EEVblog #369 - Rigol DS2000 Oscilloscope Playing Around

[EEVblog]

Dave.

[JimmySte]

Hi Dave...
Liked the video...
Just thought I'd comment of some of the issues you had in the video...
Firstly, the green leds on my DS2072 are just as bright as the red start/stop button... so maybe you didn't align something when you put the unit back together...
2nd, the updated firmware must have broken the waveform record... nice find works as you'd expect on mine... No need to change frames to update the screen
So I might wait before updating the firmware... Guess the folks I got mine from will supply it to me...
On a side note... the scope auto adjusted it's time for DLS..., I can't recall setting my location...
Can't wait for the hour+ long review...

[madworm]

Is it just a coincidence, or do these DSO firmware manufacturers ALL make the same errors... The lack of screen refresh in zoom mode brings back not so fond memories.

This again shows that one should NOT buy test gear without elaborate testing and 'playing' with it for at least a week or two. It will be very interesting to see how long it takes them to fix this.

[BravoV]

Is it just a coincidence, or do these DSO firmware manufacturers ALL make the same errors... The lack of screen refresh in zoom mode brings back not so fond memories.

Its called "Copy & Paste Engineering" .. 

[notsob]

One question. Is there a way to backup the existing firmware before upgrading, in case there is problem as dave has found, it would be good to be able to revert to a previous version whilst awaiting a fix from rigol.

[madworm]

Copy & Paste engineering... apparently.

BUT this also shows that they don't do ANY form of testing that is worth to be called that. Any at least half-sane programmer uses version control, so they should know which changes went into the new firmware release and should know which other functions are affected by these. And that allows for selective testing... and they don't even seem to do any of that.

SNAFU.

[nitro2k01]

6:50 "Pleas don't remove the U disk..." I'm wondering why they didn't spell out the word USB.
9:00 The hardware version probably refers to the board revision.
25:08 I'm wondering if this isn't all the captured waveforms overlaid on top of each other.

[PA0PBZ]

About the XY mode and Dave being disappointed that there was no indication on the BNC connectors:

A) You're lucky because they would have no means of swapping it in the firmware update
B) It is clearly indicated on the XY screen itself with the yellow 1  and blue 2

[nitro2k01]

PA0PBZ, I think you completely missed the point.

[PA0PBZ]

PA0PBZ, I think you completely missed the point.

Not unthinkable, care to educate me?

[Monkeh]

PA0PBZ, I think you completely missed the point.

Not unthinkable, care to educate me?

It's XY mode, not YX mode.

[Scarionn]

Please do a full review on this oscilloscope.
Maybe you can compare this oscilloscope with the DS1000 and with your 3000 scope!
 

[nitro2k01]

PA0PBZ, I think you completely missed the point.

Not unthinkable, care to educate me?

He criticized that the X and Y axis used the wrong inputs, not the inputs are unmarked. And this was fixed with the upgrade.

[Psi]

When Dave was talking about the trial features and their timelimit was anyone else thinking..  Hm.. i wonder, if i read out the eeprom/flash ICs before i turned it on for the first time, could i then reprogram it every month to get the features back

It's got to be stored somewhere. If you made a backup of everything before starting you could return it to that state whenever you wanted.

[tom66]

I also wonder if the different procedures might have something to do with it - if you do it the other way around does it reset the trial licences?

[mikeselectricstuff]

When Dave was talking about the trial features and their timelimit was anyone else thinking..  Hm.. i wonder, if i read out the eeprom/flash ICs before i turned it on for the first time, could i then reprogram it every month to get the features back

It's got to be stored in flash somewhere. If you made a backup of everything before starting you could return it to that state whenever you wanted.

I wonder if a simpler method might be to just stop the 32K clock oscillator - very easy to do.
 
Having trials enabled on a new instrument that stop working might raise some conflicts with consumer legislation in some countries - if it isn't made completely clear at the point of sale it could be seen as a fault occurring after purchase...
Would be interesting to see how these actually appear to the user - I think it ought to do something like show a splash screen on first powerup asking the user to confirm that they want a time-lmited free trial.

[Bored@Work]

BUT this also shows that they don't do ANY form of testing that is worth to be called that.

That or their employees have no idea how an oscilloscope is supposed to behave. Or they saw it, but the corporate culture is such that no one dared to mention it. Or it was mentioned, but ignored. All equally worrying.

Having trials enabled on a new instrument that stop working might raise some conflicts with consumer legislation in some countries - if it isn't made completely clear at the point of sale it could be seen as a fault occurring after purchase...

But Rigol is sitting in China. They wouldn't be the first Chinese company giving a flying fart about legislation outside of China. It could hit a few distributors who would then probably have to refund customers. But that requires that a consumer complains.

[DarkPrince]

Great video Dave. Would be nice to grab one of these (do like the screen). Maybe once classes are done I could save up some and get one. Give them time to work out the kinks in the firmware.

[PA0PBZ]

PA0PBZ, I think you completely missed the point.

Not unthinkable, care to educate me?

He criticized that the X and Y axis used the wrong inputs, not the inputs are unmarked. And this was fixed with the upgrade.

I know, but listen around 4:45...

[tom66]

That or their employees have no idea how an oscilloscope is supposed to behave. Or they saw it, but the corporate culture is such that no one dared to mention it. Or it was mentioned, but ignored. All equally worrying.

Or, the third option: The design people knew how the scope was supposed to operate, but the software engineers didn't. They were told to write software to a specification. If you do that,  you could easily make a simple mistake like swapping X and Y.

[ejeffrey]

Or, the third option: The design people knew how the scope was supposed to operate, but the software engineers didn't. They were told to write software to a specification. If you do that,  you could easily make a simple mistake like swapping X and Y.

Anyone could have made the software bug.  The failure is where testing failed to catch it, or caught it but failed to lead to a fix.  There are a number of plausible ways this could have happened.  Some are more troubling than others, but none of them are a good sign.  That said, the big three tend to not pay much attention to XY mode, I would be surprised if one of them let something like this slip through the cracks, especially if it was a result of a last-minute fix for a more "important" feature (rather than defective from day 1).

[mikeselectricstuff]

But Rigol is sitting in China. They wouldn't be the first Chinese company giving a flying fart about legislation outside of China. It could hit a few distributors who would then probably have to refund customers. But that requires that a consumer complains.

Rigol has a presence in Europe and USA and are pushing hard to get into these markets - There was an interview with one of their European managers in one of the trade mags recently - I think he said that only half their sales are to China.

[Bored@Work]

But Rigol is sitting in China. They wouldn't be the first Chinese company giving a flying fart about legislation outside of China. It could hit a few distributors who would then probably have to refund customers. But that requires that a consumer complains.

Rigol has a presence in Europe and USA and are pushing hard to get into these markets - There was an interview with one of their European managers in one of the trade mags recently - I think he said that only half their sales are to China.

I have some doubt about the "pushing hard" thing when it comes to Europe. Unless they have changed things my understanding is their "push" in Europe consists of said manager sitting somewhere in an office in Germany giving interviews, plus some administrative staff and a single tech to support the whole of Europe. Rigol's main web site, when switched to Europe http://eu.rigol.com/ doesn't even give a support phone number.

My understanding is the European Rigol company is more concerned with managing some inventory for distributors than with selling to consumers. And as long as they don't sell directly to consumers they wouldn't have to deal directly with consumer claims. They might probably have to hold a distributor harmless, or not, depending on the contracts they have. But would the decision makers in China care?

[Rosendorfer]

Hi Dave,  gals and guys..

If I may apart from firmware flaws and quirks.....

I'm interested in that low noise thing of DS2000 series, and have to say that I'm left with bit of disappointment, from Dave "playing with"....

Noise at 500uV being 1 division is hardly good achievement, just by simply comparing to that what DS1052e is doing in that regard, I was clearly expecting much better noise figures.
Just for reference I will attach screens from my DS1052e with bandwidth limit at 20MHz and at full 100MHz.
Any chance Dave, or somebody else with DS2000 could investigate that, bit more, Is 500uV real thing or just marketing number... without real backup.

Rosendorfer

[T4P]

You're talking 100MHz input bandwidth. This is a 200MHz And that can dependent on a boatload of issues at that level
The key thing is the resolution ... 500uV gives you lots of resolutions of mV signals and it doesn't matter if the DS1052e could do about the same noise floor or not, does it have the 500uV range? It doesn't.

6:50 "Pleas don't remove the U disk..." I'm wondering why they didn't spell out the word USB.
9:00 The hardware version probably refers to the board revision.
25:08 I'm wondering if this isn't all the captured waveforms overlaid on top of each other.

U disk ... the chinese translated to english equivalent of a USB flash drive 

[EEVblog]

I also forgot to try the HiRes mode at 500uV/div, I believe it's much improved over normal mode.
Did John South from Emona show off that in the Electronex show video?
Edit: yes he did, at 1:30


Dave.

[EEVblog]

Any chance Dave, or somebody else with DS2000 could investigate that, bit more, Is 500uV real thing or just marketing number... without real backup.

It's the real deal.
In the teardown you could see the extra regulators for the front end circuitry.
Rigol have also said they redesigned the input circuitry to achieve this low noise floor.
Would be interesting to compare directly with my Tek 2225.

Dave.

[saturation]

Vertical amp noise increases with bandwidth, so to keep it similar to 50 MHz on 200 MHz scope is noteable.  The vertical amps on the 1052e are used up to 150 MHz [ e.g. on the rarely seen 1152e variant of the 1000 series]  so design wise there is not as big a difference between them on the input section.

I'm interested in that low noise thing of DS2000 series, and have to say that I'm left with bit of disappointment, from Dave "playing with"....

Noise at 500uV being 1 division is hardly good achievement, just by simply comparing to that what DS1052e is doing in that regard, I was clearly expecting much better noise figures.

[Dread]

As an owner of Owon 8202V I was shocked as I watched this Video.  Before I go into “why” let me make it clear that I aspire to own a 200MHZ version of the DS2000.  I just don’t like having to choose between a $2500+ scope and my Wife killing me!   I needed to stay Sub $1K in order to live, but if I had ALL the money for a scope like Dave’s I would honestly still buy it even after watching Dave’s video.

Here’s what really bothered me (In Order)

1)  What!! No upgrade for the BW after purchase   I though that was one of the main upgrades!  I was hoping that one day I could buy the 70 MHz version and save up later for the 200 MHZ upgrade but now all the 70 MHz owners are stuck   This is the number one turn off for me.

2) What no Zoom at all?  Am I missing something here?  I can Zoom with my cheap scope I would expect this one to do that easily.  It was unclear to me if this feature worked before the upgrade or not but I would imagine that it just does not have this feature because Rigol would be asses to fix the XY and at the same time cripple the much more used Zoom Feature.

3)  The whole 50,000 waveform capture thing now seems to me to be more of a gimmick than a real feature.  My Owon which is supposed to have a terrible WC rate has 2Gs at 10M and it does exactly what the Rigol does except in one burst when triggered.   The DS2000 just seems to use a Trigger then Write method.  So each time the unit gets triggered it stores the waveform and stops then waits for the next trigger, etc etc until the memory is used up.    It seems to me that any scope manufacturer could just rewrite their firmware to do the same thing but they instead take one snapshot for the duration of the memory.  Now how fast the DS2000 can reset for the next snapshot may be key but Dave’s video was on a very slow event, I suspect on a fast signal like 100 MHz you would see just about no difference in what you captured on a DS2000 vs any other scope that had the same memory depth and sample rates.


I am no expert on DSO's so please if I am wrong I would like to know what assumptions I made errors on.

[johnboxall]

2) What no Zoom at all?  Am I missing something here?  I can Zoom with my cheap scope I would expect this one to do that easily.  It was unclear to me if this feature worked before the upgrade or not but I would imagine that it just does not have this feature because Rigol would be asses to fix the XY and at the same time cripple the much more used Zoom Feature.

You can zoom after sampling. Page 4-5 of manual.

[Hypernova]

The waveform capture thing looks like a cheap copy of Agilent's Segmented Memory feature, granted it's free (vs buying the SEGMEM license). With SEGMEM you choose how many segments you want and depending on how much sample memory each segment gets the scope drops the sample rate. This IMO is better as you can decide for yourself  the tradeoff between sample rate and segment count.

[taemun]

In the first bout of waveform capturing (~20:30), trigger is on Ch1, and set to 150mV and Ch1 to 500mV/div, but it seems to be that it's triggering on nothing - about half of the captured waveforms are dead flat.

Apparently, to Rigol, a rising edge trigger at 150mV is a "something happened near 150mV" trigger.

My DS1052e++ exhibits a similar waveform update problem. Not sure if there is newer firmware available, though.

[EEVblog]

1)  What!! No upgrade for the BW after purchase   I though that was one of the main upgrades!  I was hoping that one day I could buy the 70 MHz version and save up later for the 200 MHZ upgrade but now all the 70 MHz owners are stuck   This is the number one turn off for me.

Yes, very strange, and quite annoying.

2) What no Zoom at all?  Am I missing something here?  I can Zoom with my cheap scope I would expect this one to do that easily.  It was unclear to me if this feature worked before the upgrade or not but I would imagine that it just does not have this feature because Rigol would be asses to fix the XY and at the same time cripple the much more used Zoom Feature.

John South at Emona told me he's pretty sure that zoom works as it's supposed to, and there is possibly something wrong with my unit. He will confirm this when back from holiday. Can anyone else with a DS2000 confirm?
He also said there is something wrong with my dim LED's as well.


Dave.

[drieg]

There is a bug in FW 00.00.01.00.02 and earlier (similar to DS4000 units) causing rewritting some areas in memory (trial options for example) when doing Self-Cal, saving setups/ref signal, AND also doing firmware update from the mormal operation (GUI).

This bug has been fixed in FW 00.00.01.00.05 (there are still some bugs, axes in XY mode, trial options get expired after Self-Cal,..), so I suggest anyone with FW 01.00.02  to update to latest version. BUT, it is very IMPORTANT to do this update during boot process, not from the normal operation from the menu. Insert USB stick with the firmware, power-on the unit and pres shortly HELP button (immediately after POWER button before RIGOL logo appears). If it doesn't work, try again. If it still doesn't work, keep tying until it works. After you succeed, the SINGLE button will lights and update process will start shortly (CH1 LED blinks), after it's done, all LEDs light. Power-off, remove USB and power-on again.

Hint: To get full FW version go to trigger menu and set Edge, then press F7-F6-F7-Utility button combination quickly. Check additional info under Utility > System > System Info. To escape from this "special" mode, do again  F7-F6-F7-Utility under trigger menu.

Rigol firmwares starting with 01.xx.. (like in Dave's video 01.00.01.09.00) are testing firmwares.

[mikeselectricstuff]

The waveform capture thing looks like a cheap copy of Agilent's Segmented Memory feature, granted it's free (vs buying the SEGMEM license). With SEGMEM you choose how many segments you want and depending on how much sample memory each segment gets the scope drops the sample rate. This IMO is better as you can decide for yourself  the tradeoff between sample rate and segment count.

Does the Rigol not allow you to select the number of segments? If not this is a major limitation - being able to trade off sample length against number of samples is a very important facility for segmented memory acquisition.

[EEVblog]

Does the Rigol not allow you to select the number of segments? If not this is a major limitation - being able to trade off sample length against number of samples is a very important facility for segmented memory acquisition.

The Rigol feature is a bit different to Agilent's segmented memory feature.
No, I don't think you can select it on the Rigol, as the sample memory depth is automatically controlled based on timebase like on any scope. And that's the same memory depth regardless of whether the feature is turned on or off. Leave it turned off, and you just waste the memory. Turn it on and you use the wasted memory to capture extra waveforms (at the expense of update speed I presume).
The Agilent 2000X has a relatively slow 20us re-arm time for segmented memory (1us on 3000X), so is more designed for infrequent short-burst waveforms.  Not sure what the equivalent time would be on the Rigol feature.

Dave.

[marmad]

3)  The whole 50,000 waveform capture thing now seems to me to be more of a gimmick than a real feature.

I don't think you're understanding waveform update rates. Here are two good explanations of it:

cp.literature.agilent.com/litweb/pdf/5989-7885EN.pdf
www2.rohde-schwarz.com/file/1ER02_1e.pdf

To give you a concrete example of the difference between the Owon and Rigol's rates:

Scope’s blind-time percentage
= 100 x [(1/U) – W]/(1/U)
= 100 x (1 – UW)
where
U = Scope’s measured update rate
and
W = Display acquisition window = Timebase setting x number of divisions (which is 20 on the Owon and 15 on the Rigol).

For example, at a 100ns/div timebase setting:
The 30 wfrm/s update rate of the Owon means a blind-time of 99.994% - in other words, it's only showing .006% of the real-time waveform.
Assuming the Rigol does 50,000 wfrm/s at this timebase setting, the blind-time is 92.5% -  i.e., it's showing 7.5% of the real-time waveform.

What does this mean in terms of glitch hunting? According to the formula, then the average time for catching a repeating signal fault (let's say 10 glitches per second at 100ns/div) with a probability of 99.9%:
Owon: ~3 hours and 12 minutes.
Rigol: 9.2 seconds

[drieg]

Does the Rigol not allow you to select the number of segments? If not this is a major limitation - being able to trade off sample length against number of samples is a very important facility for segmented memory acquisition.

..No, I don't think you can select it on the Rigol, as the sample memory depth is automatically controlled based on timebase like on any scope.

You can select sampling memory depth under Acquire button/menu and see, how the waveform capture rate changes...

[rr100]

When Dave was talking about the trial features and their timelimit was anyone else thinking..  Hm.. i wonder, if i read out the eeprom/flash ICs before i turned it on for the first time, could i then reprogram it every month to get the features back

That's clear and if there are more than a few users be sure somebody will find a way to do it.
However I've been thinking something else: they go through all the hoops of making and delivering some features and then they write extra code to actually remove them from the majority of the customers hoping some will come with the cash asking to have them re-enabled. This really rubs me the wrong way. I know is standard practice for med-highly expensive equipment, the one that's more like a computer than like a "normal" tool. But still I can't get rid of this feeling of feeling cheated if I would buy a hammer (or a multimeter or a scope) that wouldn't work as well as it can just because I didn't pay some "extra license to enable all features".

[cidcorp]

Ok, I'm lost - where the hell is everyone getting updated firmware for the DS2000 line?

If someone can point me to a copy of the firmware it would be greatly appreciated.

I haven't done a self cal on the device yet cause I'm evaluating the add-on options, so I'd be very concerned the firmware update would reset these trials...

Side note:

The screen update issue that is shown in the video occurs here on my DS2102, same software
and hardware revisions as Dave's 200Mhz model.

[elbot]

There is a bug in FW 00.00.01.00.02 and earlier (similar to DS4000 units) causing rewritting some areas in memory (trial options for example) when doing Self-Cal, saving setups/ref signal, AND also doing firmware update from the mormal operation.

This bug has been fixed in FW 00.00.01.00.05, so I suggest anyone with this firmware to update to latest version. BUT, it is very IMPORTANT to do this update during boot process, not from the normal operation from the menu. First, iyou have to nsert USB stick with the firmware file , then power-on the unit and pres shortly HELP button. The SINGLE led will light and update process will start shortly (CH1 LED is flashing), after it's done, all LEDs light. Power-off, remove USB and power-on.

Hints: To get full FW version go to trigger menu and set Edge, then press F7-F6-F7-Utility button combination. Check additional info under Utility > System > System Info. To escape from this "special" mode, do again  F7-F6-F7-Utility under trigger menu.

Rigol firmwares starting with 01.xx.. (like in Dave's video 01.00.01.09.00) are testing firmwares.

Please let us know where to get hands on firmware updates, I'm using a DS2102 with the faulty version FW 00.00.01.00.02
and can't calibrate without loosing trial versions, it will help me alot, thanks...

Claus,

[Rosendorfer]

Hi

Just to follow with 500uV/div noise and not to go into discussions, if it is really extremely low and can be seen as important feature or is just average at best and 500uV/div is just "marketing function". Interesting would be to compare DS2000 against Hameg HMO series, as both are claiming low noise as their important feature.

Anyway interesting how Dave will follow with full review of DS2000, seems that it is bit more tricky than just showing it is well build 200MHz scope for 1000+USD .
Expectations are pretty high.

Rosendorfer



 

[Dread]

3)  The whole 50,000 waveform capture thing now seems to me to be more of a gimmick than a real feature.

I don't think you're understanding waveform update rates. Here are two good explanations of it:

cp.literature.agilent.com/litweb/pdf/5989-7885EN.pdf
www2.rohde-schwarz.com/file/1ER02_1e.pdf

To give you a concrete example of the difference between the Owon and Rigol's rates:

Scope’s blind-time percentage
= 100 x [(1/U) – W]/(1/U)
= 100 x (1 – UW)
where
U = Scope’s measured update rate
and
W = Display acquisition window = Timebase setting x number of divisions (which is 20 on the Owon and 15 on the Rigol).

For example, at a 100ns/div timebase setting:
The 30 wfrm/s update rate of the Owon means a blind-time of 99.994% - in other words, it's only showing .006% of the real-time waveform.
Assuming the Rigol does 50,000 wfrm/s at this timebase setting, the blind-time is 92.5% -  i.e., it's showing 7.5% of the real-time waveform.

What does this mean in terms of glitch hunting? According to the formula, then the average time for catching a repeating signal fault (let's say 10 glitches per second at 100ns/div) with a probability of 99.9%:
Owon: ~3 hours and 12 minutes.
Rigol: 9.2 seconds

Marmad I have noticed that you really seem to be in love with this feature, which I think is still a bit of a over sell by Agilent to try and make their consumer grade scopes look better than everyone else’s.  Personally I think it's just because so many new DSO makers are now in the market place and this has now become a big push for them because they dominate that area, but hey I am no expert so it's just my opinion.   

BTW I had read that Agilent Document before and it really came off to me as a questionable in the sense that to me as a novice in the world of DSO it seems to only addresses one part of the equation and by that I mean doesn’t it also dependent on how well your scope triggers.   It would seem to me that your scope could be capturing 10M/WFS but if it's trigger does not go off during the glitch then that’s all for naught.  Please let me know if I am wrong but to me that seems logical.

My second comment on all of this comes from 15 years of doing repair work.  For me weird Glitches were very rare in the repair world and when I did come across it; most of the time I could just use my old Analog scope and Tap or twist the board and bingo the glitch would appear or disappear.  If that plus a few other tricks did not work I typically skipped those jobs because I was not going to spend a whole day on those kind of problems, it was never going to be worth my time unless it was a darn expensive piece of gear.   I admit that these new DSO with great waveform capture might have done the trick but as I said before it was a very rare case when I came across a signal fidelity problem that was not looking me straight in the face.

[Dread]

1)  What!! No upgrade for the BW after purchase   I though that was one of the main upgrades!  I was hoping that one day I could buy the 70 MHz version and save up later for the 200 MHZ upgrade but now all the 70 MHz owners are stuck   This is the number one turn off for me.

Yes, very strange, and quite annoying.

Thanks for the Update Dave.  Please let us know what you find out.
2) What no Zoom at all?  Am I missing something here?  I can Zoom with my cheap scope I would expect this one to do that easily.  It was unclear to me if this feature worked before the upgrade or not but I would imagine that it just does not have this feature because Rigol would be asses to fix the XY and at the same time cripple the much more used Zoom Feature.

John South at Emona told me he's pretty sure that zoom works as it's supposed to, and there is possibly something wrong with my unit. He will confirm this when back from holiday. Can anyone else with a DS2000 confirm?
He also said there is something wrong with my dim LED's as well.


Dave.

[EEVblog]

BTW I had read that Agilent Document before and it really came off to me as a questionable in the sense that to me as a novice in the world of DSO it seems to only addresses one part of the equation and by that I mean doesn’t it also dependent on how well your scope triggers.   It would seem to me that your scope could be capturing 10M/WFS but if it's trigger does not go off during the glitch then that’s all for naught. 

Generally, not so.
For example, you are using the infinite persistence function to try and capture an unknown glitch you think might be there but cannot trigger off. You have to know something about the glitch in order to reliably set up the trigger to catch it on trigger, so that's what infinite persistence is for.
If that glitch occurs only once every million cycles, the Agilent 3000X can in theory catch is as quickly as 1 second. Other scopes you could wait an eternity. I have demoed this in an Agilent video.

Of course, if you do know how to trigger from the glitch, then the update rate is essentially irrelevant.

Dave.

[ejeffrey]

My second comment on all of this comes from 15 years of doing repair work.  For me weird Glitches were very rare in the repair world and when I did come across it; most of the time I could just use my old Analog scope and Tap or twist the board and bingo the glitch would appear or disappear.

Repair work is a very different beast than development.  Glitches of the type that a high waveform capture rate helps with are almost exclusively caused by design errors not hardware failures.  If they show up in the field, it means there was insufficient testing for e.g., timing variation vs. temperature and you should just send the board back to the designer.  As a corollary, a glitch that results from a design error cannot usually be fixed by swapping out with a known good board so you really have to find the problem.

[marmad]

Marmad I have noticed that you really seem to be in love with this feature, which I think is still a bit of a over sell by Agilent to try and make their consumer grade scopes look better than everyone else’s.  Personally I think it's just because so many new DSO makers are now in the market place and this has now become a big push for them because they dominate that area, but hey I am no expert so it's just my opinion.   

BTW I had read that Agilent Document before and it really came off to me as a questionable in the sense that to me as a novice in the world of DSO it seems to only addresses one part of the equation and by that I mean doesn’t it also dependent on how well your scope triggers.   It would seem to me that your scope could be capturing 10M/WFS but if it's trigger does not go off during the glitch then that’s all for naught.  Please let me know if I am wrong but to me that seems logical.

Dread: the dead/blind-time (and related waveform update rate) is not a FEATURE of a DSO - it's an inherent fault in them; i.e. the fact that the instrument must spend much more time processing the data for display than it can spend gathering it (whereas an analogue scope only needs to retrace the beam). The scope is effectively dead during this period - so it doesn't respond to triggers at all.

For me weird Glitches were very rare in the repair world and when I did come across it...

The point is this: with a slow wfrm/s rate, you could hook your scope up to a signal and stare at it for 3 solid hours and not realize that there is a glitch in the signal happening once per second.

Of course, if you do know how to trigger from the glitch, then the update rate is essentially irrelevant.

Again, this is only true if the scope is stopped, waiting to trigger from the glitch. If it's running (and has a slow waveform update rate), it will almost certainly miss the glitch. This is easily proved with a Mask test (which is often closely related to the scope's wfrm/s rate).

Edit: BTW, Dave, you mentioned in your vid that the Rigol 1000 series captures around 700-800 wfrm/s. In fact, the E series is somewhere in the 50-200 wfrm/s range, while the CA series is up to 2000 wfrm/s (hence the massive price difference).

[EEVblog]

Again, this is only true if the scope is stopped, waiting to trigger from the glitch.

That's how you usually capture things you are looking for! 

Edit: BTW, Dave, you mentioned in your vid that the Rigol 1000 series captures around 700-800 wfrm/s. In fact, the E series is somewhere in the 50-200 wfrm/s range, while the CA series is up to 2000 wfrm/s (hence the massive price difference).

I thought I measured it around the 700 mark at one point, but my memory could be wrong.

Dave.

[marmad]

That's how you usually capture things you are looking for! 

Absolutely - when you're looking for a known problem.

I thought I measured it around the 700 mark at one point, but my memory could be wrong.

It's probability theory: you can't find out how fast a scope is (wfrms/s) by running a glitch test a couple of times (as you did in your Agilent review) - you can only prove how fast it isn't.  For example (using the probability formula for a 10/s glitch with a timebase of 100ns), if my scope captures the glitch within 7 seconds - every now and then, or even a few times in a row - it doesn't prove its a 1,000,000 wfms/s scope.  But if one time it takes my scope over 10 minutes to capture the glitch, it does prove (beyond a reasonable doubt) that it isn't a 1M or 100k wfrms/s scope, because the probability of a scope that fast taking that long to capture the glitch is extremely low (99.9999...%).  If you want to try to calculate actual wfrms/s by testing the probability of capturing a random glitch, you have to run many tests - not just a few 

[EEVblog]

If you want to try to calculate actual wfrms/s by testing the probability of capturing a random glitch, you have to run many tests - not just a few 

I ran more tests than what made it into the final edit!

Dave.

[marmad]

I ran more tests than what made it into the final edit!

I'm responding to this comment in your Agilent review video:
'And I just did a second one (test) and it also took about a minute, so you can actually conclude that this Rigol, if you do the math...  that's just over 800 waveforms a second...'
This is incorrect.

I have not seen any credible evidence anywhere as to a several hundred wfrm/s rate for the Rigol E series, and in fact, all of the circumstantial evidence (wfrm/s rates of similarly priced or designed scopes -  and much more expensive scopes) points to a much lower rate.

But I'm certainly open to hearing about any highly intensive tests that were run (as I did with the Owon SDS7102 and Hantek DSO5062B).

[Rosendorfer]

Hi Dave

Any chance You will look into 12bit HR mode of DS2000 series ?
Still not many info how in real this all HR stuff on scopes is working.

Seems that it is going to be very long review.....

Rosendorfer

[AlphZeta]

6:50 "Pleas don't remove the U disk..." I'm wondering why they didn't spell out the word USB.

Aha, this is because in China USB thumb drives are referred to as "U disk" so my suspicion is that the software was written by a Chinese programmer...

[Dread]

A big thanks to Dave and Marmad, I have learned a lot more about capture rates since reading this thread.  One thing Marmad don't you agree with Dave that triggering is the way to find the glitches?
By that I mean using the pass fail method and letting the scope find the glitch? The whole idea of one glitch a second does not seem typical, it may be one every hour, how does 1mwfsc help in that case, you would be waiting for months by your calculations.

[EEVblog]

A big thanks to Dave and Marmad, I have learned a lot more about capture rates since reading this thread.  One thing Marmad don't you agree with Dave that triggering is the way to find the glitches?

Triggering is not a way to find the glitches, it's a way to examine them once you have found them.
Triggering on something you don't know exists doesn't work (or is just pot luck).
And if you are able to trigger off them, then the update rate is not relevant, because it's a single shot trigger.
Fast updating (and persistence mode + peak detect) is a way to find glitches and other things that you don't know exist, but you suspect might be there.
Having an analog scope is not magic in this regard, as you can easily miss glitches and runt pulses, even if you have the brightness turned all the way up. So in many cases, a digital scope with infinite persistence can be far superior than an analog scope in this regard. Even one with relatively poor update speed (if you leave it looking long enough)

Dave.

[Dread]

That certainly sounds like the best way Dave but don't you think the Pass/Fail could also work?
I imagine that on a low end DSO that's going to run into the same capture rate issue??
BTW Dave can you setup a standardized test on a video that anybody with a DSO and a arb generator could use to test their scope?  It would give all of us great data to publish in the videos forum thread and since you have several dso we could see how our cheapies stack up in an apples to apples comparison of wfcr.
 

[vk6zgo]

My second comment on all of this comes from 15 years of doing repair work.  For me weird Glitches were very rare in the repair world and when I did come across it; most of the time I could just use my old Analog scope and Tap or twist the board and bingo the glitch would appear or disappear.

Repair work is a very different beast than development.  Glitches of the type that a high waveform capture rate helps with are almost exclusively caused by design errors not hardware failures.  If they show up in the field, it means there was insufficient testing for e.g., timing variation vs. temperature and you should just send the board back to the designer.  As a corollary, a glitch that results from a design error cannot usually be fixed by swapping out with a known good board so you really have to find the problem.

This is why I maintain that DSOs are not the automatic instrument of choice for all Electronic testing.
The features which are essential for development work,are pretty much incidental for most repair work.

Of course,most people in the maintenance side have been confronted by non-functional new equipment.
As you logically assume "It must have worked once", it takes some time before it dawns on you that they never could have worked.

In many cases,it is not cost effective to send them back,& the "Technical Assistance" from (guess what country?)
is useless,so repair morphs into development,where all the "bells & whistles" of a nice DSO would come in handy.

These are very much the minority of cases,in most circumstances the "bash the board" trick works.

[Gunb]

That certainly sounds like the best way Dave but don't you think the Pass/Fail could also work?
I imagine that on a low end DSO that's going to run into the same capture rate issue??
BTW Dave can you setup a standardized test on a video that anybody with a DSO and a arb generator could use to test their scope?  It would give all of us great data to publish in the videos forum thread and since you have several dso we could see how our cheapies stack up in an apples to apples comparison of wfcr.

I wonder always that the discussion about wfm/s starts in many forums again and again and reveals the lack of knowledge about digiscopes' basics. The Pass/Fail test is intended to be used to find deviation of a waveform and to define a tolerance band around a given signal. It never can replace methods searching for outliers and there are a few reasons why.

Let's assume to have a simple sine and a tolerance band of 100mV necessary not to trigger the fail function caused be noise - a glitch of 99mV would not be detected even when it appears.

The other reason might be that a mask has absolutely no influence on the screen refresh rate. Coming back to the example above, even if the tolerance band would allow a 50mV band around the sine and the glitch has got an amplitude of 99mV what if it appears in the dead time between two following screens? Your mask won't help because the scope's refresh rate is to slow.

Catching a glitch with a digiscope in persistence mode always deals with the probability to detect it between two screens - and the probability to detect it raising with the screens taken per second - the wfm/s.

In the past Agilent has made some detailed reports about this issue with the calculation done for scopes with low wfm/s and higher wfm/s. It reveals simply how many hours you might need to get seldom glitches with the slow scopes.

Nothing to remark according triggering on glitches - Dave has already mentioned the details.


Kind regards
Gunb

[Lawsen]

The video almost inspired me to try one, the Rigol DS2202, the 200 MHz oscilloscope.  I liked the wide screen, reminds me of the Agilent DSOX.  Rigol of China has limited the sales of their latest Rigol DS2202 away from E Bay to protect competition from their dealerships in the U.S.A.  The Rigol DS2202 is not available in E Bay.  The Rigol DS2202 is priced $800.00 AUD in Australia as Dave Jones mentioned.  It is priced more in the U.S.A. at distributors in the U.S.A. for $1,600.00 USD.  I do like the angled design panel and the large wave form play back knob, remind me of the Yokogawa DL1502 oscilloscope, but the Yokogawa is an obsolete amber tube display oscilloscope.  There are flaws in the Rigol DS2202 as you shown in your video.  Rigol seems to be trying to update or improve their firmware.  I am happy with the Rigol 1052E.  I would like to try the little more priced Rigol, the 1102E.  I like Yokogawa, but I cannot afford them.  That 100 millionth picosecond is neat in the Rigol DS2202.  Thanks for the video, neat.  For my use, LR of geophysics metal mineral locator coil testing with signal generator, I do not need a Rigol DS2202.   

[marmad]

One thing Marmad don't you agree with Dave that triggering is the way to find the glitches?

I agree exactly with what Dave wrote in his last comment about triggering, fast updating, and analog scopes. Also with the good points made by ejeffrey about development work vs. repair work.

That certainly sounds like the best way Dave but don't you think the Pass/Fail could also work?
I imagine that on a low end DSO that's going to run into the same capture rate issue??
BTW Dave can you setup a standardized test on a video that anybody with a DSO and a arb generator could use to test their scope?  It would give all of us great data to publish in the videos forum thread and since you have several dso we could see how our cheapies stack up in an apples to apples comparison of wfcr.

I wonder always that the discussion about wfm/s starts in many forums again and again and reveals the lack of knowledge about digiscopes' basics. The Pass/Fail test is intended to be used to find deviation of a waveform and to define a tolerance band around a given signal. It never can replace methods searching for outliers and there are a few reasons why.

I'm wondering if what I wrote earlier about using a Mask test was misinterpreted. I never intended to suggest it was a method to search for glitches - in fact, just the opposite: it's a way to prove how bad the scope is at 'seeing' glitches. But this has to be taken with a grain of salt - since the Mask test adds overhead to the software before screen display, it normally decreases the wfrm/s rate (although Agilent claims that the 2000/3000X are the industry's only scopes to do the test completely in hardware, thus negating the time penalty in the 2000X). But in the case of the Owon SDS line, the wfrm/s update rate is tied to the screen refresh rate so the Mask test does not alter it substantially.

The Rigol DS2202 is priced $800.00 AUD in Australia

Huh!?! Please point me to an Australian website selling the DS2202 for $800 immediately - because it's twice as much everywhere else on the planet!

[jeeff47]

Hello everyone,

I was hoping you guys could provide me with some feedback:

I am looking at purchasing the Rigol DS2072 but I am a little hesitant as i am not very familiar with the Rigol brand. I have been doing a lot of researching and so far i am impressed with the specs and the current hardware quality of the DS2027 based off Daves tear down video and other peoples comments; but I don't know what to think of the product in terms of Software.

It appears that  the computer software isn't as nice as some of the competitors and watching your most recent video Dave, it does seem like there are some firmware bugs that still need to be worked out.

However I do like that fact that it does have the 8.5" screen, Trigger on Serial data, 2GSPS, 14MPT memory, 500uV/Div range and the record/playback function.
I also like the fact that the standard scope probes are relatively cheap in case replacements are needed.

The other model I am looking at due to my budget is the Agilent DSO1072B, it has the same bandwidth but 1GSPS and only 16K memory. From my experience the Agilent software is spot on and the products has a good reputation.
What I like about the Agilent is the waveform Zoom Simultaneously feature along with the look and feel of the product I also like the Waveform math and FFT feature. 

Does the Agilent offer a record/playback feature? Would that be apart of the "Sequence mode for easier debug" feature that they have mentioned?

I would consider the DSOX2002A  series which has the 2GPS, 8.5" screen and 100Kpt memory but that is out of my price range ad an additional $200+ I know you can get the function generator feature also but unfortunately I have to draw the line somewhere and that is at the $1000 mark.

Even though I am not overly concerned the overall size. Would the size of the Rigol be noticeable on a small work bench? As my work bench is relatively small it is something I should consider but is of low priority. 

What are peoples opinions on the quality of the Rigol scope probes vs Agilent scope probes is there a quality difference?

Aside from the spec difference between the DS2027 and the DSO1072B how do you guys think the features of the two will compare? are there any features that the Agilent will have that the Rigol wont?

Thanks in advance for all your input/opinions I really appreciate any feed.

Jeff

[tom66]

Don't buy the Agilent DSO1072B.

It's a rebadged Rigol, but with 70 MHz bandwidth instead of 50 MHz. Same as DS1052E/DS1102E, but with a huge price premium. In fact, I think they don't even have the 1 Mpt memory like the DS1052/DS1102E have.

Yes, Agilent trust Rigol for making their low end scopes.

That should give you an idea about the quality they have.

I have a DS1102E and I do like it. It's a very capable scope.

[EEVblog]

Don't buy the Agilent DSO1072B.

I agree. You are just buying an overpriced Rigol.

The DS2000 is an awesome scope for the money, it's got to be the best value scope in that $800-$1200 price range.
And I'm sure the software bugs will be fixed, if they haven't already, I'm just waiting for some updated firmware to see if it fixes it. John South said he'll have some more new firmware today.

Dave.

[Shas-O]

Hi Jeff

Record/play back is called segmented memory in the Agilent world, I have it on the 3000 series scopes at work.
It is a charged option.

Yokogawa scopes also have this, and is standard.

Dec

Hello everyone,

I was hoping you guys could provide me with some feedback:

I am looking at purchasing the Rigol DS2072 but I am a little hesitant as i am not very familiar with the Rigol brand. I have been doing a lot of researching and so far i am impressed with the specs and the current hardware quality of the DS2027 based off Daves tear down video and other peoples comments; but I don't know what to think of the product in terms of Software.

It appears that  the computer software isn't as nice as some of the competitors and watching your most recent video Dave, it does seem like there are some firmware bugs that still need to be worked out.

However I do like that fact that it does have the 8.5" screen, Trigger on Serial data, 2GSPS, 14MPT memory, 500uV/Div range and the record/playback function.
I also like the fact that the standard scope probes are relatively cheap in case replacements are needed.

The other model I am looking at due to my budget is the Agilent DSO1072B, it has the same bandwidth but 1GSPS and only 16K memory. From my experience the Agilent software is spot on and the products has a good reputation.
What I like about the Agilent is the waveform Zoom Simultaneously feature along with the look and feel of the product I also like the Waveform math and FFT feature. 

Does the Agilent offer a record/playback feature? Would that be apart of the "Sequence mode for easier debug" feature that they have mentioned?

I would consider the DSOX2002A  series which has the 2GPS, 8.5" screen and 100Kpt memory but that is out of my price range ad an additional $200+ I know you can get the function generator feature also but unfortunately I have to draw the line somewhere and that is at the $1000 mark.

Even though I am not overly concerned the overall size. Would the size of the Rigol be noticeable on a small work bench? As my work bench is relatively small it is something I should consider but is of low priority. 

What are peoples opinions on the quality of the Rigol scope probes vs Agilent scope probes is there a quality difference?

Aside from the spec difference between the DS2027 and the DSO1072B how do you guys think the features of the two will compare? are there any features that the Agilent will have that the Rigol wont?

Thanks in advance for all your input/opinions I really appreciate any feed.

Jeff

[jeeff47]

I did hear lower end Agilents were repackaged Rigols but I was under the impression that was a few years back and might not be the same anymore.

Has anybody identified which Agilent models are repackaged Rigols? Does Agilent simply load its own firmware onto the devices or does it even use Rigol firmware (with slight modifications such as changing Rigol to Agilent in text)? Has anybody purchased a Rigol scope that can be repackaged and try loading the Agilent f/w into it? (just some thoughts)
If the Software is better (more user friendly, less buggy) I could see a bit of a price increase but I dont know about double the price.

I was looking at the DS1102E but I was under the impression that it was a few years old and was hoping for something which had been released fairly recently. I also heard that it has issues with fan noise.

What are the scope probes like? Do they seem like decent quality?

Has anybody identified what the buttons on the left side of the screen do?

Dave, I am really excited to see a full review of the DS2072 when you get the opportunity. I am very interested in seeing someone operate the device as I have been informed that Rigol doesn't demo this unit.

Thanks again everyone


Jeff

- I looked in the user manual and it appears that the buttons to the left of the screen are for the
"one key measurement of 24 parameters" ie. quick measurement buttons? Has anybody tested these out? do the "increase productivity"?

[drieg]

Agilent DSO1004A = Rigol DS1064B
Agilent DSO1014A = Rigol DS1104B
Agilent DSO1024A = Rigol DS1204B

..for example. All is done by Rigol including firmware and software, there are only cosmetic changes (splash screen, menu colors). Actually the firmware of Rigol units was a bit ahead in terms of bug fixes and new features. I do agree with others, there is really no reason to spend more money on a Rigol with Agilent logo

DS2072 is really nice and very promising unit, I can only recommend..

[tom66]

I was looking at the DS1102E but I was under the impression that it was a few years old and was hoping for something which had been released fairly recently. I also heard that it has issues with fan noise.

I don't mind the fan... on mine it's quiet.

What are the scope probes like? Do they seem like decent quality?

They feel quite good and work well.

[jeeff47]

I would guess they are not but I thought I would ask... Can you use an Agilent scope probe on a Rigol unit?

[cidcorp]

I contacted Rigol North America to see if these was any information on new firmware for the DS2000 series units.  I also brought up the X <-> Y label swap.  To be honest
I likely wouldn't of noticed as it's a feature I would rarely use.  After some time the rep from the Distributor called me trying to set up a conference call between me and the tech
in the US (I'm in Canada) at Rigol.  Unfortunately I'm currently very busy during the day, and didn't get a chance to call my rep back.

I did though receive an email a couple days later in which the tech described the CH1=Y and CH2=X as by design even though it was not 'standard' convention with other manufacturers.
He also indicated that he is not aware of any firmware revisions to the scope as of this moment, but promised to look into it further.  I'm hoping this isn't a glimpse of what there is to come over the horizon, i.e.: difficult to obtain firmware updates.  No offense to Dave, but some of us don't carry the weight to get noticed, scoring none milestone revision or test firmware.

I'd like to keep the scope up to date, especially due to the model being in it's infancy.  There are ALOT of things that could be made better with relation to the UI, and adding software upgradable bandwidth would be awesome.

I'm currently evaluating the OPTIONs (trials) on the device, so would hate to update the firmware and lose the trial counters anyway.  I suspect the procedure described using the SINGLE button doesn't work above, as dave did try it on the video.

So looks like I'm out of luck for now with the firmware.  If anyone hears different, please drop me a line.

Chris

[EEVblog]

He also indicated that he is not aware of any firmware revisions to the scope as of this moment, but promised to look into it further.  I'm hoping this isn't a glimpse of what there is to come over the horizon, i.e.: difficult to obtain firmware updates.  No offense to Dave, but some of us don't carry the weight to get noticed, scoring none milestone revision or test firmware.

I'm not getting my firmware direct from Rigol, nor am I requesting it, or even in contact with them about it.
The firmware I got was asked for and requested by the local Oz rep John South from Emona on his own accord. I didn't know about new firmware or the X-Y bug until John told me.
John said he's getting another new firmware update from Rigol on Monday.

Yes, Rigol really need to fix their attitude toward firmware updates, they are very quickly getting a bad rep for it, which is a shame. No one minds firmware bugs (in fact it's good, it shows the company is actively finding, fixing and improving things), but companies need to have a clear central and public repository for it.

Dave.

[cidcorp]

I totally understand the occasional revision being done to fix a very specific problem which is not necessary for the MASS market, but when the device is so young you'd expect a great deal of revisions in the first few months slowly trickling down to none.  That point you figure you have a mature stable product.

I think the same reason they didn't include the Bandwidth upgrades as an option is the same thing that fueling them being tight fisted with the firmware revisions.  They're scared the public will end up hacking the options find an easy 'free' upgrade to the higher bandwidth models.

Hasn't Agilent stopped doing the customer activated bandwidth upgrades?  I thought I heard they were stopping last month and it would require the unit to be returned to Agilent...just rumor... might be totally false. (Referring to the DSOX2k and 3k).

But I do agree they are quickly becoming unpopular in my books... I did have high hopes for them until the DS2000 series didn't have a 4 channel option and no software upgradable bandwidth.

We'll have to see how this plays out over the coming months.
Chris

[EEVblog]

Hasn't Agilent stopped doing the customer activated bandwidth upgrades?  I thought I heard they were stopping last month and it would require the unit to be returned to Agilent...just rumor... might be totally false. (Referring to the DSOX2k and 3k).

No, that is just from certain bandwidth, e.g. 200MHz to 300MHz/500MHz, or 300/500MHz to 1GHz. They are different physical front end circuits.
Been that way from day one.

Dave.

[Bored@Work]

I'm not getting my firmware direct from Rigol, nor am I requesting it, or even in contact with them about it.
The firmware I got was asked for and requested by the local Oz rep John South from Emona on his own accord.

But isn't that exactly the problem the rest of us, the unwashed masses without the publicity, have? "We" have no one pro-actively requesting firmware updates for us. No one in the supply chain has an incentive to do that for "us".

Rigol's European office hasn't changed this a bit for customers in Europe. They are effectively hiding and are focusing on managing their warehouse. So it is back to China. It is two problems. Having to deal with Rigol China, and second, what should be a push service (Rigol pushing out new firmwares to their customers) is a pull service (having to beg Rigol for an update). And it is like pulling teeth when you have to deal with Rigol.

Yes, Rigol really need to fix their attitude toward firmware updates, they are very quickly getting a bad rep for it,

"Very quickly getting?" They have that reputation for a long time already. No support and broken promises.

[EEVblog]

But isn't that exactly the problem the rest of us, the unwashed masses without the publicity, have? "We" have no one pro-actively requesting firmware updates for us. No one in the supply chain has an incentive to do that for "us".

No. John did not solicit the firmware update for me, I did not ask for it at any stage, and he did not "pull strings" or anything of the like because I'm involved.  It had absolutely nothing to do with me or the publicity of the blog.
He either requested it for his own scopes he is selling, or Rigol simply offered it to him because he's an official Rigol dealer, and that's what you'd expect them to do with their dealers. BTW, he was the one that found the X-Y issue and got it fixed.
He simply offered me the firmware he got given because he knew I'd be doing a review soon and it might possibly fix some issues.
And BTW, I did not get the scope from him, I got it direct from Rigol NA (US), who presumably did not care enough to follow up and forward me any new firmware.

So in this case I do not have better access to firmware than anyone else, I got it like everyone else likely would, through their local Rigol dealer.
If I was a regular Emona customer buying the scope from John, he'd give me the latest firmware, because that's what good dealers do.

So those who are complaining they can't get Rigol firmware updates, I suggest you ask your local dealer you bought it from. That's one of the value-added services they usually offer, because in this case Rigol themselves seem pretty piss-poor at it directly.

Dave.

[Gunb]

Hi there,

just a moment ago I've watched the video and was curious to see if my DS4012 has got the same bug related to the zoom when recording waveforms - it has not. Works as it should, i.e., when in STOP mode playing around with activating the zoom and turning the big knob both channels are updated fine.

So, only to confirm, there higher model does not have this bug and it shouldn't be there in the DS2000 series.

Thanks for the video.

Kind regards
Gunb

[JOHNJB]

I have a new DS2102 tree weeks ago.
I lost the time in trial options at firs calibrate.
I've receive from Rigol Technologies EU GmbH,  firmware 00.00.01.00.05
It solved some bugs, but not all.
It`s an YX scope, but it runs.

For me the most important bug now is the cursor time positioning.
The cursor manual moves over the screen, not over the wave, if moves the wave, lost the cursor position,
if zooms lost the cursor position.
The cursor in track mode, it seem move over the wave, but it makes incompressible jumps.

Whats your opinion over the cursor time function, is or not a bug ?
Exist a new update over the 01.00.05 ?

John

[jeeff47]

Has anybody tested the zoom issue that Dave experienced on older firmware? Maybe it was something that was accidentally introduced after the firmware upgrade.

[EEVblog]

Has anybody tested the zoom issue that Dave experienced on older firmware? Maybe it was something that was accidentally introduced after the firmware upgrade.

Yes, I just had John South confirm this. It's a bug that was introduced in newer versions of the firmware, of which there seems to be several incarnations. It sounds as if Rigol don't have their act together in the firmware department, and are in fire-fighting mode.

Dave.

[Gunb]

Whats your opinion over the cursor time function, is or not a bug ?
Exist a new update over the 01.00.05 ?

John

If you mean the MANUAL mode then it's not a bug but what it should do: not glue to the signal.

Related to the TRACK mode I would not say it's  not a bug but you're right, it could be done better.
It depends a bit on the timebase how accurate you can set the cursors.

Same with the DS4000 series.


Kind regards
Gunb

[John South]

Hi - I'm the "culprit" behind the firware given to Dave. I also work for Emona Instruments who are the Rigol Distributors in Australia.

I'm posting here to clarify the issues with the firmware given to Dave. This was (As per what Drieg wrote) actually an engineering release not a public release. As such it was not fully tested . It was a release specifically given to me to show corrected XY display to a University here and there was a very short timeframe so it may have been rushed.

At the time I was unaware of it being an engineering release so passed it on to Dave with the introduced bug of non functional zoom in record mode.  I have a new engineering release that fixes this but am waiting on an answer from Rigol on when we expect a public release. I'll post this information here. This was my mistake - Public release firmware is more thoroughly tested. (And yes it could be posted better)

While I'm here a couple of answers to questions that came up on the record mode .

This segments the memory.
Yes you can select the number of segments you want. This is done in 2 ways.

First you can change record length.

For example if I change record length to 14k I can record up to 8128 frames 140k 508 etc.

For existing DS2000 users this is found under Acquire menu --> MemDepth
(Note : My unit has the extended memory depth option so the exact numbers might differ if it is not installed)

Second if I don't want that many I can select the End Frame (Can be anything from 1 to maximum)

For existing DS2000 owners this is found under Utility --> Record --> End Frame.

Frames can be captured by time interval  , captured on trigger or captured an a "mask" test fail.
Frames can be time date stamped and zoomed.

You can also "Analyse" the captured waveforms for "Error" frames by , after capture , selecting , a template from a frame , selecting a % value for error and running analyse. Or you can create a "mask" .

If anyone has any questions on the DS2000's or other Rigol products please feel free to ask me. I will try to keep any sales stuff out of it. 

[EEVblog]

Second if I don't want that many I can select the End Frame (Can be anything from 1 to maximum)
For existing DS2000 owners this is found under Utility --> Record --> End Frame.
Frames can be captured by time interval  , captured on trigger or captured an a "mask" test fail.
Frames can be time date stamped and zoomed.
You can also "Analyse" the captured waveforms for "Error" frames by , after capture , selecting , a template from a frame , selecting a % value for error and running analyse. Or you can create a "mask" .

Thanks John for sharing directly with us.

I didn't know about those features, sounds powerful. Pays to read the manual I guess 

Dave.

[tlu]

Hi - I'm the "culprit" behind the firware given to Dave. I also work for Emona Instruments who are the Rigol Distributors in Australia.

I'm posting here to clarify the issues with the firmware given to Dave. This was (As per what Drieg wrote) actually an engineering release not a public release. As such it was not fully tested . It was a release specifically given to me to show corrected XY display to a University here and there was a very short timeframe so it may have been rushed.

At the time I was unaware of it being an engineering release so passed it on to Dave with the introduced bug of non functional zoom in record mode.  I have a new engineering release that fixes this but am waiting on an answer from Rigol on when we expect a public release. I'll post this information here. This was my mistake - Public release firmware is more thoroughly tested. (And yes it could be posted better)

While I'm here a couple of answers to questions that came up on the record mode .

This segments the memory.
Yes you can select the number of segments you want. This is done in 2 ways.

First you can change record length.

For example if I change record length to 14k I can record up to 8128 frames 140k 508 etc.

For existing DS2000 users this is found under Acquire menu --> MemDepth
(Note : My unit has the extended memory depth option so the exact numbers might differ if it is not installed)

Second if I don't want that many I can select the End Frame (Can be anything from 1 to maximum)

For existing DS2000 owners this is found under Utility --> Record --> End Frame.

Frames can be captured by time interval  , captured on trigger or captured an a "mask" test fail.
Frames can be time date stamped and zoomed.

You can also "Analyse" the captured waveforms for "Error" frames by , after capture , selecting , a template from a frame , selecting a % value for error and running analyse. Or you can create a "mask" .

If anyone has any questions on the DS2000's or other Rigol products please feel free to ask me. I will try to keep any sales stuff out of it.

Hi John, I'm in the US and considering purchasing the DS2072 by the end of this month. Will the latest firmware fix be available to all Rigol resellers? What is the latest revision number of the firmware? I would like to ask this question to my local reseller in the states before buying.

tlu

[John South]

Hi Tlu - Yes I imagine any firmware updates will be available - it would be a matter of contacting your local distributor. As per my above post the bugs on zoom were introduced on an engineering firmware that Rigol got to me. On the previous tested release they are not there.

If you can live with Ch1 being "Y" on XY mode (just swap the inputs) then the current firmware will work fine prior to an update being available . (I needed to show a University here the correct XY as their courses are written around having students connect X to Ch1 Y to Ch 2 hence Rigol getting me an Engineering release) .

I'll post here any news.

[tlu]

Thanks John for the the respond. I will check with my local distributor. I'm also trying to get enough members here that is in the states to do a group or volume purchase in order to a discount. What is the best way you recommend on going about this as you are a reseller yourself?

The scope is very decently priced for the features it advertised and also a beautiful designed dso aesthetically in general. Hopefully, more members from here will jump on board. However, I actually do not how a group by can get arranged being we all somehow have to open up one account and pump our share into the funding. If you have any idea please throw in your two cent.

tlu

[marmad]

...then the current firmware will work fine prior to an update being available .

Hi John,

Any idea if the ability to save/recall recorded frames is something that Rigol is planning to implement in the future? Currently the option is greyed out when you are in Record/Play/Analysis mode. Of course, I know you can grab and save them via SCPI (http://turbidmedia.com/ds2000.html)- but there is no method to load them back into memory on the scope. Since you can do post-processing (measurements, bus decoding, etc) on running frames, it seems like it would be a wasted opportunity (instead of a great feature) on Rigol's part not to allow the possibility to reload a waveform you recorded last week and check further measurements, etc. on it.

Thanks, Mark

[John South]

Hi Mark - I'll check with Rigol if they have future funtionality improvements for the record mode and let you know.


John

[tlu]

Hi John,

Thanks for following up with the manufacturer of this particular feature that Mark pointed out. I've just ordered my Rigol DS2072 and will be expecting to arrive next Monday. I'm very excited about this scope and hope the feature Mark mentioned will get implemented as this is quite useful for post processing at another day. Hopefully you can convince Rigol to add this feature in their next revision of the firmware.

Much obliged,
Tim

[Sparky]

...then the current firmware will work fine prior to an update being available .

Hi John,

Any idea if the ability to save/recall recorded frames is something that Rigol is planning to implement in the future? Currently the option is greyed out ... <snip> Since you can do post-processing (measurements, bus decoding, etc) on running frames, it seems like it would be a wasted opportunity (instead of a great feature) on Rigol's part not to allow the possibility to reload a waveform you recorded last week and check further measurements, etc. on it.

Thanks, Mark

@marmad: Thank-you for thinking of this excellent addition --- the possibility of "recall" seems to be there...so now only needs to be implemented  

Hi Mark - I'll check with Rigol if they have future funtionality improvements for the record mode and let you know.
John

@John: Big thanks for following up on our functional wishes with Rigol! Many people are excited about the DS2000 series, and no doubt you can tell our enthusiasm!

<snip>
I've just ordered my Rigol DS2072 and will be expecting to arrive next Monday.
Much obliged,
Tim

@tlu: I too just placed an order! I noted earlier that you're in the US, and I responded to your group-order poll.  I was a bit on the fence then, and wasn't sure how the group purchase would work.  Recently I found Tequipment.net would give me a 5% .edu discount, and so I purchased through them.  It is possible for individuals to get a discount, if you qualify by some means, and I hope things worked out for you.

I'm very excited too!  And, we are all looking forward to trying @marmad's software!

[tlu]

@tlu: I too just placed an order! I noted earlier that you're in the US, and I responded to your group-order poll.  I was a bit on the fence then, and wasn't sure how the group purchase would work.  Recently I found Tequipment.net would give me a 5% .edu discount, and so I purchased through them.  It is possible for individuals to get a discount, if you qualify by some means, and I hope things worked out for you.

I'm very excited too!  And, we are all looking forward to trying @marmad's software!

Unfortunately, there is no individual discount or student discount. I've personally talked the the sales rep over at Tequipment. However, if the school worked at something with them than there would be a 5% discount on tier of 5 units. The rep also mentioned depending on the number of units ordered through the school (this is a must to get the discount) they can work on further discount. Unfortunately for me, my school already has dso supplied to them so I'm out of luck and had to do an individual buy but I was able to get a free dso case granted it may not fit the Rigol DS2072 unit but it is ok with me. I can use it for other things. It is their inhouse brand dso carrying case which did not look too bad. I wasn't able to get enough of my fellow students to join me in this quest for the discount so I jumped the gun and made a purchase. I'm super excited and counting down the days. Can't wait to try out Marmad's software as well. It should be here next week according to the UPS delivery schedule.

[John South]

Hi Mark - At this stage there are no immediate plans to add this functionallity (Save/Recall recorded frames) .

This is not to say it won't be added in the future.

John

[marmad]

Hi Mark - At this stage there are no immediate plans to add this functionallity (Save/Recall recorded frames) .

This is not to say it won't be added in the future.

John

Hi John,

Thanks so much for your help and speedy response. It's a shame that they have no plans to implement that feature - I could think of a number of powerful uses for it. One that springs to mind would be the time-based equivalent of using a static Reference waveform; e.g. being able to compare the output of a design you're working on (recorded a week ago) to the real-time current output which you've tweaked.

BTW, Rigol needs to come up with better names for some of their powerful features. 'Record Open' sounds rather dull - but if you use Yokogawa's name, 'History Memory', doesn't it sound way sexier? 

Anyway, just for fun, here is my current features wish-list for the DS2000 series:
1) Save/Recall of specifc channels of recorded frames at specific frame points.
2) Markers while in Delayed Sweep Mode (searching would be better - but I think unlikely)
3) SCPI :SYST:KEY command to poll the 4 menu pg up/pg dwn buttons (unused on the front panel when the menus are retracted - the scope lacks the ability to initiate contact with control software and this would be an easy way to do it - and they would also then function as programmable soft-keys).

Mark

[marmad]

*(Double posting this here - as well as the other Rigol thread - since John South might see it here and offer some feedback)*

I've been working on a new version of my Rigol software for the last couple of days, and I seem to have found the first bug in the DS2000 that is, IMO, rather a serious one:

I don't seem to be able to read the contents of the DSO memory reliably when the scope is in the STOP state - in other words, I can't get the full sample depth out of the scope (getting just the displayed data is not a problem) - which is a big deal if anyone is planning to do any post-processing with 3rd party software, including MATLAB or LabVIEW (one of the big selling points of the Rigol scope over it's competitors).

Here is the procedure to use copied from the Rigol DS2000 programming guide:

Procedures of the internal memory waveform data reading:
S1. :STOP
The internal memory waveform data can only be read in STOP state
S2. :WAV:SOURce CHAN1
Set the channel source to be read
S3. :WAV:MODE RAW
Set the waveform mode to RAW
S4. :WAV:RESet
Reset the waveform reading
S5. :WAV:BEGin
Start the waveform reading
S6. :WAV:STATus?
1) IDLE
:WAV:DATA?
:WAV: END
2) READ
:WAV:DATA?
Repeat S6

This order of commands does NOT work correctly with the DS2000 - it doesn't return or read the correct number of samples when run as a loop if the memory depth is set to anything higher than 14k.  In fact, you can test that it doesn't work correctly by just using Rigol's own software Ultra Sigma:
Just set the memory depth to >= 140k; STOP the scope; start Ultra Sigma; type in the listed SCPI commands one by one; and you will see it fail.

Perhaps the correct method has changed since the programming guide was published by Rigol - I've passed this info along to my dealer to pass to Rigol, We'll see what they have to say about it.

[tan98010]

Guys,

Currently i am trying to upgrade my Rigol DS2202 firmware, everything seems normal during the process, but the scope is bricked after it restart it self (No boot, just hang there with "RIGOL" logo appear on the screen).

P.S :i upgrade the firmware after the scope is fully booted up.

Anyone got this experience before please help me.

[Astroplio]

Do what drieg said earlier in this thread page 3.

The process can take several minutes, mine took about 25.

If you can't get the help button method to work the first time, keep trying, it is a bit tricky to get the timing of the button press right, just after the power up.

Wish you luck!

[CarlG]

That certainly sounds like the best way Dave but don't you think the Pass/Fail could also work?
I imagine that on a low end DSO that's going to run into the same capture rate issue??
BTW Dave can you setup a standardized test on a video that anybody with a DSO and a arb generator could use to test their scope?  It would give all of us great data to publish in the videos forum thread and since you have several dso we could see how our cheapies stack up in an apples to apples comparison of wfcr.

I assume that your request to Dave is regarding testing the capture rate? It's not that complicated, but on the other hand it's that simple either:

You don't need a function generator to test the capture rate, but you'll need a frequency counter. Connect it to the Trig Out of the scope and you can read the capture rate for different timebases on the counter. The famous capture rate "sweet spot" that the manufacturer claim might be a very small spot...e.g. as in this Rigol example.

Then, turn on anything you can think of: all channels (also digital if it's an MSO), turn on [serial] decoding if it exist, turn on zoom, automatic measurements,...change record length/memory depth if available, check vector/dots mode, and see what capture rate you get in a realistic situation! For instance, if you're looking at an SPI interface you're likely to have four channels enabled. Or if you're looking at a memory interface, you'd probably use the digital channels along with one or more analog channels.

And yes, as you could guess, I'm pretty fond of the Agilent scopes (the X-series but also the 6k/7k-series) just because of their high capture rates independent of the mode you're running the scope in, plus that you don't have to think about record length/memory depth. I'm not that familiar with other scopes than Agilent and Tektronix (and to some extent LeCroy), but I'd say that these Agilent scopes (NOT e.g. the 1000-series) beat any other scope you compare with (and beat them by far) except for the Rohde & Schwartz scopes. (As far as I know, R&S were the first to claim a capture rate of 1 million wfms/s. But the R&S user interface is dreadful, with single knobs for the four vertical channel attn/pos...)

However, the Agilent scopes also have their drawbacks, but I won't go into that now.

//C

[marmad]

And yes, as you could guess, I'm pretty fond of the Agilent scopes just because of their high capture rates independent of the mode you're running the scope in, plus that you don't have to think about record length/memory depth.

I'm not exactly sure what you're saying here, but if you're implying that the Agilent X series waveform capture rates don't vary greatly with the timebase scale setting and other parameters - like every other DSO - then that is incorrect.

[CarlG]

And yes, as you could guess, I'm pretty fond of the Agilent scopes just because of their high capture rates independent of the mode you're running the scope in, plus that you don't have to think about record length/memory depth.

I'm not exactly sure what you're saying here, but if you're implying that the Agilent X series waveform capture rates don't vary greatly with the timebase scale setting and other parameters - like every other DSO - then that is incorrect.

You're correct, my statement above is not strictly true. It's (obviously) not valid when the capture rate is limited by the timebase setting#1. Maybe I should have made that remark to clearify. But in my post I'm referring to [a] scope's [claimed] maximum capture rate, which isn't applicable at lower timebase settings.

So what I'm trying to say is that, at a given timebase setting which not inherently limit the capture rate, the capture rate for the Agilent scopes doesn't change when adding more channels, decoding etc, compared to single channel mode, no decoding etc. Then again, I admit I haven't verified this myself, so I might be wrong. My assumption is based on the 6000/7000-series performance, and my belief that Agilent wouldn't do the mistake of designing the X-series worse in this respect. But who knows?  It would be interesting if somebody could produce a table corresponding to the Rigol example above, with additional columns for number of channels, decoding and so on. I have the intention to produce such a table for a DSO3000-series scope later this week.

#1: E.g. for the DSOX2000-series, with a claimed 50k wfms/s, that (theoretical) limit corresponds to 1/50k = 20us, or 2us/div. In practice it might be lower, maybe 1us/div.

//C

[marmad]

It would be interesting if somebody could produce a table corresponding to the Rigol example above, with additional columns for number of channels, decoding and so on. I have the intention to produce such a table for a DSO3000-series scope later this week.

Since I produced the table you're referring to, I'd be happy to create another one with channels, decoding, etc. enabled - but I'm currently travelling so I can't do it for 3 weeks. I did do a little testing when I made the original table, and I remember that some additional settings DID affect the rate - and others did not.

I might point out that the Agilent document which lists the update rates for the 2000X, 3000X, etc. notes:
"Tables 1, 2, and 3 show side-by-side measured waveform update rates of competitively priced 100-MHz, 500-MHz, and 1-GHz bandwidth scopes respectively. The test began by defaulting each scope’s setup condition. Only one channel of the scope was turned on for these tests."
While I don't own an Agilent X-Series to test, I would surmise that the above comment tends to imply that non-default setups - or more than one channel turned on - might negate the listed waveform update rates.

[marmad]

If you can't get the help button method to work the first time, keep trying, it is a bit tricky to get the timing of the button press right, just after the power up.

After power on, all of the LEDs flash for ~1 second. During that flash, you have to press AND let go of the 'Help' button.

[CarlG]

It would be interesting if somebody could produce a table corresponding to the Rigol example above, with additional columns for number of channels, decoding and so on. I have the intention to produce such a table for a DSO3000-series scope later this week.

Since I produced the table you're referring to, I'd be happy to create another one with channels, decoding, etc. enabled - but I'm currently travelling so I can't do it for 3 weeks. I did do a little testing when I made the original table, and I remember that some additional settings DID affect the rate - and others did not.

Sounds great!

I might point out that the Agilent document which lists the update rates for the 2000X, 3000X, etc. notes:

Also great! I tried to find such a document, but obviously not hard enough...thanks!
EDIT: Oh, that document...I've seen it being referred to all over the forum but didn't think of it

I made some preliminary measurement on an MSO7104B today but got inconsistent results, so I have to investigate further...but for the moment I'm inclined to withdraw my statement, at least partially, and agree with you.  I'll be back

[Hydrawerk]

So when is Rigol company going to release the new DS2000 firmware to fix the bugs?

[KuchateK]

I wonder if a simpler method might be to just stop the 32K clock oscillator - very easy to do.

Anyone tried this? Do you think that stopping this clock would break any important features?

[MikeR]

*(Double posting this here - as well as the other Rigol thread - since John South might see it here and offer some feedback)*

I've been working on a new version of my Rigol software for the last couple of days, and I seem to have found the first bug in the DS2000 that is, IMO, rather a serious one:

I don't seem to be able to read the contents of the DSO memory reliably when the scope is in the STOP state - in other words, I can't get the full sample depth out of the scope (getting just the displayed data is not a problem) - which is a big deal if anyone is planning to do any post-processing with 3rd party software, including MATLAB or LabVIEW (one of the big selling points of the Rigol scope over it's competitors).

Here is the procedure to use copied from the Rigol DS2000 programming guide:

Procedures of the internal memory waveform data reading:
S1. :STOP
The internal memory waveform data can only be read in STOP state
S2. :WAV:SOURce CHAN1
Set the channel source to be read
S3. :WAV:MODE RAW
Set the waveform mode to RAW
S4. :WAV:RESet
Reset the waveform reading
S5. :WAV:BEGin
Start the waveform reading
S6. :WAV:STATus?
1) IDLE
:WAV:DATA?
:WAV: END
2) READ
:WAV:DATA?
Repeat S6

This order of commands does NOT work correctly with the DS2000 - it doesn't return or read the correct number of samples when run as a loop if the memory depth is set to anything higher than 14k.  In fact, you can test that it doesn't work correctly by just using Rigol's own software Ultra Sigma:
Just set the memory depth to >= 140k; STOP the scope; start Ultra Sigma; type in the listed SCPI commands one by one; and you will see it fail.

Perhaps the correct method has changed since the programming guide was published by Rigol - I've passed this info along to my dealer to pass to Rigol, We'll see what they have to say about it.

I can also confirm this behavior with the 00.01.00.00.03 firmware. 1400 points "all the time".

Curiously theres a bandwidth license option in there also, not sure if thats for the DS2072.
The firmware seems to be a menagerie of code supporting many different models.

-M

[marmad]

I can also confirm this behavior with the 00.01.00.00.03 firmware. 1400 points "all the time".

Curiously theres a bandwidth license option in there also, not sure if thats for the DS2072.
The firmware seems to be a menagerie of code supporting many different models.

-M

Mike - you don't seem to reading the collected info that we've gathered here. I steered you yesterday to the post which contained not only the instructions for getting your firmware version - but the bug list for the all the various firmware versions. The bug I reported above (which you quoted) is NOT in FW 01.00.00.03 - so you must be doing something wrong in reading the memory. I suggest you go and read the bug list we've collected first before posting reports of bugs  - and then, if you still believe you have an unreported bug, post it in the Rigol REVIEW thread.

Mark