Author Topic: Lab power supply over-voltage when turning on  (Read 2909 times)

0 Members and 1 Guest are viewing this topic.

Offline raresvinteaTopic starter

  • Contributor
  • Posts: 15
Lab power supply over-voltage when turning on
« on: June 19, 2019, 08:16:01 am »
Hello, we have some lab power supplies that have a strange and annoying issue. When you turn it on, the set voltage is 12v and it overshoots with 72V for a small time and kills boards or if you turn it on it remain on 65V and we need to turn it off again...
How can i manage this?

-limit somehow this peak
-detect over voltage and disconnect the output
-make a simple on off switch, first turn on the power supply and push a button that will turn on the output and it will compare the output voltage with a set limit and stays off.
 

Offline pevonous

  • Newbie
  • Posts: 1
  • Country: us
Re: Lab power supply over-voltage when turning on
« Reply #1 on: June 19, 2019, 01:20:05 pm »
Have you tired to open it up and check inside of it, sometimes, there has a problem of MOSFEET issue on it. I don't know if it can help you or not.
 

Offline raresvinteaTopic starter

  • Contributor
  • Posts: 15
Re: Lab power supply over-voltage when turning on
« Reply #2 on: June 19, 2019, 01:22:42 pm »
This symtom is rare, but when it is present, blew the board
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: Lab power supply over-voltage when turning on
« Reply #3 on: June 20, 2019, 03:07:05 am »
That sort of thing is not as rare as it should be.  It indicates a poor design.
 
The following users thanked this post: Yansi

Offline DaJMasta

  • Super Contributor
  • ***
  • Posts: 2393
  • Country: us
    • medpants.com
Re: Lab power supply over-voltage when turning on
« Reply #4 on: June 20, 2019, 04:48:12 am »
Could certainly be a design issue with the PSU, but make sure you're not adding too much inductance to the front end, it can magnify the issue if the switch on time is pretty quick.  Otherwise, if you're manually switching the output, maybe just putting a DPDT switch between the output of the supply and the input of your device and leaving it disconnected while the supply is first switched on, then connecting it later.  It's a stopgap, but it's a simple solution if you're not using the programmable switching features the supply may have.

Another option would be trying to load the output and seeing if the response changes.  Even a 10k load resistor in parallel with your load or something may be enough to change the way the supply responds initially to that output is more in line with what it should be, or perhaps an NTC thermistor in series with the power line to take the edge off the spike.  There's also the option of a TVS diode in parallel rated a bit above your 12V requirement to take the edge off.

How are you testing this, though?  72V is a LOT for a normal linear lab power supply who's normal transformer taps used for a 12V output level for a power supply... it likely cannot be connected to that voltage internally unless the maximum supply voltage exceeds 72V, so I suspect some sort of probing issue.  Again inductance is a concern if the power on is a fast edge, but this isn't typically how lab supplies operate.  Are you referencing your device to a ground different ground?  Is the ground connected to the power supply floating ground before the power is switched on?

While some of the earlier measures could alleviate the symptoms, I really wonder if that voltage is actually present and whether it should even be possible under normal operation.
 

Offline H713

  • Regular Contributor
  • *
  • Posts: 180
  • Country: us
Re: Lab power supply over-voltage when turning on
« Reply #5 on: June 20, 2019, 07:19:20 am »
Inductive spiking can easily reach 72 volts. Any chance you have a schematic? It's hard to troubleshoot something if you don't know what it is. Is this a single power supply doing this, or do you have multiple? If it's just a single one out of many, then I would expect a fault, if they all do this, then we're looking at a bad or misused design.
 

Offline raresvinteaTopic starter

  • Contributor
  • Posts: 15
Re: Lab power supply over-voltage when turning on
« Reply #6 on: June 20, 2019, 07:23:57 am »
More or less the load is represented by a headunit, with power supplies and so on.
Max voltage of the PS is 65V and when it does the problem it shits the output to the max, the power supply is a EA-PS 9065-20A, it's a german one. We have others like EA-PS 7032-10A that have a similar issue. I am thinking to do a simple but reliable device to compare the output with a voltage obtained from a zenner?! and if it's higher to shut off a mosfet that can handle 10A
 

Offline capt bullshot

  • Super Contributor
  • ***
  • Posts: 3033
  • Country: de
    • Mostly useless stuff, but nice to have: wunderkis.de
Re: Lab power supply over-voltage when turning on
« Reply #7 on: June 20, 2019, 08:02:39 am »
EA aren't the best lab power supplies from my experience, I tend to avoid them. Some of them might be OK, but some others look rather Korean or Russian inside.
Turn-on transients is something that is bad, but happens even to Keysight with their E36312 (and similiar) supplies. So to me it's good practice to first turn on the supply and then connect the load. Normally I don't leave a DUT connected to a lab power supply when I shut down the lab.

Anyway, this one has a OV protection, that should do what you want (typically it shorts the output when it triggers and you have to restart the supply). Have you tried the OVP setting?
Safety devices hinder evolution
 

Offline raresvinteaTopic starter

  • Contributor
  • Posts: 15
Re: Lab power supply over-voltage when turning on
« Reply #8 on: June 20, 2019, 08:05:57 am »
I've set the OVP because my colleagues didn't do this.

EA-PS 7032-10A But this one doesn't have a button to switch on/off the load, i would to a small pcb with protection and a small push button to avoid unplug every time the DUT
 

Offline MT.Shariq

  • Newbie
  • Posts: 2
  • Country: sa
Re: Lab power supply over-voltage when turning on
« Reply #9 on: June 20, 2019, 09:15:28 am »
Have you opened the unit for a look inside?
 

Offline raresvinteaTopic starter

  • Contributor
  • Posts: 15
Re: Lab power supply over-voltage when turning on
« Reply #10 on: June 20, 2019, 10:04:14 am »
Nope, I will take some time to open
 

Online Siwastaja

  • Super Contributor
  • ***
  • Posts: 9328
  • Country: fi
Re: Lab power supply over-voltage when turning on
« Reply #11 on: June 20, 2019, 12:46:05 pm »
Simple,

A broken-by-design, piece-of-shit trap garbage device. The options are,

1) Throw it away, buy a proper one,
2) Only use it for secondary, non-critical purposes, add a warning sticker
3) Take it apart, reverse-engineer, analyze the circuitry, find the problem and develop a fix. Can be a great learning experience.
 

Offline Conrad Hoffman

  • Super Contributor
  • ***
  • Posts: 2077
  • Country: us
    • The Messy Basement
Re: Lab power supply over-voltage when turning on
« Reply #12 on: June 20, 2019, 06:03:19 pm »
That's a completely unacceptable flaw. Come up with a fix or throw it away, lest it cost you ten times what the supply cost in terms of wrecked boards and failed projects.
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1048
  • Country: ca
Re: Lab power supply over-voltage when turning on
« Reply #13 on: June 20, 2019, 08:25:20 pm »
If the current limit is set to minimum, does the output voltage also go to zero?  Many lab supplies do this, and I use this feature to set the output voltage to zero without shutting the mains switch off.  With any luck you may find that with the current limit set to minimum, the supply does not generate a voltage transient.
 

Offline H713

  • Regular Contributor
  • *
  • Posts: 180
  • Country: us
Re: Lab power supply over-voltage when turning on
« Reply #14 on: June 21, 2019, 06:30:21 am »
The lazy solution is just to add a load switch and a warning sticker. IMO, it is a feature that all bench power supplies should have. A big transient on power up is something that is ideally avoided in power amplifiers and absolutely completely unacceptable in a bench power supply, especially one of that magnitude. One other solution is to have a relay and a 1 second delay timer so the load is connected after the power switch is turned on. Many power amplifiers use this solution to avoid the loud "thump" when it is turned on (this is actually part of the speaker protection circuit- last thing speakers want to see is DC).

At least for me, any power supply that does this goes off the bench immediately. The right solution to this problem is to go backwards and figure out why the hell it does this in the first place, and correct the design. Is this a linear or switching supply?
 

Offline perieanuo

  • Frequent Contributor
  • **
  • Posts: 914
  • Country: fr
Re: Lab power supply over-voltage when turning on
« Reply #15 on: June 21, 2019, 08:55:55 am »
Repair it.
Third option is just not realistic, what if 220V goes off and on when you turn your head?
Option two, same thing.you supply a 3v3 board with 3.3v and suddenly your ps gives 12.you blow the board...how you can limit the overvoltage???
Repair or remaster the supply, don't look for shortcuts
Sanatate :)


Envoyé de mon iPad en utilisant Tapatalk
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Lab power supply over-voltage when turning on
« Reply #16 on: June 21, 2019, 09:35:09 am »
Simple,

A broken-by-design, piece-of-shit trap garbage device. The options are,

1) Throw it away, buy a proper one,
2) Only use it for secondary, non-critical purposes, add a warning sticker
3) Take it apart, reverse-engineer, analyze the circuitry, find the problem and develop a fix. Can be a great learning experience.
This!

If it's new enough to be covered by the warranty, simply return it. The seller will be legally obliged to: exchange it for a new, working one, repair it, or refund you.
 
The following users thanked this post: Yansi, Siwastaja, 2N3055

Offline Chris56000

  • Frequent Contributor
  • **
  • Posts: 913
  • Country: gb
Re: Lab power supply over-voltage when turning on
« Reply #17 on: June 21, 2019, 08:10:31 pm »
Hi!

Apart from the hassle of arguing it out with his original Supplier, with all the aggro of waiting for a replacement to be supplied, the OP could well do that and find out the replacement that comes has an even worse overshoot, and the Supplier/OEM denies it's a design fault, which it most assuredly is!

The reason a turn-on overshoot occurs is because the control error amplifier is operative before the rest of the PSU has started to provide a feedback-voltage to it's voltage-control input.
 
By putting a time-delay on the reference voltage for the PWM controller, at the moment of power-on, the Ref. volts will be at, or near chassis or primary 0V potential, so the PWM control amplifier will give maximum negative-feedback control output at the instant of power-on, helping to prevent an unwanted high-voltage surge from the SMPSU's power-transformer.

As the reference voltage builds up from chassis (or 0V primary potential), the PWM Control Amplifier's negative feedback will reduce until the normal amount used in the circuit design is reached, by which time the output from the SMPSU power transformer will have built up enough for the output-voltage feedback loop to take over.

I have attached some notes from Keith Billing's "Switch-Mode Power Supply Handbook" which discusses the theory and gives a simple practical circuit the OP can try, and it works if the Control Amplifier is part of a '494, '3524, 741s, discretes, etc., etc., as long as both it's reference and error-voltage control I/Ps are accessible, in either linear or SM-based power-supply units!

If the OP opens up his unit and advises us what PWM controller is in use, we can suggest a circuit idea for him to try! I've assumed, from the OP's description of the overshoot-spike, that his PSU unit is likely to be a switch-mode based one - a turn-on overshoot spike of 70V certainly suggests so!

Chris Williams

PS!

I believe Siglent had a nasty issue with turn-on overshoot with one of their models and had to cajoled into developing a cure for it!

https://www.eevblog.com/forum/testgear/siglent-spd3303d-review/
« Last Edit: June 21, 2019, 08:49:01 pm by Chris56000 »
It's an enigma that's what it is!! This thing's not fixed because it doesn't want to be fixed!!
 

Offline ifrenide

  • Regular Contributor
  • *
  • Posts: 57
  • Country: dz
Re: Lab power supply over-voltage when turning on
« Reply #18 on: June 21, 2019, 08:41:50 pm »
oops. this is not a problem. I have a power supply E S reland close to you power supply. with output voltage 30v , 5 Amps max. when I fully turn the "fine" knob it show me 32.5 v which is normal,and when I turn it anti clockwise  the voltage drops to 28 v . so I fine adjusted to 30v. turn fine knob fully anti clockwise and you will get about 55v or litle bit more. what you call it problem is an option  :popcorn:
 

Offline ifrenide

  • Regular Contributor
  • *
  • Posts: 57
  • Country: dz
Re: Lab power supply over-voltage when turning on
« Reply #19 on: June 21, 2019, 09:35:36 pm »
I'm sorry I didn't read all what you say.
voltage is adjusted with pot .so your pot is not working at all, that's why the unit shows you the max output voltage , it is like you don't have a pot and it's just be shorted.try to replace it .
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: Lab power supply over-voltage when turning on
« Reply #20 on: June 21, 2019, 09:36:15 pm »
The reason a turn-on overshoot occurs is because the control error amplifier is operative before the rest of the PSU has started to provide a feedback-voltage to it's voltage-control input.

It is not always because the error amplifier is inoperative.  It can also be caused by integrator windup in the error amplifier.  Some designs do not require anything extra to prevent problems during startup or shutdown.

In the example below from a Tektronix PS503A, Q15 and Q115 (marked red) clamp the outputs during shutdown to prevent output voltage spikes (1) but no special precautions were necessary for startup because the reference at Q25 and Q30 (marked green) does not start until the error amplifiers are already operating.  The lack of additional external frequency compensation in this design prevents excessive integrator windup.

The Tektronix PS501, which is basically the positive half of the PS503A, does not even have the shutdown clamp circuit because there is no negative output to interact with the positive output.  (1)

(1) I suspect the shutdown clamps were added to the PS503A design after discovering that there was an issue.  I doubt they were originally designed in.  I do not know if they were in the PS503 because I do not have those schematics.
« Last Edit: June 21, 2019, 09:40:59 pm by David Hess »
 

Offline lordvader88

  • Frequent Contributor
  • **
  • !
  • Posts: 935
  • Country: ca
Re: Lab power supply over-voltage when turning on
« Reply #21 on: June 21, 2019, 09:44:32 pm »
Have you tired to open it up and check inside of it, sometimes, there has a problem of MOSFEET issue on it. I don't know if it can help you or not.
I don't know much, but I was thinking a MOSFET too because I had 1 do something similar in a load tester. And also learned better linear PSU's use MOSFET's...can't wait to make 1.


But there's other things too, maybe there is no FET
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Lab power supply over-voltage when turning on
« Reply #22 on: June 25, 2019, 10:25:20 am »
Hi!

Apart from the hassle of arguing it out with his original Supplier, with all the aggro of waiting for a replacement to be supplied, the OP could well do that and find out the replacement that comes has an even worse overshoot, and the Supplier/OEM denies it's a design fault, which it most assuredly is!
Then send it back again and demand a refund.

I know what you mean: fixing something can be educational and rewarding, but it's not the buyer's job to fix a defective product. I also think doing so would be a wrong, since it would allow the seller to get away with selling defective products. Consumers shouldn't have to put up with this crap.
 
The following users thanked this post: Yansi, Siwastaja

Online Siwastaja

  • Super Contributor
  • ***
  • Posts: 9328
  • Country: fi
Re: Lab power supply over-voltage when turning on
« Reply #23 on: June 25, 2019, 11:56:44 am »
I know what you mean: fixing something can be educational and rewarding, but it's not the buyer's job to fix a defective product. I also think doing so would be a wrong, since it would allow the seller to get away with selling defective products. Consumers shouldn't have to put up with this crap.

Yes. IMHO, people are too shy when it comes to completely normal things like making things right by the person/company that fucked it up. I have manufactured and sold things; I have made mistakes; I have never felt bad when they have been pointed out to me, and I'm asked to make it right, in one way or another. It's normal customer service.

There are companies and people unable to be responsible, but you don't need to assume that. Assume they want to set it right, and be polite and clear, and most often everything goes well, and you don't need to suffer with faulty products. Unless you genuinely want to see it as an learning opportunity, of course.

The same applies to reselling & retail. They are making a profit for a reason. It's part of their business risk that some units fail and come back, and it's their job to communicate it further to the manufacturer.
« Last Edit: June 25, 2019, 12:09:21 pm by Siwastaja »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf