Author Topic: 40 year old power supply  (Read 5122 times)

0 Members and 1 Guest are viewing this topic.

Offline MartiniTopic starter

  • Regular Contributor
  • *
  • Posts: 89
40 year old power supply
« on: April 24, 2016, 09:56:44 am »
A friend gave me a power supply he made in high school in the 70's. It hadn't been powered on for decades.

I opened it and checked for any obvious short, checked the capacitance of the two electrolytics (don't have an ESR-meter yet), checked the forward voltage of the rectifying diodes, removed the oxydation from the fuse and plugged it in.

It didn't blow up ;D

The voltage selection does work and it's reasonnably well calibrated.

But as soon as I put a load accros the terminals, I get a pretty high voltage drop:




The AC is properly transformed and rectified (there is no ripple accros the 1000 µF cap that smoothes the DC after the bridge rectifier, it shows a stable 30+ volts).
After that, I don't know what happens ;D



Do these symptoms shout out any obvious fault or should I keep investigating?
 

Offline Puffie40

  • Regular Contributor
  • *
  • Posts: 54
  • Country: ca
  • Irregular Logic
Re: 40 year old power supply
« Reply #1 on: April 24, 2016, 10:20:23 am »
Since it's homemade, the first thing I'd do is trace out the circuitry and find out what you have. Unless it was a circuit copied from a set of plans, there is the chance you are dealing with a design flaw, and that will be a major hair puller to find.

The fact that the voltage drops a fair bit when a load is put on suggests a problem with the regulation feedback.
 

Offline MartiniTopic starter

  • Regular Contributor
  • *
  • Posts: 89
Re: 40 year old power supply
« Reply #2 on: April 24, 2016, 10:39:01 am »
Since it's homemade, the first thing I'd do is trace out the circuitry and find out what you have. Unless it was a circuit copied from a set of plans, there is the chance you are dealing with a design flaw, and that will be a major hair puller to find.
It was done in school, under the supervision of a teacher, every kid made one and they apparently all worked.
I can't rule out a schematic/build issue but that doesn't seem to be the likeliest as my friend happily used  it for years.


The fact that the voltage drops a fair bit when a load is put on suggests a problem with the regulation feedback.
I'll dig out on that side then. The only thing I can see for the moment, is an ASZ15 on a big heatsink.
 

Offline MartiniTopic starter

  • Regular Contributor
  • *
  • Posts: 89
Re: 40 year old power supply
« Reply #3 on: April 25, 2016, 09:41:59 am »
I did draw the schematics, it took me this much to notice a fuse was blown between the 220 R and the collector of the power transistor |O
I replaced it with a higher value one (I know, I know, but I didn't have a 0.5 A one) and powered it up. Voltage stayed at 0, no matter what position the potentiometer was in. I moved things a bit, thinking I might have waken up a bad solder joint somewhere while reading the circuitry...it worked for two seconds and two Zener diodes went kaboum! :palm:


But let's start with the first step. Do my schematics make sense to you?

« Last Edit: April 25, 2016, 09:43:37 am by Martini »
 

Offline MartiniTopic starter

  • Regular Contributor
  • *
  • Posts: 89
Re: 40 year old power supply
« Reply #4 on: April 27, 2016, 02:23:35 pm »
I found a bad solder joint, so I reflew (?) it. Replaced (temporarily) the two dead Zener diodes by a 470 ? resistor, replaced the fuse and...voilà !

I drew 300 mA at 10V with a limited voltage drop of 1 V (mostly because of the 470 ? resistor in lieu of the Zener diodes I assume) so I'm gonna call it a victory :)



I noticed some ripple on the output today. Is it safe to assume it's because of the high ESR of the electrolytics? Capacitance was within spec.
(Building an ESR-meter actually is my next project)


Lastly, I'd appreciate comments on my schematics. Do they make sense to you?
 

Offline mij59

  • Frequent Contributor
  • **
  • Posts: 693
  • Country: nl
Re: 40 year old power supply
« Reply #5 on: April 27, 2016, 03:42:01 pm »
At 300 mA output current, the ripple voltage on the 1000 uF cap will be about 3V, you might want to replace it  any way.

The schematic make sense, the potentiometer feeds a two stage emitter follower, voltage regulation will not be the best.
 

Offline MartiniTopic starter

  • Regular Contributor
  • *
  • Posts: 89
Re: 40 year old power supply
« Reply #6 on: April 27, 2016, 04:34:42 pm »
Thanks !


At 300 mA output current, the ripple voltage on the 1000 uF cap will be about 3V, you might want to replace it  any way.
I didn't check it, I will.
(The ripple I'm speaking about is at the outpu, of course).


Quote
The schematic make sense, the potentiometer feeds a two stage emitter follower, voltage regulation will not be the best.
It's not meant to be but I'm interesting in knowing why (assuming I replace the resistor with a stable zener)?
 

Offline mij59

  • Frequent Contributor
  • **
  • Posts: 693
  • Country: nl
Re: 40 year old power supply
« Reply #7 on: April 27, 2016, 05:05:15 pm »
For simplicity assume that the voltage on the wiper of the potentiometer with respect to the positive output terminal is constant, lets call this voltage Vw.

The output voltage will be  Vw minus the base emitter voltage of the two transistors, an increase of the output current will  result in a higher base emitter voltage, therefore the output voltage will drop.
 

Offline MartiniTopic starter

  • Regular Contributor
  • *
  • Posts: 89
Re: 40 year old power supply
« Reply #8 on: April 27, 2016, 09:56:27 pm »
an increase of the output current will  result in a higher base emitter voltage
Oh, it will?
 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1938
  • Country: us
Re: 40 year old power supply
« Reply #9 on: April 28, 2016, 09:27:19 am »
I'd put a big capacitor, 470uf,  from + supply to base of transistor on pot to do some active filtering.
 

Offline MartiniTopic starter

  • Regular Contributor
  • *
  • Posts: 89
Re: 40 year old power supply
« Reply #10 on: April 28, 2016, 03:08:30 pm »
I will check how much ripple there is there.
What's the math behind the size of that cap?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf