Author Topic: Linear lab bench power supply - Advise Needed  (Read 2652 times)

0 Members and 1 Guest are viewing this topic.

Offline UltraRCTopic starter

  • Contributor
  • Posts: 18
  • Country: nz
Linear lab bench power supply - Advise Needed
« on: January 17, 2020, 10:07:59 pm »
A week ago I posted on reddit asking for some advice but i figured here would be a better place.  I think i'm getting close to putting this thing together but I still have a few questions/problems. I am aware that there are likely many problems with the way this is drawn/designed and that when I posted on reddit, people suggested that I simply buy a pre-made regulator unit off of eBay or something, but that seems dull and I want to learn how to build it my self.

What I have drawn so far is at the bottom of the post.

The circuit is based around an IRFZ44N mosfet used as a series pass device and regulated by an opamp. The current control is essentially a copy paste from the eev blog PSU design series although a few of the resister values will need changing as i'm using two 1R resistors in parrallel as the current sense so its 0.5R. When the current gets over a threshold set by R9 the mosfet gate is pulled low by Q2, BC547. I'll probably also add a resistor in series with R9 to limit the max current to some current value which will depend on how hot the heat sink gets while testing (probably 1A).

While I was doing some bread board prototyping the opamp/mosfet voltage regulation combo did a good job of regulating the output and closely followed the voltage of the pot, but stating at max voltage and slowly lowering the voltage with the pot, when the output approaches 0V it quickly jumps up to near max output voltage even though the pot is set to 0V. I think it has something to do with how the opamp is configured but i'm unsure whats causing this.

I am planning to use an arduino board to measure and display the current set, current sense and voltage sense values on an LCD but otherwise not involved in the control circuitry. Can I just use a few of the analog pins to measure 0-15V? Or is that too high of a voltage for the chip to deal with?

I will add current limit and power on indicator LEDS to the design later but other than that  was just looking for more experienced eyes to point out any issues or suggestions for the design before I start soldering it together.

Thanks, Reuben
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Linear lab bench power supply - Advise Needed
« Reply #1 on: January 17, 2020, 11:12:25 pm »
What type of op-amp are you using? It needs to have an input common mode voltage range that includes its negative supply voltage(pin 4).
Some common types are LM358 and LM324.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline UltraRCTopic starter

  • Contributor
  • Posts: 18
  • Country: nz
Re: Linear lab bench power supply - Advise Needed
« Reply #2 on: January 17, 2020, 11:25:07 pm »
Yep its an LM324N
 

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9964
  • Country: us
Re: Linear lab bench power supply - Advise Needed
« Reply #3 on: January 17, 2020, 11:42:45 pm »
There's an 'oopsie' known as inversion under certain input conditions.

http://www.ti.com/lit/an/sloa277/sloa277.pdf

Section 2.2 starts to talk about this.
« Last Edit: January 18, 2020, 12:15:08 am by rstofer »
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Linear lab bench power supply - Advise Needed
« Reply #4 on: January 17, 2020, 11:44:53 pm »
I can't see why it should have that problem. I can see some unrelated problems. R8 and all op-amps should be powered directly by the 15V.
The CV feedback loop is likely to be unstable and oscillate.
Take some voltage readings on the op-amp.
An oscilloscope will often spot problem causes that a DMM can miss.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline UltraRCTopic starter

  • Contributor
  • Posts: 18
  • Country: nz
Re: Linear lab bench power supply - Advise Needed
« Reply #5 on: January 17, 2020, 11:59:46 pm »
I can't see why it should have that problem. I can see some unrelated problems. R8 and all op-amps should be powered directly by the 15V.
The CV feedback loop is likely to be unstable and oscillate.
Take some voltage readings on the op-amp.
An oscilloscope will often spot problem causes that a DMM can miss.

When you say "R8 and all op-amps should be powered directly by the 15V", do you mean that they should be connected to the 15V before the current sense resistor?

I'll build up the voltage control circuit and take some measurements
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Linear lab bench power supply - Advise Needed
« Reply #6 on: January 18, 2020, 12:08:06 am »
I can't see why it should have that problem. I can see some unrelated problems. R8 and all op-amps should be powered directly by the 15V.
The CV feedback loop is likely to be unstable and oscillate.
Take some voltage readings on the op-amp.
An oscilloscope will often spot problem causes that a DMM can miss.

When you say "R8 and all op-amps should be powered directly by the 15V", do you mean that they should be connected to the 15V before the current sense resistor?

I'll build up the voltage control circuit and take some measurements
Yes, the supply voltage will be more stable.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline UltraRCTopic starter

  • Contributor
  • Posts: 18
  • Country: nz
Re: Linear lab bench power supply - Advise Needed
« Reply #7 on: January 18, 2020, 01:12:03 am »
Thank You very much, progress is being made. Because I needed more opamps to for the circuit i'm now using an LM324 quad opamp instead of the dual LM833 I was using previously. Using the LM324 seems to have eliminated the issue of the output jumping up when the pot is set to 0V so thats solved. I read part of that PDF and didn't really understand what I was reading, but that's fine because the problem is gone now.

I have tested the voltage control section of the circuit on breadboard and tested with a multimeter and oscilloscope. The only oscillation I noticed was when it was set to zero volts with a small DC motor across the output. 

The next thing that I am considering is that many people on the internet seem to hold the opinion that the use of a mosfet as a series pass device is inferior to a BJT. If this is the case I found a TIP31C in my parts bin, would this part be better than the IRFZ44N for this application?

Thanks, Reuben
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Linear lab bench power supply - Advise Needed
« Reply #8 on: January 18, 2020, 02:45:51 am »
Thank You very much, progress is being made. Because I needed more opamps to for the circuit i'm now using an LM324 quad opamp instead of the dual LM833 I was using previously. Using the LM324 seems to have eliminated the issue of the output jumping up when the pot is set to 0V so thats solved. I read part of that PDF and didn't really understand what I was reading, but that's fine because the problem is gone now.

I have tested the voltage control section of the circuit on breadboard and tested with a multimeter and oscilloscope. The only oscillation I noticed was when it was set to zero volts with a small DC motor across the output. 

The next thing that I am considering is that many people on the internet seem to hold the opinion that the use of a mosfet as a series pass device is inferior to a BJT. If this is the case I found a TIP31C in my parts bin, would this part be better than the IRFZ44N for this application?

Thanks, Reuben
MOSFETs that are optimised for switching are susceptible to Secondary Breakdown(google it). I wouldn't be worried about the IRFZ44N in your application.   
When checking for stability problems, look at the op-amp's output also.

https://electronics.stackexchange.com/questions/361407/dc-operation-for-irfz44-mosfet-soa-curve-max-current-at-40v
« Last Edit: January 18, 2020, 02:48:59 am by xavier60 »
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline not1xor1

  • Frequent Contributor
  • **
  • Posts: 716
  • Country: it
Re: Linear lab bench power supply - Advise Needed
« Reply #9 on: January 18, 2020, 05:38:47 pm »
there are various problems with that design:

- there is no need for a pre-regulator (L7815cp) as opamp plus power device are enough to provide both current and voltage regulation

- the filter capacitor (C1) may be too small while the output capacitor is too large (C4)
  BTW you should provide some information about the transformer: (no load output voltage and rated VA) and required feature (min-max V and I) to get helpful advice

- a switching mosfet is unsuitable as pass device both because it is not specified for linear operation and because, as source follower, it would require a gate voltage around 4-5V higher than the maximum output voltage.
   Besides that LM324 is not a rail-to-rail output opamp, so with 15V input voltage (opamp supply and drain voltage) at 1A load and taking into account the drop through R2||R4 you would hardly get more than 7-8V of maximum output voltage.

You should take care of opamp max supply voltage, min-max input range, min-max output capability at various load, proper transformer derating, power device SOA, feedback compensation, etc...
TIP31C might be ok, but you should drive it through another BJT like BD135 or similar. IMHO you should take inspiration from this Jaycee's design (read the whole thread).
 

Online mikerj

  • Super Contributor
  • ***
  • Posts: 3383
  • Country: gb
Re: Linear lab bench power supply - Advise Needed
« Reply #10 on: January 18, 2020, 07:58:10 pm »
there are various problems with that design:

- there is no need for a pre-regulator (L7815cp) as opamp plus power device are enough to provide both current and voltage regulation

The regulator is effectively providing the reference voltage.
 

Offline not1xor1

  • Frequent Contributor
  • **
  • Posts: 716
  • Country: it
Re: Linear lab bench power supply - Advise Needed
« Reply #11 on: January 19, 2020, 05:45:19 pm »
there are various problems with that design:

- there is no need for a pre-regulator (L7815cp) as opamp plus power device are enough to provide both current and voltage regulation

The regulator is effectively providing the reference voltage.

that 7815 in the circuit of the original poster works as a preregulator and while it is also used for the reference voltage of the current and voltage regulation circuits, there are plenty of better, smaller and cheaper alternatives than a 1.5A series regulator as a voltage reference.
 

Online mikerj

  • Super Contributor
  • ***
  • Posts: 3383
  • Country: gb
Re: Linear lab bench power supply - Advise Needed
« Reply #12 on: January 19, 2020, 05:51:42 pm »
there are various problems with that design:

- there is no need for a pre-regulator (L7815cp) as opamp plus power device are enough to provide both current and voltage regulation

The regulator is effectively providing the reference voltage.

that 7815 in the circuit of the original poster works as a preregulator and while it is also used for the reference voltage of the current and voltage regulation circuits, there are plenty of better, smaller and cheaper alternatives than a 1.5A series regulator as a voltage reference.

I fully agree, but your first post implies this device could be removed without considering the implications.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17429
  • Country: us
  • DavidH
Re: Linear lab bench power supply - Advise Needed
« Reply #13 on: January 20, 2020, 04:29:01 am »
At only 15 volts, secondary breakdown of the power MOSFET will not be an issue.

There are a couple of problems:

1. The high output capacitance impedes good current limiting response.  No more than about 22 to 100 microfarads per amp should be necessary.  High performance designs with the best current limiting get by with less than 0.1 microfarads per amp.

2. C5 severely limits bandwidth and should not be necessary.

3. Q2 adds voltage gain within the current control loop and this gain varies with operating point making frequency compensation of the current control loop difficult; replace it with a diode or PNP emitter follower.  If you absolutely want to keep it, then add a resistor in series with the emitter to control and limit transconductance.

4. It is not impossible to make a current control loop work with two cascaded operational amplifiers but there are better ways to do it with only one stage.  Or make the instrumentation amplifier faster.
 

Offline not1xor1

  • Frequent Contributor
  • **
  • Posts: 716
  • Country: it
Re: Linear lab bench power supply - Advise Needed
« Reply #14 on: January 20, 2020, 06:09:21 pm »
I fully agree, but your first post implies this device could be removed without considering the implications.

I've been misunderstood   ;D
as I suggested to trash the whole circuit (and to adopt Jaycee's design instead) I hardly see any implication by 7805 removal  :)
 

Offline UltraRCTopic starter

  • Contributor
  • Posts: 18
  • Country: nz
Re: Linear lab bench power supply - Advise Needed
« Reply #15 on: January 21, 2020, 07:55:35 am »
Thanks again for the help

I have done a bit of reading, thinking and tinkering and have made the following changes (Sorry but the component prefixes have changed in the new schematic because I pressed a wrong button):

- Added current limiting indicator LED
- Added power on indicator LED
- Opamps, voltage set and current set pots now powered from before the current shunt
- Changed output cap from 1000uF to 100uF
- Added 10K resister in series with the emmitter of Q2

Questions:
- Will Q3 interfere with the loop?
- I read that I may need a negative supply for the opamp, is this needed?
- Also thinking that the inputs of U2 may need switching?

Thanks, am excited to get this thing working
« Last Edit: January 21, 2020, 08:50:42 am by UltraRC »
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Linear lab bench power supply - Advise Needed
« Reply #16 on: January 21, 2020, 09:25:01 am »
Thanks again for the help

I have done a bit of reading, thinking and tinkering and have made the following changes (Sorry but the component prefixes have changed in the new schematic because I pressed a wrong button):

- Added current limiting indicator LED
- Added power on indicator LED
- Opamps, voltage set and current set pots now powered from before the current shunt
- Changed output cap from 1000uF to 100uF
- Added 10K resister in series with the emmitter of Q2

Questions:
- Will Q3 interfere with the loop?
- I read that I may need a negative supply for the opamp, is this needed?
- Also thinking that the inputs of U2 may need switching?

Thanks, am excited to get this thing working
Q2 is over driven, 10K for R15 is too high anyway to pull the Gate voltage down.
R11 needs to be 10K and about 100Ω for R15. R15 needs to be just low enough so that the Gate can be pulled below the MOSFETs threshold voltage.
The CC loop will still likely be unstable. Adding compensation feedback to U2 might help.
Preferably, U2 should be configured as inverting with series RC feedback and be able to pull down the Gate voltage without adding extra gain.
So far I have not been able to figure out a way that will pull the Gate voltage low enough without causing other complications.

Might need to consider an alternate output configuration that is easier to control and has a much lower dropout voltage so that  over 14V max output will be possible.
 I have good success with the PNP Sziklai Transistor Pair type output,  https://circuitdigest.com/electronic-circuits/sziklai-transistor-pair

Extra: A negative rail can make things easier but not essential with careful design.
« Last Edit: January 21, 2020, 09:32:14 am by xavier60 »
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf