Author Topic: Could you review regulator stage of bench PSU design  (Read 934 times)

0 Members and 1 Guest are viewing this topic.

Offline virinomTopic starter

  • Newbie
  • Posts: 1
  • Country: tr
Could you review regulator stage of bench PSU design
« on: August 24, 2020, 09:14:01 pm »
Hi all,
I am designing 0-50V 0-10A CV, CC digital bench power supply. By saying digital I am referring to controlling output transistors with DAC from MCU and setting DAC voltage by reading output values with MCU using ADC.
I started with power stage. In Multisim simulation, I used a potentiometer instead of DAC and I don't have ADC stage. I'm planning to use MJ15003 in circuit but I used MJ15024G for simulation. Could you please look at my power stage and inform me If there is any major design error or If you have any recommendation.
I am putting Multisim link below, you can also find schemetic in attachments.
https://www.multisim.com/content/wRrPWn9FGKN44TeTAtQD3h/power-supply/
Thank you

« Last Edit: August 24, 2020, 09:16:18 pm by virinom »
 

Offline kallek

  • Regular Contributor
  • *
  • Posts: 59
  • Country: fi
Re: Could you review regulator stage of bench PSU design
« Reply #1 on: August 26, 2020, 07:13:12 pm »
Hi, and welcome!
You need analog feedback from output. Digital feedback is too slow. I also suggest to start with lower voltage and current. Designing that kind of linear power supply is big challenge to everyone.
 

Offline Doctorandus_P

  • Super Contributor
  • ***
  • Posts: 4008
  • Country: nl
Re: Could you review regulator stage of bench PSU design
« Reply #2 on: August 28, 2020, 09:45:32 pm »
50V 10A is pretty ambitious.
A lineair supply would need to be able to dissipate 500W in worse case condition.
That is like a big CPU cooler for each of your 4 power transistors.

Do you have a good reason for this power requirement?

In my experience, I can do most projects with 30V 500mA power supply, and when more power is needed it often does not have to be well regulated.

You are also likely to want to have multiple power supplies. So even if you first design / build a small one, then you will still use it after you built the big monster.

I also agree with kallek. Use a uC and dac to make a setpoint, and use analog circuitry to regulate the output voltage.
 

Offline dietert1

  • Super Contributor
  • ***
  • Posts: 2473
  • Country: br
    • CADT Homepage
Re: Could you review regulator stage of bench PSU design
« Reply #3 on: August 28, 2020, 10:14:37 pm »
There is analog feedback. The U1 gain stage has G = 6, so at 2 V input it outputs 12 V. The Q3/Q2 gain stage has G = 17, so that gives 204 V instead of 50 V. And something is wrong with R7.

Regards, Dieter
 

Online wraper

  • Supporter
  • ****
  • Posts: 18016
  • Country: lv
Re: Could you review regulator stage of bench PSU design
« Reply #4 on: August 28, 2020, 10:24:42 pm »
There is analog feedback. The U1 gain stage has G = 6, so at 2 V input it outputs 12 V. The Q3/Q2 gain stage has G = 17, so that gives 204 V instead of 50 V. And something is wrong with R7.

Regards, Dieter
It is not a feedback he talks about. Digital feedback is not impossible as such. However not with a regular MCU, nor it would be trivial. Very complicated stuff to implement. Here is simplified linear PSU where you can see the feedback.

« Last Edit: August 28, 2020, 10:28:17 pm by wraper »
 

Offline dietert1

  • Super Contributor
  • ***
  • Posts: 2473
  • Country: br
    • CADT Homepage
Re: Could you review regulator stage of bench PSU design
« Reply #5 on: August 29, 2020, 03:25:02 am »
The feedback signal for R6 should be taken from the output of the emitter followers. And R6 should be 3.1666 K.

Regards, Dieter
« Last Edit: August 29, 2020, 03:33:47 am by dietert1 »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf