Rigol DP832 - Firmware list and bugs

[alank2]

Thanks for letting me know.  Just out of curiosity - is the firmware between the 832 and 832A the same or different?

[staze]

I've had this happen once as well. In my case: Latest Firmware with all options enabled. Nothing connected at all - ie: The power cable was the only cable coming from the unit. No channels were turned on. The PSU had been running for about 3 Minutes at the time from a completely cold start. That was about a 6 Months ago and has never happened since.

McBryce.

My reset issue just occurred again. I was working on something else and heard the fans ramp to full speed like it does on power up, and noticed the outputs were off. They were on previously, but with no load...

Very odd. Thinking I might contact Rigol about this.

[jc101]

I have a weird issue with a DP832A and it's ethernet port and would be interesting if anyone else has issues?

When in DHCP mode, regardless of the lease duration set on the server, the DP832A requests a new IP every 5 minutes.  So it seems to be ignoring the lease duration, although this could be linked to...

Every so often the supply will drop off the network, either a reboot or a reset of the IP parameters via the front panel will fix this.  Also the display will display the LAN connected message at power up OK, than also randomly display this again from time to time.  Even with a static IP address assigned this will happen too.

I have swapped the ethernet leads, different port on the switch etc. and the problem persists.  The unit is running 1.14 but I think it did it with 1.13 too.

It's a real pain as I control the PSU via ethernet for some semi-automated testing of items and it's *really* annoying when it happens as it means I have to restart the tests.  The DP832A will complain it is under remote control, but not respond to commands or even a ping when it drops off.  You need to 'release' it using the Back key when this happens before you can talk to it again.  Even when not talking to the unit via IP and just using the front panel the unit will stop responding on the LAN side too.

Anyone else seen anything strange with the ethernet port or is this a developing issue with my DP832A?

[NF6X]

I haven't had my DP832 powered on and under observation enough to be at all sure, but seeing a random LXI splash message out of the corner of my eye sounds a bit familiar.

[jc101]

I haven't had my DP832 powered on and under observation enough to be at all sure, but seeing a random LXI splash message out of the corner of my eye sounds a bit familiar.

I suppose that is good news in that it looks like a software thing then.

I have been pondering changing it for a.n.other 2 or 3 channel PSU to get 0-60v and a high current, might have a look around and see what will fit the bill.  Defeats the point of scripting tests if you have to keep stopping to reset the PSU to get the comms back.

[LaurentR]

I haven't had my DP832 powered on and under observation enough to be at all sure, but seeing a random LXI splash message out of the corner of my eye sounds a bit familiar.

I suppose that is good news in that it looks like a software thing then.

I have been pondering changing it for a.n.other 2 or 3 channel PSU to get 0-60v and a high current, might have a look around and see what will fit the bill.  Defeats the point of scripting tests if you have to keep stopping to reset the PSU to get the comms back.

I have never seen any issue with my DP832 resetting or getting on/off the network.

Several of us have scripted this PSU, e.g. to analyze the acuracy of the DAC and ADC and have had no issue. In one of my tests, I did a full 1mV sweep of the whole 30V range, which is almost 24hrs of runtime, with no issue.

Obviously, some people have issues with their DP832, but I wouldn't panic and automatically assume this is wide spread and broken on all units. I am not sure what alternative PSU there is that will give you more perceived stability and specs in this price range.

[jc101]

I haven't had my DP832 powered on and under observation enough to be at all sure, but seeing a random LXI splash message out of the corner of my eye sounds a bit familiar.

I suppose that is good news in that it looks like a software thing then.

I have been pondering changing it for a.n.other 2 or 3 channel PSU to get 0-60v and a high current, might have a look around and see what will fit the bill.  Defeats the point of scripting tests if you have to keep stopping to reset the PSU to get the comms back.

I have never seen any issue with my DP832 resetting or getting on/off the network.

Several of us have scripted this PSU, e.g. to analyze the acuracy of the DAC and ADC and have had no issue. In one of my tests, I did a full 1mV sweep of the whole 30V range, which is almost 24hrs of runtime, with no issue.

Obviously, some people have issues with their DP832, but I wouldn't panic and automatically assume this is wide spread and broken on all units. I am not sure what alternative PSU there is that will give you more perceived stability and specs in this price range.

I haven't always had issues, but it may well be a symptom of something else going on internally, some days are better than others.  There is no panic, but a wider voltage range and higher current would also be handy for the things I get up to these days.  I've had it 2.5 years now and works perfectly with the exception of the ethernet issue, which is just getting really annoying.

[McBryce]

Have you excluded all other possibilities? ie: swapped computers / routers etc to make sure it's not some other part of the setup that's causing the issue?

McBryce.

[jc101]

Have you excluded all other possibilities? ie: swapped computers / routers etc to make sure it's not some other part of the setup that's causing the issue?

McBryce.

Yes, swapped leads and even switches around, the issue always remains with the DP832A.  It will drop out using DHCP or with a static IP assigned too.

[Macbeth]

Have you tried it with USB instead of Ethernet? I'm ashamed to say I haven't even tried the Ethernet option myself, but haven't had issues with USB.

[jc101]

Have you tried it with USB instead of Ethernet? I'm ashamed to say I haven't even tried the Ethernet option myself, but haven't had issues with USB.

I have to admit I haven't, I tend to do everything via ethernet as it's just easier in my setup.
I will probably have to move over to USB, assuming that is OK, to keep things running until I get round to sorting things out.

[staze]

Rigol got back to me and asked some follow ups, including shipping information. Made sure I was running the latest firmware, etc.

It's very odd. I wish I knew what was causing it. It sure sounds like the LM317 issue, but mine is well after they fixed that. It also sounds a bit like some software watchdog bug...

Anyway, I'll post back when I hear more. If it was easier to reproduce, I'd be troubleshooting a lot more, but just because it runs hours without an issue doesn't mean it's fixed. =/ And since it's under warranty, I don't want to crack the seal to monitor the LM317 or CPU reset pin...

[kandrey89]

Not a lot of DP832A threads on EEVblog.

I was wondering if these issues have been fixed in the latest PCB/design/firmware revisions?
1. Easy to overload pass mossfet
2. Do people still prefer the DP832 instead of DP832A? Does DP832 have all the updates PCB/design wise?
3. Where's the DP832A hack thread to unlock more resolution? I saw the "Sniffing Rigol's Internal I2C Bus" thread, but that's like >160 pages, not realistic to go through that.

[McBryce]

Not a lot of DP832A threads on EEVblog.

I was wondering if these issues have been fixed in the latest PCB/design/firmware revisions?
1. Easy to overload pass mossfet
2. Do people still prefer the DP832 instead of DP832A? Does DP832 have all the updates PCB/design wise?
3. Where's the DP832A hack thread to unlock more resolution? I saw the "Sniffing Rigol's Internal I2C Bus" thread, but that's like >160 pages, not realistic to go through that.

1 - The 832 and 832A use exactly the same PCB, so all PCB/Design updates should be on both. The MOSFET issue has been resolved on the latest models as far as I know.
2 - The preference between 832 and 832A is mainly just visual. A hacked 832 offers everything that the A offers without the Eye-cancer inducing colours.
3 - The hack doesn't require any I2C sniffing or opening the device. Just use the riglol software to create the keys you need.

McBryce.

[CustomEngineerer]

3. Where's the DP832A hack thread to unlock more resolution? I saw the "Sniffing Rigol's Internal I2C Bus" thread, but that's like >160 pages, not realistic to go through that.

There are no options on the DP832A to be hacked. The DP832A is just a DP832 with all options already enabled.

[staze]

Rigol issued me an RMA to send my unit in. Thankfully, rather than shipping this beast, I'm going to be in the area of their office (Portland Oregon suburb) in the next month or two and they said I could just drop it off. =)

I'll post back when they've fixed it... and if I get any pictures of the Rigol NA offices. =)

[SupraWez]

Thanks for letting me know.  Just out of curiosity - is the firmware between the 832 and 832A the same or different?

According to this link it would appear the FW is the same for all DP units :-

http://int.rigol.com/Support/SoftDownload/3

[Scottjd]

Hi Laurent, it looks like you are the creator of the cal script and I appreciate your work and sharing this.
I'm just got a new DP832 that is showing different reading and tolerates, it's said it was calibrated 6 months ago. So I plan to try your script since I have a 3 month old calibrated hp 34401A.
One of the factors for buying this was that I could see the current being used on my DUT without adding burden voltage inline.

So one fothe downfall of most bench meters is the high burden voltage when measuring current draw. I still need to learn more on SCPICOMMNADS and install the computer needed libraries So for calibrating the current side (I) I was wondering if you think it's possible that I could modify the script to read voltage for the current reading. What I'm looking at doing is getting the Agilent current shunt that shows 1mV per 1mA for more accurate measurements with the 34401A. So my thought was to modify the script to pull e mV measurements and apply the difference to the current calibration factors on the DP832.
What are the chances of this working after I do the needed research and learn more on this method of communication to my Equiptment? Are the needed command available for such a change to the script, or would I be waisting my time trying to get this to work?

After reading reviews I expected that even my new unit might be out of calibration to what I'm looking to use it for. So I'm happy to say your script was a big part of my decision in buying the DP832. I like the fact that I can have my 34401A calibrated and then use that to calibrate the DP832 without having to spend on getting a second device calibrated. I'm also banking in the age of the 34401A and that it's done drifting, I'll find out if that's true when I send it out in 8 months for my next yearly calibration and compare the data from the original calibration. If it has stopped drifting then a yearly cal may not be needed for someone like me who is just a hobbiest. If the 34401A starts disagreeing with my Fluke 289 then I will have to figure out what meter went off, but for now as long as they both agree within the tolerance expected then I can trust the meter to use for calibrating other Equiptment like the DP832 and save money not having to pay for calibrations.

Thanks,
Scott

Looks great (making me have Keithley envy)!

Note that for the "readings within 1mV", if you do a full sweep, you'll discover the DAC is a bit all over the place in some spots. See this thread for more background:
https://www.eevblog.com/forum/testgear/noob-dac-linearity-in-the-dp832/
So across the sweep, it's actually much closer to the quoted 10mV accuracy.

[LaurentR]

Hi Laurent, it looks like you are the creator of the cal script and I appreciate your work and sharing this.
I'm just got a new DP832 that is showing different reading and tolerates, it's said it was calibrated 6 months ago. So I plan to try your script since I have a 3 month old calibrated hp 34401A.
One of the factors for buying this was that I could see the current being used on my DUT without adding burden voltage inline.

So one fothe downfall of most bench meters is the high burden voltage when measuring current draw. I still need to learn more on SCPICOMMNADS and install the computer needed libraries So for calibrating the current side (I) I was wondering if you think it's possible that I could modify the script to read voltage for the current reading. What I'm looking at doing is getting the Agilent current shunt that shows 1mV per 1mA for more accurate measurements with the 34401A. So my thought was to modify the script to pull e mV measurements and apply the difference to the current calibration factors on the DP832.
What are the chances of this working after I do the needed research and learn more on this method of communication to my Equiptment? Are the needed command available for such a change to the script, or would I be waisting my time trying to get this to work?

After reading reviews I expected that even my new unit might be out of calibration to what I'm looking to use it for. So I'm happy to say your script was a big part of my decision in buying the DP832. I like the fact that I can have my 34401A calibrated and then use that to calibrate the DP832 without having to spend on getting a second device calibrated. I'm also banking in the age of the 34401A and that it's done drifting, I'll find out if that's true when I send it out in 8 months for my next yearly calibration and compare the data from the original calibration. If it has stopped drifting then a yearly cal may not be needed for someone like me who is just a hobbiest. If the 34401A starts disagreeing with my Fluke 289 then I will have to figure out what meter went off, but for now as long as they both agree within the tolerance expected then I can trust the meter to use for calibrating other Equiptment like the DP832 and save money not having to pay for calibrations.

Thanks,
Scott

Looks great (making me have Keithley envy)!

Note that for the "readings within 1mV", if you do a full sweep, you'll discover the DAC is a bit all over the place in some spots. See this thread for more background:
https://www.eevblog.com/forum/testgear/noob-dac-linearity-in-the-dp832/
So across the sweep, it's actually much closer to the quoted 10mV accuracy.

Scottjd, what you are trying to do is not complicated, and everything you'll need is already in the script, so you'll just need a bit of Matlab editing to get it to work.
* First, you should install Matlab, the Keysight IO Libraries and make sure you can run the script and go through the first IDN calls. I used to have a 34401A and the script ran on it in draft form. Note that I was using the serial port and had all kinds of issues with that setup.
* Second, the current script probably doesn't work out of the box on the 34401A anymore, but it should be easy to make it do so. You just have to recognize the 34401A IDN and skip the is34461A section, which uses 34461A-only commands (temperature, uptime... which are not critical to the test).
* Finally, for the 3.2A reading, you have to modify a bit of code. When you detect that the current step is >3A, do the voltage init and reading as done during the voltage stepping and do the math.

Note that at 3.2A, you are unlikely to blow the fuse and the 34401A should be able to measure it properly. I believe one of the users in this thread has done it.

I don't really have time to experiment at home, but I can try to help if you bump into any issue.

Laurent

[Scottjd]

Thanks for the responce,

As for blowing a fuse with the > 3.2A current  measurement with the 34401A it won't be an issue if I'm using the more acurate current shut, hence the 34401A will be measuring mV, voltage and not amperage.
But yes, I've done short times of 3.2A current measurements with the 34401A, but this is were its the most inaccurate and can apply 2V of burden voltage, it's why I'm looking at using the shut in the voltage side instead for calibration and it more acurate over the 3.2A range.

I won't be using the serial cominication, I have a keysight GBIP to USB adapter so I shouldn't see these cominication errors.

Thanks for the heads up in these potential issues.
Scott

Hi Laurent, it looks like you are the creator of the cal script and I appreciate your work and sharing this.
I'm just got a new DP832 that is showing different reading and tolerates, it's said it was calibrated 6 months ago. So I plan to try your script since I have a 3 month old calibrated hp 34401A.
One of the factors for buying this was that I could see the current being used on my DUT without adding burden voltage inline.

So one fothe downfall of most bench meters is the high burden voltage when measuring current draw. I still need to learn more on SCPICOMMNADS and install the computer needed libraries So for calibrating the current side (I) I was wondering if you think it's possible that I could modify the script to read voltage for the current reading. What I'm looking at doing is getting the Agilent current shunt that shows 1mV per 1mA for more accurate measurements with the 34401A. So my thought was to modify the script to pull e mV measurements and apply the difference to the current calibration factors on the DP832.
What are the chances of this working after I do the needed research and learn more on this method of communication to my Equiptment? Are the needed command available for such a change to the script, or would I be waisting my time trying to get this to work?

After reading reviews I expected that even my new unit might be out of calibration to what I'm looking to use it for. So I'm happy to say your script was a big part of my decision in buying the DP832. I like the fact that I can have my 34401A calibrated and then use that to calibrate the DP832 without having to spend on getting a second device calibrated. I'm also banking in the age of the 34401A and that it's done drifting, I'll find out if that's true when I send it out in 8 months for my next yearly calibration and compare the data from the original calibration. If it has stopped drifting then a yearly cal may not be needed for someone like me who is just a hobbiest. If the 34401A starts disagreeing with my Fluke 289 then I will have to figure out what meter went off, but for now as long as they both agree within the tolerance expected then I can trust the meter to use for calibrating other Equiptment like the DP832 and save money not having to pay for calibrations.

Thanks,
Scott

Looks great (making me have Keithley envy)!

Note that for the "readings within 1mV", if you do a full sweep, you'll discover the DAC is a bit all over the place in some spots. See this thread for more background:
https://www.eevblog.com/forum/testgear/noob-dac-linearity-in-the-dp832/
So across the sweep, it's actually much closer to the quoted 10mV accuracy.

Scottjd, what you are trying to do is not complicated, and everything you'll need is already in the script, so you'll just need a bit of Matlab editing to get it to work.
* First, you should install Matlab, the Keysight IO Libraries and make sure you can run the script and go through the first IDN calls. I used to have a 34401A and the script ran on it in draft form. Note that I was using the serial port and had all kinds of issues with that setup.
* Second, the current script probably doesn't work out of the box on the 34401A anymore, but it should be easy to make it do so. You just have to recognize the 34401A IDN and skip the is34461A section, which uses 34461A-only commands (temperature, uptime... which are not critical to the test).
* Finally, for the 3.2A reading, you have to modify a bit of code. When you detect that the current step is >3A, do the voltage init and reading as done during the voltage stepping and do the math.

Note that at 3.2A, you are unlikely to blow the fuse and the 34401A should be able to measure it properly. I believe one of the users in this thread has done it.

I don't really have time to experiment at home, but I can try to help if you bump into any issue.

Laurent

[LaurentR]

Thanks for the responce,

As for blowing a fuse with the > 3.2A current  measurement with the 34401A it won't be an issue if I'm using the more acurate current shut, hence the 34401A will be measuring mV, voltage and not amperage.
But yes, I've done short times of 3.2A current measurements with the 34401A, but this is were its the most inaccurate and can apply 2V of burden voltage, it's why I'm looking at using the shut in the voltage side instead for calibration and it more acurate over the 3.2A range.

I won't be using the serial cominication, I have a keysight GBIP to USB adapter so I shouldn't see these cominication errors.

Thanks for the heads up in these potential issues.
Scott

Note that I don't understand your point on the burden voltage. While calibrating current, it is in CC mode and the burden voltage should not matter.

[Scottjd]

Sorry, I'll explain that a bit more. My old Tripple linear power supply (although it seemed better isolated and had easy features like one button series and parallel and did up to 5A per channel, or 10A -60V depending how you had it configured) worked ok, but the display showing the current draw and voltage was off. So I found my self contanty wiring other meters in line to get an accurate load draw in the device, but in doing so cussed a voltage drop. It's the main reason I wanted a precision power supply so I dint have to spend the star time figuring out how much voltage was drop happened depending on the amount of current being used by the DUT. With the DP832 I hope I won't need the 4 other DMMs wired up when providing 2 channels of different voltages to a single device. It became tired some and took a lot of space up on the bench. Not to mention I only own one DMM that had the 50,000 count and acuracy that I considered acceptable when testing, that's the fluke 289. Other meters like the bryman 235 are great, but lack the resolution sometimes for the finer precision items I'm working on.

So when I was planning out the calibration I kind of figured I would run into the same burden voltage, but now that I'm thinking about it again to calibrate the current factor it wouldn't matter if the voltage was affected on that line through the meter I'm using for calibration since I won't have a DUT hooked up. So it might of been the habit of thinking alway having to put another meter in line since I had to think like that when testing a device.

So since I know the 34401A can take short time with 3.2A (I just did this today using the min/max on testing something by mistake, something i was testing  had a short and caused the spike at 3.2145A) then it sounds like I can just follow the normal procedure and use the 34401A on DCI for the calibration figure adjustments, and forget about changing the code to use the shunt for the voltage reading on the shunt.

FYI, I know I should have had OCP on, but I'm still learning this PSU so the 3.2A reading won't happen in the future unless I need a component to heat up so I can see what's failing with the Flir.

Thanks,
Scott


 author=LaurentR link=topic=24919.msg960804#msg960804 date=1465794678]

Thanks for the responce,

As for blowing a fuse with the > 3.2A current  measurement with the 34401A it won't be an issue if I'm using the more acurate current shut, hence the 34401A will be measuring mV, voltage and not amperage.
But yes, I've done short times of 3.2A current measurements with the 34401A, but this is were its the most inaccurate and can apply 2V of burden voltage, it's why I'm looking at using the shut in the voltage side instead for calibration and it more acurate over the 3.2A range.

I won't be using the serial cominication, I have a keysight GBIP to USB adapter so I shouldn't see these cominication errors.

Thanks for the heads up in these potential issues.
Scott

Note that I don't understand your point on the burden voltage. While calibrating current, it is in CC mode and the burden voltage should not matter.
[/quote]

[Scottjd]

So I noticed today that you still can not install option remotely on firmware 00.01.14 once you enable the LAN option.
I thought I could save myself some time and enable the LAN, then use SSH to install the other option.
So since this firmware was last updated on April of 2015, I guess its safe to say they are not going to fix this bug.

Are any of the other remote syntax command lines messed up that I should know about now before I start writing or scripting commands. Or it is only the :SYST:OPT:INST that does not work?

[firstcolle]

hi,

i've recent bought a rigol dp832. i experiece that overvolt protection switch off the channel a second after the threshold is raised.. it's correct???
i tought that this kind of protection is controlled before setting the voltage output.. but what i've seen is that the voltage output go higher than the overvolt set for 1 second and than the channel si switched off..

no problem if the voltage is increased by 10/100mw.. but if is rised by 1 volt or 10 volt this can cause some chips burned....

[TurboTom]

Actually I was quite disappointed about this behavior as well. I thought to myself, "Why do they permit an output voltage to be enabled that's clearly above the limit threshold". Well, in a second thought this means, if the output would stay disabled in this mode, you could never charge a battery if you want the O/V function to cut the output when the battery reached charge end voltage. I guess it's difficult to make everyone happy... Anyway, altogether I consider the DP832 a pretty decent piece of kit even though it may still have its peculiarities 

Cheers,
Thomas