Author Topic: Quad output bench power supply  (Read 1639 times)

0 Members and 1 Guest are viewing this topic.

Offline kaffineTopic starter

  • Newbie
  • Posts: 6
  • Country: us
Quad output bench power supply
« on: November 17, 2024, 06:00:32 am »
Work is looking to upgrade our bench power supplies for the hardware debug team.

Prefer a quad output but a triple output would likely work.  We need to be able to have the power turn on sequenced, turn on 3.3V management power rail then turn on the 12V and 5V and 3.3V rails ~250mS later. 

A few years ago we were using cheap brand of the week bench power supplies after we blew up a few boards due to the overshoot on power on the boss went out and ordered some Keysight E3633A power supplies.  While the Keysight supplies have a much better output they are not well liked by those using them and the single output is very limiting having 3 of them on my bench takes up a lot of room. 



Requirements for the power supply

3 outputs preferably 4.
be able to sequence turn on of outputs preferably using front panel controls
Be able to output 12V at 10 amps on at least 1 output
Remote control prefer ethernet interface however serial will work.
Be able to supply 48V putting 2 outputs in series is ok

Preferer front ports with multi use banana jack / spade lug however open to other configurations.
would be nice to be able to power cycle power supply and have it come up with output voltages or a foot pedal control (I'd be surprised if this is an option anymore though)

Right now I am looking at Rohde & Schwarz HMP4040 and Instek PSW-1080L114.  I'm expecting the price to be in the $3k-4k USD range.  I'm realizing how out of touch I am when it comes to test equipment most places I have work the test equipment has been from the 80s and 90s I'm not even sure who all makes good test equipment anymore.  Any recommendations are greatly appreciated.  Thank you.


 
 

Offline tautech

  • Super Contributor
  • ***
  • Posts: 29813
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Quad output bench power supply
« Reply #1 on: November 17, 2024, 06:25:59 am »
SPD4306X should be worth a look:
https://siglentna.com/power-supplies/spd4000x-series-programmable-linear-dc-power-supply/

We have a thread on these for further info:
https://www.eevblog.com/forum/testgear/new-siglent-spd4000x-series-power-supply/

Member and Youtuber Defpom gets his soon and plans a review and teardown.
Avid Rabid Hobbyist.
Some stuff seen @ Siglent HQ cannot be shared.
 

Offline mahi

  • Regular Contributor
  • *
  • Posts: 128
  • Country: 00
Re: Quad output bench power supply
« Reply #2 on: November 17, 2024, 10:11:40 am »
Since you are considering the Rohde & Schwarz HMP4040...

We need to be able to have the power turn on sequenced, turn on 3.3V management power rail then turn on the 12V and 5V and 3.3V rails ~250mS later.

There's no way to do this from the front panel (*). You can set a group of outputs to turn on/off together, but you cannot set a delay between them.

You can achieve this via SCPI remote control (optional interface required!).

*: It might be possible via the arbitrary waveform functionality. I tried the arbitrary waveform functionality directly on the device, but couldn't get it to keep the output voltage after the cycle ended (it either repeats the cycle, or goes back to 0V). Maybe with a bit more experimentation, but it's extremely frustrating to enter the arbitrary waveform directly on the device. It's supposedly much easier with the HMExplorer/EasyARB software for PC, but Rohde & Schwarz recently blocked all software/documentation downloads for non-commercial customers, so I can't test it anymore. If you are really interested in the HMP4040, get in touch with R&S and explain your needs.

Update: A friendly forum member sent me the HMExplorer/EasyARB software so I could test this. Yes, it is possible to achieve this with the arbitrary waveform functionality. See attached screenshots for settings. It should be noted that the outputs are not really switched on with a delay. They are enabled simultaneously, but channel 2 is set to 0 V / 0 A for 250 ms, and then changed to the desired output voltage. Once the cycle ends, the outputs remain active at the set voltages. Not sure why I could not get that to work from the device itself, as the settings are identical as far as I can see.

Quote
A few years ago we were using cheap brand of the week bench power supplies after we blew up a few boards due to the overshoot on power on the boss went out and ordered some Keysight E3633A power supplies.

I'm not sure what kind of power-on overshoot you mean, but the HMP4040 is not without overshoot either. More info in Lab Power Supply Turn ON and OFF Characteristics.

Quote
Requirements for the power supply
3 outputs preferably 4.

The HMP4040 got you covered.

Quote
be able to sequence turn on of outputs preferably using front panel controls

See earlier. If the sequence you want is possible via the arbitrary waveform feature, yes, you can use the front panel to switch on/off the outputs.

Quote
Be able to output 12V at 10 amps on at least 1 output

No issue with the HMP4040. The maximum power is 160 W per channel and a combined total of 384 W for all channels.

Quote
Remote control prefer ethernet interface however serial will work.

There are 3 optional interfaces for the HMP4040: USB/LAN interface (HO732), RS-232/USB (HO720) and GPIB (HO740). The interfaces are rather pricey, though...

The web interface is... disappointing. You can only see a screenshot of the display and control the device via SCPI commands. No nice real-time full front panel interface like the modern R&S oscilloscopes have. The SCPI control is more useful from a command line or scripts.

I have attached screenshots of the web interface. Note that the screenshots say "HAMEG" - that's because I retrofitted my R&S HMP4040 with the HO732 interface of an older Hameg power supply. I have no reason to assume that the current HO732 is more advanced.

Quote
Be able to supply 48V putting 2 outputs in series is ok

The HMP4040 can do up to 128 V in series and 40 A in parallel, but there's no built-in series/parallel operation. You must do it with external wiring.

Quote
Preferer front ports with multi use banana jack / spade lug however open to other configurations.

The HMP4040 front panel accepts only 4 mm banana jacks. There's no way to insert loose wires or spade lugs. On the back are screw terminals, though.

Quote
would be nice to be able to power cycle power supply and have it come up with output voltages or a foot pedal control (I'd be surprised if this is an option anymore though)

There's no foot pedal control input, but since you can control the power supply via SCPI you could make one...

If you power cycle the power supply, the outputs default to OFF. There's no way to change that. But the last used group of outputs is selected by default, so all you have to do is press the Output button to restore the last situation.
« Last Edit: November 18, 2024, 05:54:37 am by mahi »
 

Offline glitchcatcher

  • Contributor
  • Posts: 35
  • Country: at
Re: Quad output bench power supply
« Reply #3 on: November 17, 2024, 11:36:15 am »
Hi,
I would like to suggest the Aim-TTi MX100Q S2 although it doesn't have 10A on a single output. All the other requirements should be met:

https://www.aimtti.com/go/mx/index.php?p=index-mx100q

Requirements for the power supply

Be able to output 12V at 10 amps on at least 1 output

Quote
I'm not even sure who all makes good test equipment anymore. 
Aim-TTi definitely is one of them. I have lots of their equipment and while some of it might look slightly outdated, it works flawlessly and is rock-solid. Also feature and spec-wise they are impressive especially for the price points.
I visited them at the Electronica fair in Munich earlier this week and they have some nice new products upcoming. They are also updating the looks of their gear to the more modern looking color displays.
clandestine one-man UNI-T fan club
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 28429
  • Country: nl
    • NCT Developments
Re: Quad output bench power supply
« Reply #4 on: November 17, 2024, 12:20:15 pm »
Work is looking to upgrade our bench power supplies for the hardware debug team.

Prefer a quad output but a triple output would likely work.  We need to be able to have the power turn on sequenced, turn on 3.3V management power rail then turn on the 12V and 5V and 3.3V rails ~250mS later. 

Preferer front ports with multi use banana jack / spade lug however open to other configurations.
would be nice to be able to power cycle power supply and have it come up with output voltages or a foot pedal control (I'd be surprised if this is an option anymore though)

Right now I am looking at Rohde & Schwarz HMP4040 and Instek PSW-1080L114.  I'm expecting the price to be in the $3k-4k USD range.  I'm realizing how out of touch I am when it comes to test equipment most places I have work the test equipment has been from the 80s and 90s I'm not even sure who all makes good test equipment anymore.  Any recommendations are greatly appreciated.  Thank you.
AFAIK the R&S has some current limit issues at turn on and the PSW-1080L114 is a switching PSU which is what you don't want in an R&D environment. A switching PSU pushes out a lot of noise which can interfere with measurements. The 12V 10A requirement is something you might want to look for in a seperate PSU.

If I where you I'd look at the GW Instek GPP-4323 or Keysight E36313A Both are linear (low noise) power supplies. In some areas the GPP-4323 is sold with sockets but a version with normal bindings posts does exist.
« Last Edit: November 17, 2024, 02:31:22 pm by nctnico »
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Online Martin72

  • Super Contributor
  • ***
  • Posts: 7019
  • Country: de
  • Testfield Technician
Re: Quad output bench power supply
« Reply #5 on: November 17, 2024, 01:19:04 pm »
Quote
GW Instek GPP-4323

This has also been on my list for a while.
First of all, one for me privately and I think that's good enough, some are ordered for work.

Offline mawyatt

  • Super Contributor
  • ***
  • Posts: 4120
  • Country: us
Re: Quad output bench power supply
« Reply #6 on: November 17, 2024, 02:08:12 pm »
A few years ago we were using cheap brand of the week bench power supplies after we blew up a few boards due to the overshoot on power on the boss went out and ordered some Keysight E3633A power supplies.  While the Keysight supplies have a much better output they are not well liked by those using them and the single output is very limiting having 3 of them on my bench takes up a lot of room. 

One reason we've always insisted on quality supplies that don't produce "surprise glitches" and have stable low noise outputs. This is especially true in an R&D environment where one doesn't need to be dealing with faulty PS issues. We lost some very expensive custom chips sets long ago to a PS glitch, and learned a lesson the hard way!!

We acquired the GPP-4323 a few years back and have used it often, with no issues other than recall it requires a special Windows Driver for remote communication by RS232/USB. It can't deliver 10amps tho, with paralleled outputs this is ~6amps.

Best
Curiosity killed the cat, also depleted my wallet!
~Wyatt Labs by Mike~
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 28429
  • Country: nl
    • NCT Developments
Re: Quad output bench power supply
« Reply #7 on: November 17, 2024, 06:21:11 pm »
For more power and higher currents / voltages (to satisfy the 12V / 10A requirement) the GW Instek GPP-3610H and Keysight E36232A look very interesting. Both are single output though.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline MarkL

  • Supporter
  • ****
  • Posts: 2302
  • Country: us
Re: Quad output bench power supply
« Reply #8 on: November 17, 2024, 06:55:11 pm »
Hi,
I would like to suggest the Aim-TTi MX100Q S2 although it doesn't have 10A on a single output. All the other requirements should be met:

https://www.aimtti.com/go/mx/index.php?p=index-mx100q
...
Also take a look at the Aim-TTi MX180TP.  It's in the same family and checks all the same boxes with power-on sequencing and so on.   It's only triple output, but that was stated as possibly ok.

It can fulfill the 12V @ 10A requirement without disabling other outputs, and can also do 48V on a single output.  All the output combinations are listed in the link below.  The table on that page is a little confusing and it does not mean, for example, "Range 1" sets all the outputs according to that row.  The outputs are independent except as noted.   The "TP" version has the RS232/USB/Ethernet-LXI interfaces installed, the "T" does not.

  https://www.aimtti.com/go/mx/index.php?p=index-mx180t
 
The following users thanked this post: glitchcatcher

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 28429
  • Country: nl
    • NCT Developments
Re: Quad output bench power supply
« Reply #9 on: November 17, 2024, 07:02:16 pm »
Keep in mind that the TTi PSUs are switching PSUs. Their noise performance isn't stellar. I wouldn't want those on my bench for doing R&D work.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline MarkL

  • Supporter
  • ****
  • Posts: 2302
  • Country: us
Re: Quad output bench power supply
« Reply #10 on: November 17, 2024, 07:41:59 pm »
Keep in mind that the TTi PSUs are switching PSUs. Their noise performance isn't stellar. I wouldn't want those on my bench for doing R&D work.
The MX series are switching supplies with a linear post-regulator.  It depends on what you're doing, but the noise is not bad.  Ripple & Noise from the spec sheet, 20MHz BW limited:

  Conditions: O/P 1&2 loaded at 15V/10A, O/P 3 loaded at 5.5V/3A, or for O/P 1 loaded at 15V/20A, 30V/12A, or 60V/6A (O/P 2 disabled):

    All outputs typically <2mVrms, <15mV pk-pk; 3mVrms max.

    O/P 1 loaded at 120V/3A: typically <3mVrms, <20mV pk-pk; 6mVrms max.

The OP did not specify noise requirements, but for development bench use I agree a full switcher is not a good choice.
 
The following users thanked this post: glitchcatcher

Offline KungFuJosh

  • Super Contributor
  • ***
  • Posts: 3166
  • Country: us
  • TEAS is real.
Re: Quad output bench power supply
« Reply #11 on: November 17, 2024, 07:48:07 pm »
Siglent SPD4306X:

CH1: Voltage 0 to 15 V, Current 0 to 1.5 A
CH2: Voltage 0 to 30 V, Current 0 to 6 A
CH3: Voltage 0 to 30 V, Current 0 to 6 A
CH4: Voltage 0 to 15 V, Current 0 to 1 A

CH2 and CH3 series and parallel connections - parallel up to 12A, series up to 60V

https://siglentna.com/wp-content/uploads/dlm_uploads/2024/08/SPD4000X_DataSheet_EN01B.pdf
https://siglentna.com/wp-content/uploads/dlm_uploads/2024/08/SPD4000X_UserManual_EN01B.pdf
"Right now I’m having amnesia and déjà vu at the same time. I think I’ve forgotten this before." - Steven Wright
 

Offline kaffineTopic starter

  • Newbie
  • Posts: 6
  • Country: us
Re: Quad output bench power supply
« Reply #12 on: November 18, 2024, 01:05:56 am »
Thanks for the suggestions. 

The old power supplies we were running are HY3005F-3 with various brand names Dr Meter is the one I can remember.  The issue we had with them is when set for 5V when power cycled they would over shoot by enough that it burnt up a few things.  Reading the linked thread it seems the overshoot issue with the HMP4040 is when transitioning from CV to CC it isn't overshooting the CV limit just overshooting the CC limit.  We don't use CC much mostly for locating shorts and we set the voltage to 0.5 and current limit to 2 amps to essentially make a 4 wire ohm meter. 

As far as switching vs linear while it might be nice to have a linear supply low noise while troubleshooting all of the chassis the boards would be installed in are running switching supplies. 

A good portion of our  boards wont boot with only 5 amps and a few wont boot with 7 amps and they are wanting more and more power.



 

Offline J-R

  • Super Contributor
  • ***
  • Posts: 1321
  • Country: us
Re: Quad output bench power supply
« Reply #13 on: November 18, 2024, 07:09:29 am »
Have all the power supplies power up, enable output, then feed through a relay board that you can control independently.

 

Offline Chicane

  • Contributor
  • Posts: 25
  • Country: us
Re: Quad output bench power supply
« Reply #14 on: November 18, 2024, 11:43:59 am »
UNI-T UDP3305S
4ch (2x 0-30V, 5A ; 1x 0-6V 3A ; 1x5V 3A), 330W, programmable through LAN and RS232, SCPI remote control. It has good reviews and you can even find teardowns on Youtube. Best part is, you can buy 7-8 of those for the price of one R&S. Saw a review of that tested it and no overshoots. It appeared to be very well designed and made. 
« Last Edit: November 18, 2024, 11:48:41 am by Chicane »
 

Offline Lochstreifen

  • Newbie
  • Posts: 6
  • Country: de
Re: Quad output bench power supply
« Reply #15 on: November 18, 2024, 01:53:00 pm »
Aim TTI MX series can be configured to power up with outputs enabled in its previous setting. However, in this case sequencing is not applied.
There are also four channel models. These go only up to 6A though.
https://www.aimtti.com/product-category/dc-power-supplies/aim-mxseries
« Last Edit: November 18, 2024, 02:00:28 pm by Lochstreifen »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf