Author Topic: Simple DIY Linear PSU  (Read 17889 times)

0 Members and 1 Guest are viewing this topic.

Offline panossTopic starter

  • Frequent Contributor
  • **
  • Posts: 327
  • Country: gr
Re: Simple DIY Linear PSU
« Reply #50 on: March 07, 2020, 02:24:45 pm »
I like to place larger areas of copper print under power resistors.`
Are you referring to SMD resistors?
I 'm asking because through hole resistors are on the top side of the PCB, while the copper on the bottom.
« Last Edit: March 07, 2020, 02:27:39 pm by panoss »
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Simple DIY Linear PSU
« Reply #51 on: March 07, 2020, 02:32:52 pm »
I like to place larger areas of copper print under power resistors.`
Are you referring to SMD resistors?
No, the axial resistors for R13, R14 and R18. As well as keeping resistance low, the extra wide print helps take the heat away.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Simple DIY Linear PSU
« Reply #52 on: March 08, 2020, 12:46:45 am »
With the early version PCBs  for my bench supply I was inputting unreg on one end and outputting on the opposite end of the PCB.
With later versions I put the input and output close together at one end of the PCB to avoid having long tracks carrying high current across the PCB.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Simple DIY Linear PSU
« Reply #53 on: March 08, 2020, 07:28:04 am »
About R15. I picked its value out of the air. Some consideration needs to be given to its possible high dissipation.
1 watt would be a safe choice and try to keep it away from electrolytic capacitors to avoid heat degradation.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Simple DIY Linear PSU
« Reply #54 on: March 09, 2020, 08:50:27 am »
My bench supply's PCB illustrates my layout suggestions.
C6 does not strictly need to be 1000uF.
« Last Edit: March 09, 2020, 10:25:33 am by xavier60 »
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 
The following users thanked this post: panoss

Offline panossTopic starter

  • Frequent Contributor
  • **
  • Posts: 327
  • Country: gr
Re: Simple DIY Linear PSU
« Reply #55 on: March 10, 2020, 05:54:02 pm »
I 'm in the process of building the circuit and I thought of trying something in the simulation.
BUT I found out that I had made a mistake:

On Reply #32 I had written that I made the simulation with D45H11.
Well...that 's not true...I had not connected the D45H11 model to the transistor. :palm:
So the simulator used a 'Generic PNP transistor' and not a D45H11...

So, I connected the model to the transistor and run it, but it wouldn't run because of errors.
So, for Q2 I thought of using a Darlington transistor (BDX54), because of it 's high gain.

I changed the value of R10 (from 300 Ohms) back to it's original value, 1k.
And looks like working fine!
« Last Edit: March 10, 2020, 06:35:37 pm by panoss »
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Simple DIY Linear PSU
« Reply #56 on: March 11, 2020, 08:17:44 am »
If you are adding one of those cheap Volt Amp panel meters to the project, it should not be powered directly from the 9V rail. A shunt regulator should be added to prevent the LEDs from causing current fluctuations through the CS resistor.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline panossTopic starter

  • Frequent Contributor
  • **
  • Posts: 327
  • Country: gr
Re: Simple DIY Linear PSU
« Reply #57 on: March 14, 2020, 09:49:32 am »
Ok, I built it!
But, I 've made some modifications because these are the parts I had in hand.
I avoided getting out of the house (corona virus, you know).

So, I replaced Q2 with a DIY darlington pair, with BC557 & S8550.
Q3=Q4=D209L.
D5=BAT41 (instead of BAT85)

I did some tests, with no current limit:
RL= no load, Vout=16.79V
RL= 20 Ohms, Vout=14.93V, Iout=0.71A
RL= 10 Ohms, Vout=13.59V, Iout=1.26A
« Last Edit: March 14, 2020, 11:52:43 am by panoss »
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Simple DIY Linear PSU
« Reply #58 on: March 14, 2020, 10:02:31 am »
Ok, I built it!
But, I 've made some modifications because these are the parts I had in hand.
I avoided getting out of the house (corona virus, you know).

So, I replaced Q2 with a DIY darlington pair, with BC557 & S8550.
Q3=Q4=D209L.
D5=BAT41 (instead of BAT85)

I did some tests, with no current limit:
RL= no load, Vout=16.79V
RL= 20 Ohms, Vout=14.93V, Iout=0.71A
RL= no load, Vout=13.59V, Iout=1.26A
Both of those PNP transistors have fairly high gain.
The CV loop is likely oscillating, can you check?
Use just one PNP transistor.

Extra; The D209L has very low gain so a lot of load will be placed on the undersized PNP driver. So don't power anything valuable with it yet.
« Last Edit: March 14, 2020, 10:10:42 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 xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Simple DIY Linear PSU
« Reply #59 on: March 14, 2020, 10:14:16 am »
Do you have any scrapped audio power amplifiers? They might have suitable transistors.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline panossTopic starter

  • Frequent Contributor
  • **
  • Posts: 327
  • Country: gr
Re: Simple DIY Linear PSU
« Reply #60 on: March 14, 2020, 11:42:45 am »
I have no oscilloscope so I can't check for oscillation.

I made a list of the Power NPN transistors I have (these without any number of pieces are 1 piece):
E13009L  2PCS
E13009   2PCS  (they are a lot smaller than E13009L)
BU426V   
BU508A
ON4046 (PHILIPS)
BU508DF
C5027-R  2PCS
TT2146   2PCS
D13007K  2PCS
BD241C 5PCS
C5763
D1408  2PCS
BD139


CORRECTION in my previous post:
I did some tests, with no current limit:
RL= no load, Vout=16.79V
RL= 20 Ohms, Vout=14.93V, Iout=0.71A
RL= 10 Ohms, Vout=13.59V, Iout=1.26A
« Last Edit: March 14, 2020, 11:59:20 am by panoss »
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Simple DIY Linear PSU
« Reply #61 on: March 14, 2020, 12:06:36 pm »
Try the two D1408 with a BC557.
Also measure the voltage drop across R10 for each load.

Extra: If the BC557 fails, try the S8550 although it has a low voltage rating and is likely to fail also.
« Last Edit: March 14, 2020, 12:10:04 pm by xavier60 »
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline panossTopic starter

  • Frequent Contributor
  • **
  • Posts: 327
  • Country: gr
Re: Simple DIY Linear PSU
« Reply #62 on: March 14, 2020, 12:14:30 pm »
I have already removed the BC557 (as it was more convenient) so I will firstly try with the S8550.
 

Offline panossTopic starter

  • Frequent Contributor
  • **
  • Posts: 327
  • Country: gr
Re: Simple DIY Linear PSU
« Reply #63 on: March 14, 2020, 12:58:03 pm »
Unfortunately one of the pins of one of the D1408s was half-broken and I broke it totally...
Maybe I  can use one D1408 with some other transistor?
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Simple DIY Linear PSU
« Reply #64 on: March 14, 2020, 01:02:52 pm »
Unfortunately one of the pins of one of the D1408s was half-broken and I broke it totally...
Maybe I  can use one D1408 with some other transistor?
Use one for now and don't go much more than 1A.
I notice that the voltage has been dropping by about 2.5V per amp of load. There could be a wiring mistake.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 
The following users thanked this post: panoss

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 15157
  • Country: de
Re: Simple DIY Linear PSU
« Reply #65 on: March 14, 2020, 01:12:00 pm »
Combining different power transistors in parallel is not that great and would need more emitter resistance.


The BD241C should be OK, at least for lower power (about 20 W per transistor).

Less current and only 1 transistor should be OK too.
 
The following users thanked this post: panoss

Offline panossTopic starter

  • Frequent Contributor
  • **
  • Posts: 327
  • Country: gr
Re: Simple DIY Linear PSU
« Reply #66 on: March 14, 2020, 01:14:50 pm »
I notice that the voltage has been dropping by about 2.5V per amp of load. There could be a wiring mistake.
I 've split it in two PCBs, in the attachment you will find my Eagle (7.7.0) files:
1. main pcb.
2. power section, contains the power transistors' section
 

Offline panossTopic starter

  • Frequent Contributor
  • **
  • Posts: 327
  • Country: gr
Re: Simple DIY Linear PSU
« Reply #67 on: March 14, 2020, 02:49:20 pm »
-I removed the BC557 so now Q2 = S8550
-I removed the D209Ls and I put one D1408

I get the same measurements.

I use an LM358. It's voltage at pin 8(in respect to pin 4) is 9.69V.
With or without load.
« Last Edit: March 14, 2020, 02:55:52 pm by panoss »
 

Offline iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Simple DIY Linear PSU
« Reply #68 on: March 14, 2020, 03:12:37 pm »
OT: It has been demonstrated many times here the LTspice gives pretty good results with DIY linear PSU simulations..  :-+
Readers discretion is advised..
 
The following users thanked this post: panoss

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Simple DIY Linear PSU
« Reply #69 on: March 14, 2020, 03:59:14 pm »
-I removed the BC557 so now Q2 = S8550
-I removed the D209Ls and I put one D1408

I get the same measurements.

I use an LM358. It's voltage at pin 8(in respect to pin 4) is 9.69V.
With or without load.
What are the CV op-amp's input voltages for the 3 loads? Also Q1's Base voltage.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline panossTopic starter

  • Frequent Contributor
  • **
  • Posts: 327
  • Country: gr
Re: Simple DIY Linear PSU
« Reply #70 on: March 14, 2020, 04:30:46 pm »
By 'input voltages from LM358' you mean voltage on pin 5 (and maybe 6), right?
Here they are:
RL=no load,  Vpin5=4.63V  Vpin6=3.18, Q1's Vb=8.97V 
RL=20 Ohm,  Vpin5=4.67V  Vpin6=2.26, Q1's Vb=8.55V   
RL=10 Ohm,  Vpin5=4.70V  Vpin6=1.72, Q1's Vb=8.30V   
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Simple DIY Linear PSU
« Reply #71 on: March 14, 2020, 04:45:04 pm »
By 'input voltages from LM358' you mean voltage on pin 5 (and maybe 6), right?
Here they are:
RL=no load,  Vpin5=4.63V  Vpin6=3.18, Q1's Vb=8.97V 
RL=20 Ohm,  Vpin5=4.67V  Vpin6=2.26, Q1's Vb=8.55V   
RL=10 Ohm,  Vpin5=4.70V  Vpin6=1.72, Q1's Vb=8.30V
So the voltage across R10 is about 4V?
What is the unregulated input voltage?
Try doing the test at a lower output voltage setting.

When the Inverting input voltage is less than the Non-inverting input voltage, the output of the op-amp should go to its maximum voltage. About 1.4V less than supply voltage for the LM358.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline panossTopic starter

  • Frequent Contributor
  • **
  • Posts: 327
  • Country: gr
Re: Simple DIY Linear PSU
« Reply #72 on: March 14, 2020, 04:52:56 pm »
So the voltage across R10 is about 4V?
It 's 3.26V, no matter the load.

What is the unregulated input voltage?
18.50V

Try doing the test at a lower output voltage setting.
A regulated output from an SMPS would not be a good power supply, right?
 

Offline panossTopic starter

  • Frequent Contributor
  • **
  • Posts: 327
  • Country: gr
Re: Simple DIY Linear PSU
« Reply #73 on: March 14, 2020, 05:00:41 pm »
When the Inverting input voltage is less than the Non-inverting input voltage, the output of the op-amp should go to its maximum voltage. About 1.4V less than supply voltage for the LM358.

RL=no load, Vpin7=8.52V   Vpin8=9.69V
RL=20 Ohm,  Vpin5=8.13    Vpin8=9.30V
RL=10 Ohm,  Vpin5=7.87    Vpin8=9.04V
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Simple DIY Linear PSU
« Reply #74 on: March 14, 2020, 05:05:51 pm »
No problem using a SMPS so long as the voltage is correct and holds up when loaded.
The Collector of Q1 should be close to 18V?
It seems that the output stage has too little gain or a fault.
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