEA.
 

Offline adamboon

  • Contributor
  • Posts: 26
  • Country: au
John from Emona just emailed me this update:

Quote from: John South
HI Adam - Rigol will release new firmware to fix this and other bugs by the end of July.

I have found them (emona) great to deal with so far - and will stick with the Rigol (and await the update).
 
The following users thanked this post: nrxnrx

Offline bitseeker

  • Super Contributor
  • ***
  • Posts: 3248
  • Country: us
  • Lots of engineer-tweakable parts inside!
Thanks for the great news, Adam. And thanks to John at Emona for the great customer service. Maybe Rigol was already going to release an update, maybe not. Either way, it doesn't hurt to have more vendors influence change.
Got GAS? Get TEA.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf

 

http://opalkelly.com/