Author Topic: Keysight EDU36311A current reading bug?  (Read 3023 times)

0 Members and 1 Guest are viewing this topic.

Offline AmarukTopic starter

  • Contributor
  • Posts: 46
  • Country: us
Keysight EDU36311A current reading bug?
« on: November 27, 2021, 08:01:21 pm »
Hi all,
I recently got myself a Keysight EDU36311A three-channel power supply. I really love it!! However, when I was working on a project the other day I noticed something strange with regards to the current readings in the display. It seems to me that channel 2 and 3 fail to monitor the current at all times... if the current goes down to 0 due to a the load being disconnected, it fails to capture it! For some reason it seems to capture it the first time (see video below) but if the disconnections are repeated it fails on channel 2 and 3. Channel 1 works just fine. So odd. What can be causing this? Firmware bug? Hardware issue?

Here is a quick experiment I set up that shows the problem. In the video I have the three channels all set to 5V with a CC limit of 175mA. The channels are connected to three 100 ohm (5W) power resistors that will draw 50mA when connected to the power supply. Watch the video and pay attention to the current readings when I disconnect the live wires from the three channels. Channel 2 and 3 gives the impression that there is current flowing thru air...Has anyone else experienced this?



« Last Edit: November 27, 2021, 11:47:07 pm by Amaruk »
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 14076
  • Country: de
Re: Keysight EDU36311A current reading bug?
« Reply #1 on: November 27, 2021, 08:19:32 pm »
The main suspect would be a software issue. A pure HW issue is rather unlikely - maybe a combination, like loosing the connection (e.g. an interface going to an error state that is not cleared) to the ADC and than no longer updating.

 

Offline AmarukTopic starter

  • Contributor
  • Posts: 46
  • Country: us
Re: Keysight EDU36311A current reading bug?
« Reply #2 on: November 27, 2021, 09:01:15 pm »
The main suspect would be a software issue.

Good point! Thank you! I forgot that say that I am using the latest firmware version for the instrument which is from 4/4 2021 (Keysight document with firmware info: https://www.keysight.com/us/en/lib/software-detail/instrument-firmware-software/edu36311a-firmware.html).

If anyone else has this power supply it would be great if you can confirm the issue. Thanks!
« Last Edit: November 27, 2021, 09:03:23 pm by Amaruk »
 

Offline WattsThat

  • Frequent Contributor
  • **
  • Posts: 764
  • Country: us
Re: Keysight EDU36311A current reading bug?
« Reply #3 on: November 27, 2021, 09:39:31 pm »
Something is very odd with channels 2 and 3 each showing exactly 50 millamps regardless of the channel being on or off, which was the displayed value when the channel was turned off. Channel 3 also has the odd behavior of showing 23 milliamps instead of zero. Sure looks like a failure to update either the mux  or adc.

I would try different output voltages so you have different currents in each channel to see if the behavior changes.

As stated by Kleinstein, it is likely a firmware issue. You could always try rolling back to a previous firmware to see if the behavior changes.
 

Offline AmarukTopic starter

  • Contributor
  • Posts: 46
  • Country: us
Re: Keysight EDU36311A current reading bug?
« Reply #4 on: November 27, 2021, 11:38:59 pm »
You could always try rolling back to a previous firmware to see if the behavior changes.

Unfortunately, this is the only firmware version that has been released so far for this power supply so I can't try another version.

I will contact Keysight and see what they say about this. Since the current output reading is an important function for a power supply I would imagine they will fix this in the next firmware release. I have worked with HP/Agilent/Keysight instruments for three decades and they always stand behind their products. That is why they are always on the top of my wishlist for new gear.
 

Offline Anthocyanina

  • Frequent Contributor
  • **
  • Posts: 317
  • Country: 00
  • The Sara
Re: Keysight EDU36311A current reading bug?
« Reply #5 on: November 28, 2021, 04:34:24 am »
Have you used the benchvue software to see if it shows the same there?
 

Offline electr_peter

  • Supporter
  • ****
  • Posts: 1300
  • Country: lt
Re: Keysight EDU36311A current reading bug?
« Reply #6 on: November 28, 2021, 09:26:26 am »
Can you wire multimeter in series to show actual current reading? Does 50mA current come back after you connect leads again? What happens if you connect leads and change the load to draw more/less current?
FW issue is likely in this case - it is as if current screen is not updated when no current is flowing (and screen is stuck on old value).
 

Offline AmarukTopic starter

  • Contributor
  • Posts: 46
  • Country: us
Re: Keysight EDU36311A current reading bug?
« Reply #7 on: November 28, 2021, 03:49:52 pm »
Can you wire multimeter in series to show actual current reading?

Here is a follow-up video with a multimeter in series. This time I have a 2 ohm power resistor (100W) and the channels set to 1V with a CC of 800mA. With this video we can conclude that:
1. The power supply is putting out the correct voltage at all times.
2. The current readings are not updated correctly
3. The firmware has a bug

As you can see in the video, both channel 2 and 3 has major issues and fail this test. Channel 1 has a brief (slow update) issue here but worked fine in our earlier experiment with the 100 ohm resistors at 5V. But before I had the camera out I got channel 1 fail the test just like you see on channel 2 and 3 here... Anyway, I am sure Keysight will take a close look at this issue and release a fix soon.


« Last Edit: November 28, 2021, 03:52:51 pm by Amaruk »
 

Offline Keysight DanielBogdanoff

  • Supporter
  • ****
  • Posts: 777
  • Country: us
  • ALL THE SCOPES!
    • Keysight Scopes YouTube channel
Re: Keysight EDU36311A current reading bug?
« Reply #8 on: November 29, 2021, 05:09:08 am »
Interesting find, thank you for documenting! We passed this along to the support team, I'll send you a PM and loop you in as well.
 

Offline WattsThat

  • Frequent Contributor
  • **
  • Posts: 764
  • Country: us
Re: Keysight EDU36311A current reading bug?
« Reply #9 on: November 29, 2021, 06:03:19 am »
On second thought, I hope this is actually a hardware fault. Such a basic software fault should never have gotten out the door at KS.

In the OP’s last video, do I really see a 100W resistor mounted to a block of wood? If so, I wonder how many watts are required to achieve ignition?
 
The following users thanked this post: nctnico

Offline electr_peter

  • Supporter
  • ****
  • Posts: 1300
  • Country: lt
Re: Keysight EDU36311A current reading bug?
« Reply #10 on: November 29, 2021, 10:08:01 am »
Channel 1 current measurement value update rate also seems suspect on disconnection (waits too long on some values). Anyway, I hope Keysight will sort that out.
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 14076
  • Country: de
Re: Keysight EDU36311A current reading bug?
« Reply #11 on: November 29, 2021, 11:03:47 am »
A possible problem could be something like ignoring negative readings. With enough offste the reading may stay negative for a long time.
 

Offline Kean

  • Supporter
  • ****
  • Posts: 2053
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: Keysight EDU36311A current reading bug?
« Reply #12 on: November 29, 2021, 11:21:22 am »
Seems like a bad filtering implementation that ignores readings (possibly negative as Kleinstein suggests) and thus stops updating the display.

As it is rather surprising this has not been reported/caught before, it is possible it is a rare issue and only happens when the channels have a significant zero current offset, or some other hardware issue.
« Last Edit: November 29, 2021, 12:23:13 pm by Kean »
 

Offline HighVoltage

  • Super Contributor
  • ***
  • Posts: 5453
  • Country: de
Re: Keysight EDU36311A current reading bug?
« Reply #13 on: November 29, 2021, 12:05:56 pm »
When the E36300A series power supply was released, it had a similar problem and it was hardware related and Keysight fixed all the problems nicely.

I would think that this is an isolated problem on your EDU36311A. It is hard to believe that such simple problem would not be detected by the Keysight testing team.
There are 3 kinds of people in this world, those who can count and those who can not.
 

Offline AmarukTopic starter

  • Contributor
  • Posts: 46
  • Country: us
Re: Keysight EDU36311A current reading bug?
« Reply #14 on: November 29, 2021, 12:36:37 pm »
Have you used the benchvue software to see if it shows the same there?
Great idea! I would have done that in a flash but only use Mac and Linux computers at home. However, I am looking to get a Windows PC sometime soon. The framwork laptop looks great...

In the OP’s last video, do I really see a 100W resistor mounted to a block of wood? If so, I wonder how many watts are required to achieve ignition?

 :-DD that's a good one! I mostly have regular 0.25W metal film resistors in my drawers so this biggie came in handy for this quick test (0.5W). I normally use it for 4-wire resistance tests and its size and the wood base just makes it easier to handle. But a very good point that you bring up though!

Interesting find, thank you for documenting! We passed this along to the support team, I'll send you a PM and loop you in as well.
Again, one of the reasons I am a big fan of Keysight! Thank you Daniel!

Channel 1 current measurement value update rate also seems suspect on disconnection (waits too long on some values). Anyway, I hope Keysight will sort that out.
Yes, I am sure they will. I also think that the update rate seen here on ch 1 is on the slow side but at least it updates correctly after a few seconds.


A possible problem could be something like ignoring negative readings. With enough offste the reading may stay negative for a long time.
Interesting theory! This is a drastic change in current so you might be onto something!

When the E36300A series power supply was released, it had a similar problem and it was hardware related and Keysight fixed all the problems nicely.
I would think that this is an isolated problem on your EDU36311A. It is hard to believe that such simple problem would not be detected by the Keysight testing team.
Yes, I was also surprised to see this and the power supply really shines of quality in all other aspects!
« Last Edit: November 29, 2021, 01:18:15 pm by Amaruk »
 

Offline 2N3055

  • Super Contributor
  • ***
  • Posts: 6451
  • Country: hr
Re: Keysight EDU36311A current reading bug?
« Reply #15 on: November 29, 2021, 01:49:56 pm »
When the E36300A series power supply was released, it had a similar problem and it was hardware related and Keysight fixed all the problems nicely.

I would think that this is an isolated problem on your EDU36311A. It is hard to believe that such simple problem would not be detected by the Keysight testing team.

And E36100A series was retired within 6 months of release and E36100B series was silently released to fix hardware deficiencies.

Keysight sometimes releases hardware that is not up to their reputation. They are not magical, it's only normal. Even old, mythical HP had revisions, and field corrections and such. And devices were much simpler..
There are numerous problems reported. 34465A had problems with AC measurements. 34470 has weird autocal problem. MSOX3000A/2000 had firmware corruption problems...
Like everybody else, they have to rush products to market to keep up with the competition. And they have same problems as everybody in software development, fast time to market, and "ship, we'll fix it later" mentality is something you cannot avoid.
But they usually admit and fix their mistakes.  Sometimes not, if it is something small and limited scope.
 

Offline J-R

  • Frequent Contributor
  • **
  • Posts: 952
  • Country: us
Re: Keysight EDU36311A current reading bug?
« Reply #16 on: November 29, 2021, 05:44:04 pm »
And E36100A series was retired within 6 months of release and E36100B series was silently released to fix hardware deficiencies.

What issue are you referring to?  Please provide some links.
 

Offline 2N3055

  • Super Contributor
  • ***
  • Posts: 6451
  • Country: hr
Re: Keysight EDU36311A current reading bug?
« Reply #17 on: November 29, 2021, 06:57:30 pm »
And E36100A series was retired within 6 months of release and E36100B series was silently released to fix hardware deficiencies.

What issue are you referring to?  Please provide some links.

Did you miss this topic?

https://www.eevblog.com/forum/testgear/e36300-series-programmable-dc-power-supplies-(e36311a-e36312a-e36313a)/msg1584658/#msg1584658
« Last Edit: November 29, 2021, 07:00:23 pm by 2N3055 »
 

Offline TheSteve

  • Supporter
  • ****
  • Posts: 3743
  • Country: ca
  • Living the Dream
Re: Keysight EDU36311A current reading bug?
« Reply #18 on: November 29, 2021, 08:28:18 pm »
And E36100A series was retired within 6 months of release and E36100B series was silently released to fix hardware deficiencies.

What issue are you referring to?  Please provide some links.

Did you miss this topic?

https://www.eevblog.com/forum/testgear/e36300-series-programmable-dc-power-supplies-(e36311a-e36312a-e36313a)/msg1584658/#msg1584658

I think the confusion may be that you said "And E36100A series was retired within 6 months of release and E36100B series was silently released to fix hardware deficiencies."
I guess you're actually talking about the E363xxA series not E361xxA. And while the hardware in the E363xx series was updated there is no official B model.
VE7FM
 
The following users thanked this post: 2N3055

Offline tooki

  • Super Contributor
  • ***
  • Posts: 11341
  • Country: ch
Re: Keysight EDU36311A current reading bug?
« Reply #19 on: November 29, 2021, 09:05:59 pm »
What happens if you don’t disable channel 1 before testing the other channels? Right now, you’ve tested (in essence) two different situations: channel 1’s behavior with all channels turned on, vs. channels 2 and 3’s behaviors with some channels turned off.

Not saying it makes the demonstrated behavior OK — it clearly isn’t — but it’s a bit of an odd test.
« Last Edit: November 29, 2021, 09:08:32 pm by tooki »
 

Offline AmarukTopic starter

  • Contributor
  • Posts: 46
  • Country: us
Re: Keysight EDU36311A current reading bug?
« Reply #20 on: November 29, 2021, 09:14:18 pm »
What happens if you don’t disable channel 1 before testing the other channels?
Good point. The reason I turn the channels off is to see if the wrong readings stay up when the channel is turned off as they indeed do some times (see channel 2 and 3 in the first video).

As for finer details into exactly when things work and don't work is beyond the scope of my test. I am sure the engineers over at Keysight will test this further if they need to in order to solve this.

However, I first stumbled across this when I only used channel 2 (channel 1 and 3 were off). The circuit I was testing had an on/off power switch and the readings on the EDU36311A confused me... Channel 2 had its voltage on at all times and I thought the power switch in my circuit was broken as the power supply indicated that current continued to flow into my circuit after the switch turned the circuit off...

Thus, to highlight the issue I simulate that use-case by completely disconnecting the cable to the power supply as that makes it clear that no current can flow.
« Last Edit: November 29, 2021, 09:22:52 pm by Amaruk »
 

Offline AmarukTopic starter

  • Contributor
  • Posts: 46
  • Country: us
Re: Keysight EDU36311A current reading bug?
« Reply #21 on: December 02, 2021, 04:36:33 pm »
When the E36300A series power supply was released, it had a similar problem and it was hardware related and Keysight fixed all the problems nicely.

Do you know what the the approximate timeline was for that hardware issue?

I have found the thread on here that talks about the 36300A power-on transients but it is not clear to me when it was first reported to Keysight, re-calls issued by Keysight, and updated/repaired supplies sent out to customers. I just purchased my supply about a month ago so I am hoping this will not take too long to fix. If the EDU36311A issue we talk about here is related to software one would assume that things will be easier and an updated firmware will be posted shortly.

Just a side-note; Am I the only one with this power supply on this forum (aside from Dave)? I know it is a new model and many of you have the more expensive (non EDU) models but If anyone is looking to get a supply from a major brand, this is an excellent model to get. The big display is fantastic and all three channels are full channels (e.g. with adjustable CC on all three channels)!
« Last Edit: December 02, 2021, 04:38:30 pm by Amaruk »
 
The following users thanked this post: tooki

Offline AmarukTopic starter

  • Contributor
  • Posts: 46
  • Country: us
Re: Keysight EDU36311A current reading bug?
« Reply #22 on: December 03, 2021, 09:03:25 pm »
Just an update on the status with this issue:

  • Keysight were able to reproduce the issue with the current readings reported on in this thread
  • Their engineering team went to work on it and in about 48 hours they sent me a beta version of the next firmware to test
  • I upgraded the firmware and I am happy to report that the beta version of the firmware fixed the problem!!!

A big thanks go out to Keysight for showing that they stand behind their products like this!

 
The following users thanked this post: Kean, HighVoltage, tooki, TheSteve, Fgrir, eplpwr


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf