Author Topic: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)  (Read 451916 times)

0 Members and 1 Guest are viewing this topic.

Offline Aigor

  • Contributor
  • Posts: 29
  • Country: it
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #325 on: May 25, 2016, 12:11:37 pm »
I'm waiting my pcb set  and other stuff to try to build your amazing creature, when i'm ready i'll post my build
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #326 on: May 25, 2016, 12:12:59 pm »
I'm waiting my pcb set  and other stuff to try to build your amazing creature, when i'm ready i'll post my build

What revision you are decided to build?

Offline Aigor

  • Contributor
  • Posts: 29
  • Country: it
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #327 on: May 25, 2016, 12:40:03 pm »
I'm waiting my pcb set  and other stuff to try to build your amazing creature, when i'm ready i'll post my build

What revision you are decided to build?

I'm courageous i decide to try rev5
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #328 on: May 25, 2016, 01:41:15 pm »
Ok, welcome to club  :clap:. Your eventual feedback is appreciated.
Anyway please note that Arduino shield will be definitely modified in layout (not in functionality). But don't worry I hope that group buy milestone will be reached and I can send you one shield when it will be available (ok, you have to resolder everything, but that could be a nice exercise :)).
There is also few minor changes in BOMs that I still didn't upload on the github. So many things to do, and not so many time.

Offline Aigor

  • Contributor
  • Posts: 29
  • Country: it
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #329 on: May 25, 2016, 01:49:39 pm »
Ok, welcome to club  :clap:. Your eventual feedback is appreciated.
Anyway please note that Arduino shield will be definitely modified in layout (not in functionality). But don't worry I hope that group buy milestone will be reached and I can send you one shield when it will be available (ok, you have to resolder everything, but that could be a nice exercise :)).
There is also few minor changes in BOMs that I still didn't upload on the github. So many things to do, and not so many time.
Thank you  ;D I'm going to improve my knowledge building your amazing project.
I would like to a report a strange thing happened when i tried to upload firmware into arduino due as you showed in video.
You must use IDE 1.6.9 otherwise you get a error, compiling and uploading went fine, but if i look into serial monitor, it seems blocked into ethernet init and nothing happens even if i wait more then 30seconds.
But for now it's' not a problem, i'm going to try after complete built.
 

Offline Aigor

  • Contributor
  • Posts: 29
  • Country: it
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #330 on: May 25, 2016, 05:42:09 pm »
Damn! i have same tft screen  :-X  do you use low level call in you code to use only SSD1298? maybe it's enough add specific library if the functions are the same for both controller
Can i try without PSU arduino shield?
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #331 on: May 25, 2016, 06:46:04 pm »
Damn! i have same tft screen  :-X  do you use low level call in you code to use only SSD1298? maybe it's enough add specific library if the functions are the same for both controller

Take it easy, possibility to work with more models gives us more flexibility. That "only" requires more time to do everything right. This TFT is not directly supported by UTFT that is some kind of standard library for ebay/alibaba models but we already had the similar situation with UTouch from the same author. As we successfully moved from it maybe it's time to do the same thing with the TFT part.

Can i try without PSU arduino shield?

In what sense? With the latest firmware you can run Arduino shield even without display. So this issue with "wrong" display is not a show stopper if you are going to start PCB assembly right away.
 
The following users thanked this post: Aigor

Offline Aigor

  • Contributor
  • Posts: 29
  • Country: it
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #332 on: May 25, 2016, 07:55:37 pm »

In what sense? With the latest firmware you can run Arduino shield even without display. So this issue with "wrong" display is not a show stopper if you are going to start PCB assembly right away.
I mean try code on arduino with TFT connected but without your PSU arduino shield
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #333 on: May 25, 2016, 09:51:39 pm »
That will require a lots of wiring but you can try. You have to follow schematics on sheet 2 and 3 for the Arduino shield (JP3 to JP5 is assigned to TOUCH_SCLK, TOUCH_DIN and TOUCH_DOUT).

Instead of loading firmware you can also use some of the UTouch examples (e.g. UTFT_Demo_320x240). Please take in account that TFT required 5 V for supply but is 3.3 V logic level device!

Offline Aigor

  • Contributor
  • Posts: 29
  • Country: it
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #334 on: May 25, 2016, 10:06:55 pm »
Thank you! I will try in the weekend

Inviato dal mio NX507J utilizzando Tapatalk

 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Preliminary AUX PS PCB correction ...
« Reply #335 on: May 26, 2016, 09:46:29 am »
Phew, yesterday when testing a combination of new the Arduino shield and power board started I realize that triac triggering on AUX PS does not work. The problem is with wiring that is made in this revision. Fortunately that issue was not so problematic to fix. One PCB cut, two holes for new wiring and it's now works.





 
The following users thanked this post: Aigor

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #336 on: May 26, 2016, 10:14:33 am »
Thank you! I will try in the weekend

Inviato dal mio NX507J utilizzando Tapatalk

Regarding ILI9341 library it seems there is many solutions available. Only on github search returns 109 results. If anyone already have something to recommend even only for testing please let me know.

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
TFT display with ILI9341 ...
« Reply #337 on: May 26, 2016, 01:17:12 pm »
He, he, TFT display with different controller wasn't be so great issue. There is a hack in UTFT library that is described on Arduino forum. One line in UTFT.cpp has to be changed from:

Code: [Select]
word dsy[] = {319, 399, 319, 319, 319, 319, 219, 219, 399, 159, 127, 319, 479, 799, 319, 319, 319, 0, 0, 319, 799, 479, 319, 219, 159, 319, 319, 479, 479, 479, 159, 159};
to ...

Code: [Select]
word   dsy[] = {319, 319, 319, 319, 319, 319, 219, 219, 399, 159, 127, 319, 479, 799, 319, 319, 319, 0, 0, 319, 799, 479, 319, 219, 159, 319, 319, 479, 479, 479, 159, 159};
Now we have to add into firmware selection between SSD1289 and ILI9341 controller. In the meantime you can test sketch that is in the attachment.

 
The following users thanked this post: Aigor

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Strange behavior of reset IC...
« Reply #338 on: June 04, 2016, 09:06:17 am »
A lots of things are still happening in parallel and here I'd like to present where I spent many hours by "catching a ghost". Maybe someone more experienced would spot that at once but I still do know why it's happening but fortunately I find a simple (dirty?) solution for that.
This issue is introduced with adding TPS3705-33 power supervisor and reset IC. It offers power-on (and power-failure) reset and also watchdog functionality that can be used as extreme measure if firmware for some reason blocked.



Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Testing "low ripple" mode of operation
« Reply #339 on: June 04, 2016, 04:32:33 pm »
Since the Low ripple functionality is now also implemented on the firmware side we can test it on the real equipment. Idea about bypassing switching pre-regulator comes from void_error (thank again for that). When "Low ripple" mode of operation is selected the power pre-regulator is bypassed by setting so-called Duty cycle of the SMPS controller to 100 %. That will disable switching frequency and therefore that otherwise hard to filter component disappear entirely from the channel output. In that case the remaining noise  should mainly come from the low power bias switching pre-regulator (LM5574) and the Arduino Due board.

Max. output current for pre-regulator mosfet (SUD19P06) is limited in firmware to 0.8 A. That value comes from its SOA chart:



Maybe I should add some reserve and lower it to 10-20 %. Your input here is welcome. 50 V point is chosen since that is DC voltage at the power board input (LRS-150-48 module). As max. power dissipation for post-regulator's mosfet mounted on heatsink a 20 W is chosen. In real life I'm going to put that heatsink near 60 mm fan that should easily handle 40 W for two channels.



This time as a measurement point a 4mm to BNC adapter is used and no additional capacitors are connected on the output (only output capacitor is 15u in parallel with 470n+1R). Let's starts with grounded probe input:



When power is switched off:



OUTPut OFF means that SMPS pre-regulator is completely shut down and post-regulator's mosfet bias is switched off (Arduino Due is off course active):



Connected load is 16R4 power resistor that with output voltage set to 7 V consume about 0.43 A. That is within limits for low ripple mode of operation, but first how it looks when LRIPple if OFF:



and finally when LRIPple is ON:



It works! :)
While in this mode of operation and set parameters note that dissipation on post-regulator mosfet is about 18.5 W! In the same time pre-regulator mosfet is cold since with Rds(on) of 60 m? is dissipate 25 mW.

If you compare measured noise with LRIPple OFF (SMPS is turned on) with that from e.g. post #310 you can see that this one are much more dirty. That is possibly due to the different method of signal probing. I also found that grounding or lack of it can influence greatly appearance on the scope of that noise. The Arduino shield board has 10n capacitors connected from power output terminals to PE (protective earth line), adding the same combination at the input of the pre-regulator (after common-mode choke) can also make a change at least when AC/DC adapter module as in this example is used.

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Finalizing revision 5 PCB layout
« Reply #340 on: June 07, 2016, 06:51:05 am »
We are approaching slowly finalization of the Revision 5 PCBs. All new functionality that is added on the preliminary PCBs that Electricar sent to me seems to work. During testing phase some new ideas came and I'd like to include it in the final revision. Before that I'd like to hear your opinion about few things that will affect final PCB layout.

1) Power boards has to be placed closer to each other as announced in post #284. A heatsink for post-regulator mosfet is mounted on that board and PCB orientation that is suggested in mentioned looks like this:



... but it is not optimal since pre-regulator power mosfet and diode are SMD components and air-flow over them is marginal. I was planned to add a "thermal bridge" below PCB area where mentioned components are located but I'm not sure if that is better then rotating PCB that its surface is in "air tunnel" like on the diagram below. Such orientation decrease amount of air that reach heatsink but it's not so critical since max. dissipation cannot exceed 20 W per channel. This is important to know for proper location of 26-pin connectors on the Arduino shield.



2) Fan blowing orientation: what is better that air is sucked out of enclosure or vice-versa?

3) Enclosure holes: please take a look once again on the proposed enclosure in post #284. The power boards will be located on the right side (looking from the front) and power transformer or two AC/DC modules on the left side. What do you think is there too many ventilation holes on top and bottom plates? How is that related to topic 2)?

4) I'd like to reduce the number of connecting cables to the minimum. On the power board was connector for NTC that should be somehow mounted or glued with silicon onto heatsink. But, the more critical area is the surface of the PCB near pre-regulator power mosfet (Q1). I'm thinking about using a SMD NTC mounted close to that area as on the picture below. Does it makes sense?



5) I'd like to offer flexible power board input that people could choose between AC (main transformer) and DC (AC/DC module) input as shown on the following picture:



That section could reside on the power board but also it can be moved to the AUX PS that is planned to be mounted on the enclosure rear panel. In that case instead of four SMD rectifiers diodes we can use cheaper THT bridge that can be fixed to the rear panel for cooling. Also with such constellation bulk capacitors (that are mounted vertically to the PCB) will not dictates distance between power boards and in that way a total width of the power boards section that possibly can also decrease needed width of the enclosure.

Thanks everyone in advance for your valuable inputs.

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Heatsink for power board ...
« Reply #341 on: June 07, 2016, 07:08:03 am »
6) I'd like to use easy to obtain extruded profile heatsink for power board and have a difficulties with width and length when fins orientation is as shown in topic 1) in previous post. Many of them comes with 100 mm width but I need 70 to 76 mm. Also if width is right then they are too short: usually 100 mm, but I'd like to have one from 120 to up to 150 mm. The closer candidate (that is also very affordable) is RAD-A4463/120.

Offline Electronix

  • Contributor
  • Posts: 39
  • Country: de
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #342 on: June 07, 2016, 05:59:41 pm »
Hi Prasimix,

I ask me, if it is not possible to directly regulate the output voltage of your switching supply if you make some changes on the board of that power supply ?
Normally such kind of ps have a pot where you can adjust the output voltage between a upper and lower limit. I can imagine that these range is only controlled by a few resistors. So if you change the voltage from the pot over a op like I do on the cosel supply's it should also work for you. This will make the pre regulator obsolete  :D

Cheers

Sven
 

Offline Electronix

  • Contributor
  • Posts: 39
  • Country: de
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #343 on: June 07, 2016, 07:23:27 pm »
Hi Prasimix,

I have another question. I have made on my prototype some measurements and find out when I set voltage to 6.000 volts and add a 2.2 ohm power resistor on the output that than the output voltage goes a little bit lower to 5.994 volts. Have you also made same observations on your design ?
Maybe on my prototype it is a grounding problem. I don't know I'm searching for the problem now.

Cheers

Sven
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #344 on: June 08, 2016, 05:18:35 am »
Hi Prasimix,

I ask me, if it is not possible to directly regulate the output voltage of your switching supply if you make some changes on the board of that power supply ?
Normally such kind of ps have a pot where you can adjust the output voltage between a upper and lower limit. I can imagine that these range is only controlled by a few resistors. So if you change the voltage from the pot over a op like I do on the cosel supply's it should also work for you. This will make the pre regulator obsolete  :D

Cheers

Sven

Hi Sven, not so sure because the price of single Cosel unit is comparable with completed dual channel PSU. I'm talking about suppliers such as Mouser (Farnell nor TME cannot offer it) and I think that BOM should not rely on used/discontinued units found on eBay auctions. It seems that Cosel cannot offer a lot for DIY-class of project or maybe I missed something?

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #345 on: June 08, 2016, 05:23:53 am »
Hi Prasimix,

I have another question. I have made on my prototype some measurements and find out when I set voltage to 6.000 volts and add a 2.2 ohm power resistor on the output that than the output voltage goes a little bit lower to 5.994 volts. Have you also made same observations on your design ?
Maybe on my prototype it is a grounding problem. I don't know I'm searching for the problem now.

Cheers

Sven

Well, we are talking about 2.7 A and 6 mV voltage drop. I can bet that if you try that on multi-thousand dollars/euros professional unit that you can get the same result. For that you have to use remote sense inputs and connect it wisely (even wrong sequence of connecting power, sense and load terminal could results in 1-2 mV drop).

Offline Electronix

  • Contributor
  • Posts: 39
  • Country: de
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #346 on: June 08, 2016, 04:07:30 pm »
"Hi Sven, not so sure because the price of single Cosel unit is comparable with completed dual channel PSU. I'm talking about suppliers such as Mouser (Farnell nor TME cannot offer it) and I think that BOM should not rely on used/discontinued units found on eBay auctions. It seems that Cosel cannot offer a lot for DIY-class of project or maybe I missed something?"

Hi prasimix,

maybe we misunderstood. I mean not that you use a cosel ps. I mean to modify the meanwell ps that you have to be able to adjust the output voltage.

Regarding the 6 milivolt it will be fine if someone can make a test which have a good power supply so that we can compare our self made ps with a professional power supply.

Cheers

Sven
 

Offline prasimixTopic starter

  • Supporter
  • ****
  • Posts: 2022
  • Country: hr
    • EEZ
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #347 on: June 08, 2016, 04:20:38 pm »
Hi prasimix,

maybe we misunderstood. I mean not that you use a cosel ps. I mean to modify the meanwell ps that you have to be able to adjust the output voltage.
Ok, now I understand. Mean well has a potentiometar that allows you to set output voltage I think between 44 and 54 V. Not so sure if that also mean that you can go down to e.g. 4-5 V. Maybe someone have such experience.

Regarding the 6 milivolt it will be fine if someone can make a test which have a good power supply so that we can compare our self made ps with a professional power supply.

Cheers

Sven

I still think it's not possible to achieve that without remote sense even with much better power supply. It's simply an Ohm's law in action, connection cable and connection points adds some resistance that are out of reach of CV error amp and 2.7 A is not small current. It's a question of "only" 2.2 miliohms.

Offline electricar

  • Regular Contributor
  • *
  • Posts: 89
  • Country: ch
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #348 on: June 08, 2016, 04:50:06 pm »
I still think it's not possible to achieve that without remote sense even with much better power supply. It's simply an Ohm's law in action, connection cable and connection points adds some resistance that are out of reach of CV error amp and 2.7 A is not small current. It's a question of "only" 2.2 miliohms.

Today I made some measurements with our Keysight E3634A but unfortunately I forgot the spreadsheet at work. But I can say that the drop on the Keysight was worse, even though I measured directly on the power supply output, not at the load. I my opinion the E3634A is a great power supply and with the comparisons I already made, I have to say that I'm impressed every time of prasimix power supply.
I will post the exact values tomorrow.
 

Offline Electronix

  • Contributor
  • Posts: 39
  • Country: de
Re: DIY SCPI programmable dual channel bench PSU 0-50V/3A
« Reply #349 on: June 09, 2016, 04:35:17 am »
Hi prasimix,

maybe we misunderstood. I mean not that you use a cosel ps. I mean to modify the meanwell ps that you have to be able to adjust the output voltage.
Ok, now I understand. Mean well has a potentiometar that allows you to set output voltage I think between 44 and 54 V. Not so sure if that also mean that you can go down to e.g. 4-5 V. Maybe someone have such experience.

Regarding the 6 milivolt it will be fine if someone can make a test which have a good power supply so that we can compare our self made ps with a professional power supply.

Cheers

Sven

I still think it's not possible to achieve that without remote sense even with much better power supply. It's simply an Ohm's law in action, connection cable and connection points adds some resistance that are out of reach of CV error amp and 2.7 A is not small current. It's a question of "only" 2.2 miliohms.


But I'm using already remote sense. I find out that when I set 6.200 volts I got 6200,5 mV on remote sense op and under load also I have 6,2005 mV on the ic. But I have under load only 6,154 volts on the dummy load which I have connected. Strange.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf