Author Topic: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)  (Read 451861 times)

0 Members and 2 Guests are viewing this topic.

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #600 on: February 04, 2017, 06:59:03 am »
Thanks for this Kean. I add a notice to shunt in series proposal. Introducing the second current monitor doesn't look attractive either.

Online Kean

  • Supporter
  • ****
  • Posts: 2050
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #601 on: February 04, 2017, 07:06:31 am »
Introducing the second current monitor doesn't look attractive either.

Is that because of the cost of the current monitor?  I guess in quantity it still adds a significant cost.
No problem, then you should use a dual analog switch for selecting between the shunt kelvin connections (both upper and lower sides).
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #602 on: February 04, 2017, 07:11:26 am »
Yes, but I'm trying to imagine what will happen when range switching is performing with load connected and output is in CC (or even CV) mode? First it seems that analog switches/multiplexers have to be "make-before-break" too, but is that enough to not introduce glitches?

Online Kean

  • Supporter
  • ****
  • Posts: 2050
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #603 on: February 04, 2017, 07:52:05 am »
Ah, OK.  I wasn't considering that the mode switching could happen while operating, and how that would effect the control loop.

I just took a look at the Agilent 6612B service manual to see how they do it (very nice PSU for testing low current devices).  It appears they always use one shunt for the control loop, and the other is purely used for monitoring and has a clamping circuit across it.  Note that the design is for much lower currents but it may be useful to analyse.  If you're interested check out page 51 for schematic, and 52 for description of operation.  Can be downloaded from Keysight website if you provide your contact details, or various places on the web like http://exodus.poly.edu/~kurt/manuals/manuals/HP%20Agilent/HP%20POWER%20SUPPLIES/CUSTOMERCARE%20(D)/service/66312Aser.pdf
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #604 on: February 04, 2017, 08:03:36 am »
Hm, this is interesting, so precise CC mode is not possible. I'm wondering why is that? Possibly Agilent/Keysight have a good reason for such approach :-//. Are you aware of their any other product that include multi-range and could possibly have schematic exposed at least in the form of building blocks?

Online Kean

  • Supporter
  • ****
  • Posts: 2050
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #605 on: February 04, 2017, 08:57:39 am »
I'm not aware of any, but I haven't really looked.  Not saying it isn't possible, but maybe not as important as precise monitoring.
Maybe open another topic asking this question - you may get more eyes on a new topic than deep in a thread like this...
 

Offline Strada916

  • Frequent Contributor
  • **
  • Posts: 252
  • Country: au
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #606 on: February 04, 2017, 09:12:32 am »
I have a circuit that uses a dpdt switch dual range. Current metering. I'll have a look when I get home later

Sent from my D6503 using Tapatalk

The Bone, the Off-White, the Ivory or the Beige?
 
The following users thanked this post: prasimix

Online Kean

  • Supporter
  • ****
  • Posts: 2050
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #607 on: February 04, 2017, 09:30:00 am »
If switching range during operation, won't you also need to switch to a different input voltage for the control loop?
Otherwise you'll have to scale the current monitor output to match the different ranges.
Either way, I suspect it will be tricky to do without a glitch, especially if in current limiting mode at the time.
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 14067
  • Country: de
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #608 on: February 04, 2017, 09:58:15 am »
Having the MOSFET R_On in the measurement path is bad, as the R_on has a high TC (AFAIK something like 6000 ppm/K). So this would be acceptable only if R_on is less than 1/100 of the shunt - so not really practical.

Many of the other solutions proposed also don't work. If you need a 4 wire shunt for the lower range as well, you essentially have to switch over both sense lines. Adding a CMOS switch to the sense inputs is usually not a problem - 6 digit DMMs like the Keithley2000 use this to do Autozero and for the shunts as well. Just avoid large thermal gradients in this circuit area.

The easy solution is to use the larger of the shunts in 2 wire mode, connected to the voltage sensing contact of the lower shunt. This how it is usually done in the voltmeters up the the very good ones. E.g. look here:
https://www.eevblog.com/forum/metrology/repair-and-experiments-thread-hp-3458a-u2/msg1124455/#msg1124455
The thread about Daves new planed meter also has quite a lot about shunt switching.

Switching the shunt seem to be hard to describe in words, so I attach a cride drawing for the series connection option:
 
The following users thanked this post: Kean, prasimix

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #609 on: February 04, 2017, 10:04:51 am »
I started to understand a beauty of this approach. Thanks Kleinsten, Dave...

)

... and Jaromir for point out Dave's video. I'll implement this concept.

Online Kean

  • Supporter
  • ****
  • Posts: 2050
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #610 on: February 04, 2017, 10:15:58 am »
Most of that info relates to multimeters, and not potential issues doing this in a control loop.
I'd be interested to see if anyone can find a design that goes further than the Agilent 6612B.
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #611 on: February 04, 2017, 10:23:23 am »
I'll test it and let you know. Here is how it will looks like if I understand it correctly. Of course Q1 don't need to be so beefy, it eventually make sense if we don't want to increase BOM:



The great thing about this approach is that if you are for any reason not satisfied with multiple range you can simple remove lower range (leave LO resistor or replace it with 0R jumper) and keep the highest one on the PCB.

Offline Strada916

  • Frequent Contributor
  • **
  • Posts: 252
  • Country: au
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #612 on: February 04, 2017, 10:31:22 am »


Sent from my D6503 using Tapatalk
This is part of a circuit used in one of my analogue PSU's the sense is off the 22k (R20 and R22) resistors. They use a switch to short the excluded resistor.

Although Dave's approach as posted above is more elegant.
« Last Edit: February 04, 2017, 10:36:08 am by Strada916 »
The Bone, the Off-White, the Ivory or the Beige?
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 14067
  • Country: de
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #613 on: February 04, 2017, 11:07:18 am »
Changing the shunt in the current control loop will also change the gain of the loop. This could be a potential problem, if the current loop is tuned very fast. Often this is not needed, as the CC performance is limited by the output capacitors anyway. So it is well possible that the control can accept an extra gain of a factor of 10. The extra shunt will also increase the series resistor seen and this way make current regulation possibly a little easier. For high frequency range the higher shunt this way reduces the loop gain back to where is was before.

To be on the safe side one might want to check stability of the CC mode loop with the higher shunt and maybe adjust for lower speed. If really needed one would have to switch compensation too. The higher shunt could also effect the CV mode regulation a little, as more voltage drop needs to be compensated.

Switching the shunt should be done under user control, usually when adjusting the set current.
 
The following users thanked this post: Kean, jbb

Offline Strada916

  • Frequent Contributor
  • **
  • Posts: 252
  • Country: au
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #614 on: February 04, 2017, 11:27:18 am »
I might be wrong but I don't agree with you Kleinstein.

Sent from my D6503 using Tapatalk

The Bone, the Off-White, the Ivory or the Beige?
 

Offline artag

  • Super Contributor
  • ***
  • Posts: 1061
  • Country: gb
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #615 on: February 04, 2017, 11:35:00 am »
The HP6632B has a dual current range. It isn't obvious from the control, but that's how it works internally.

http://cp.literature.agilent.com/litweb/pdf/5962-8119.pdf

The high current range is always in circuit, the low current range has FETs to bypass it and diodes to clamp it. Each sense resistor has its own amplifier.
 

Online Kean

  • Supporter
  • ****
  • Posts: 2050
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #616 on: February 04, 2017, 03:17:07 pm »
The HP6632B has a dual current range. It isn't obvious from the control, but that's how it works internally.

http://cp.literature.agilent.com/litweb/pdf/5962-8119.pdf

The high current range is always in circuit, the low current range has FETs to bypass it and diodes to clamp it. Each sense resistor has its own amplifier.

Actually, just like the 6612B I mentioned earlier in the thread, on the 6632B the high current range monitor (IMON_H) is always used for the constant current control loop.  The low current range monitor (IMON_L) is only used for "readback" and for enabling the clamping FETs.  This is described reasonably clearly in the "Principles of Operation", but a bit more difficult to follow on the schematics where you'll see only IMON_H enters the current control circuits.
 

Offline texaspyro

  • Super Contributor
  • ***
  • Posts: 1407
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #617 on: February 04, 2017, 05:32:51 pm »
Having better current resolution for lower currents is a worthy goal,  but I think that any practical / cost effective implementation (particularly with dual current shunts) will open quite a few cans of worms.  Getting those worms under control and back into their cans will be quite difficult and introduce more compromises to the unit's performance.   
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 14067
  • Country: de
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #618 on: February 04, 2017, 05:54:48 pm »
I have not followed all of the task, but AFAIK this supply uses a low output impedance power stage. This makes current regulation more difficult - so one really has to check if it still works with a larger shunt and thus a modified CC loop. Chances are it still works.

The shunt chosen is rather small, likely to get a good transient response. However this somewhat limits the performance at low currents in CC mode. So a switchable shunt makes some sense. It is not only the measurement where a larger shunt would be nice to have, but also the control.
 

Offline neslekkim

  • Super Contributor
  • ***
  • Posts: 1305
  • Country: no
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #619 on: February 05, 2017, 03:42:12 pm »
This is a just to the point. Arduino IDE is selected for two reasons: it's open source and "plug&play". I'd like to see another more powerful MCU that is directly supported from environment such as Arduino IDE. I was entertained in one moment with idea to squeeze down the TM4C1294 Connected LaunchPad to Arduino Due/Mega form factor and make it pin compatible with Due or simply put it directly on what is currently called Arduino Shield board. That MCU is equipped with many interesting stuff, and together with more capable display such as one with FT80x controller that could be an attractive combination even for more demanding projects. Again, the question is how to make it simple for use, and Lauchpad has Energia IDE but I'm not so sure that has a so dynamic community like Arduino does.

Talking about SBCs I see more attractive some open source solution such as BeagleBone instead of RPi.

This one is powerfull, 180Mhz M4, with floatingpoint, and uses same arduino IDE, lots of code, and have even more performant display libraries:
https://www.pjrc.com/store/teensy36.html

https://forum.pjrc.com/threads/26305-Highly-optimized-ILI9341-%28320x240-TFT-color-display%29-library

 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #620 on: February 05, 2017, 06:27:08 pm »
Thanks for your suggestion, this one looks very powerful and has attractive price. For the beginning someone should spent some time to see if our firmware can be without many errors compiled when that board is selected from Arduino IDE. 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Going down to calibrated zero ...
« Reply #621 on: February 14, 2017, 10:12:34 am »
Finally I find some time to test one detail that I ignored for too long time. It's how to go down to zero with programmed output voltage or current when channel is calibrated. Liv warns me about that long time ago that if DAC has not bipolar output (i.e. can generate negative voltage) that it's great chance that if DAC is programmed to zero that you cannot get zero on output due to various parts offsets. The simple solution is to shift reference ground of control loops op-amps to few tens of millivolts. In that case with programmed voltage set to zero output value will become negative and in process of calibration firmware will find a non-zero DAC value that relate to zero on the output.
Insisting on going down to zero possibly does not make a lot of sense for output voltage as for limiting current because if e.g. your resolution is set to 10 mA and during calibration you cannot reach e.g. less then 15 mA, firmware will lock your range to 20 mA that is a huge amount for many small loads and CC mode of operation is needed.
I tried that on Power board r5B9 (that changes is added in version r5B10 and later) where you have to do the following:



... on the PCB that looks like this where R42 ground is not cut but resistor is moved from its original position and soldered with one end directly to pin 5 of IC6. On current monitor side R60 and R61 is replaced with single resistor that is soldered on R61 position. Only cut on the PCB that is required is to isolate pin 5 on IC7 from ground. Voltage divider 10K+220R produce about 50 mV derived from +2.5 V voltage reference:



Making this change is not enough. You have to instruct firmware that when channel is not calibrated that take into account ground offset otherwise you'll get strange output values at the beginning of the scale. For example for set 0 V you can expect more then 800 mV or for 0 A a more then 100 mA will be measured without connected load :).

If you have Power board r5B9 and add above described hack you have to add into your conf_user.h the following code:

Code: [Select]
#undef CHANNELS
#define CHANNELS \
    CHANNEL(1, CH_BOARD_REVISION_R5B10_PARAMS, CH_PINS_1, CH_PARAMS_40V_5A), \
    CHANNEL(2, CH_BOARD_REVISION_R5B10_PARAMS, CH_PINS_2, CH_PARAMS_40V_5A) \

When this hack is applied you are not able to reach full scale (i.e. 40 V) with R47, R57 set to 20K. Use e.g. 24K or 24K9 to ensure full scale for calibrated output.
« Last Edit: February 14, 2017, 10:31:13 am by prasimix »
 
The following users thanked this post: Kean

Offline ondreji

  • Contributor
  • Posts: 43
  • Country: au
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #622 on: February 18, 2017, 03:42:33 am »
I'm considering to back this project but the case is too big. The shelf is 27cm depth thus I have to fit everything 23cm or so (max 25cm). I've managed to put everything into 230 x 240 x 130 (LxWxH).

The grey rectangle represent maximum outside enclosure dimension. I assumed 50mm height (incl heatsink) for power modules. Additional 2cm (per module) can be gained by rotating aux supply. The ultimate question is: would everything  fit into my box? I would say yes...

PS. It's probably too late but it would be great ho have support level with all the electrical related parts: Core Modules (1 x AUX power supply module, 2 x power boards, and 1 x Arduino shield), power board heatsinks, Arduino Due, LCD module, and Mean Well PSUs (not sure about including wire harness and binding posts)
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #623 on: February 18, 2017, 07:14:50 am »
Hi, are you talking about depth or width? Suggested modules constellation is not possible with current Arduino shield at least not without adding extender 26-pin flat cables. Also if Power boards are far away from each other you'll need two instead of one cooling fan. That will also need additional fan speed controller.
Currently only what is possible is to rotate AC/DC power modules, increase enclosure depth and try to stay within width of Arduino shield that is 220 mm.

When time comes I'll try to redesign Arduino shield for "tower"/vertical enclosure where on lower section can have 2 x Power modules and 2 x AC/DC power modules or mains transformer on the upper level. AUX PS could stay on the rear panel.

Anyway, with the current width PSU is a solid base to put a DSO/MSO on top :)



Offline ondreji

  • Contributor
  • Posts: 43
  • Country: au
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)
« Reply #624 on: February 18, 2017, 10:17:58 am »
Suggested modules constellation is not possible with current Arduino shield at least not without adding extender 26-pin flat cables. Also if Power boards are far away from each other you'll need two instead of one cooling fan. That will also need additional fan speed controller.
Currently only what is possible is to rotate AC/DC power modules, increase enclosure depth and try to stay within width of Arduino shield that is 220 mm.

It looks that I should keep Arduino & two power boards together as they are now. I'll go for taller enclosure to accomodate the shield on tow of two AC/DC PSUs side by side (this will add around 4-5cm) under it. AUX power supply will go to the space where PSU are at the moment.... Thanks!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf