Author Topic: 36V symmetric power supply  (Read 12498 times)

0 Members and 1 Guest are viewing this topic.

Offline Richard Crowley

  • Super Contributor
  • ***
  • Posts: 4319
  • Country: us
  • KJ7YLK
Re: 36V symmetric power supply
« Reply #25 on: March 11, 2016, 02:34:58 am »
If you are replacing a faulty SMPS with a linear power supply for a power amplifier there is no reason to regulate the voltage.  A power amplifier is a special kind of "voltage regulator" which follows the voltage of the audio input waveform.  We traditionally use unregulated power supplies for power amplifiers because regulating the supply rails is redundant and unnecessary and fiddly and expensive.  There is NO benefit to regulating the rails, and a number of disadvantages.

There appear to be websites dealing with repair of that specific supply. For example....
http://www.neufeld.newton.ks.us/electronics/?p=232
http://www.badcaps.net/forum/showthread.php?t=10910

There are replacement power supplies which may well be less expensive than having custom transformers wound and building your own PS from scratch..
http://www.fullcompass.com/prod/270616-Alesis-AL9-79-B601
 

Offline kodi

  • Contributor
  • Posts: 19
Re: 36V symmetric power supply
« Reply #26 on: March 11, 2016, 09:50:28 am »
If you are replacing a faulty SMPS with a linear power supply for a power amplifier there is no reason to regulate the voltage.  A power amplifier is a special kind of "voltage regulator" which follows the voltage of the audio input waveform.  We traditionally use unregulated power supplies for power amplifiers because regulating the supply rails is redundant and unnecessary and fiddly and expensive.  There is NO benefit to regulating the rails, and a number of disadvantages.

There appear to be websites dealing with repair of that specific supply. For example....
http://www.neufeld.newton.ks.us/electronics/?p=232
http://www.badcaps.net/forum/showthread.php?t=10910

There are replacement power supplies which may well be less expensive than having custom transformers wound and building your own PS from scratch..
http://www.fullcompass.com/prod/270616-Alesis-AL9-79-B601

Actually I know if I would be doing pure power amp with LM3886 I would just slap 150-200VA transformer with some MURs and 10mF on rails. It's PSRR is much higher than those puny 317/337s. However as I said I don't have the schematics for the whole board. I don't know what else is there, so the simplest solution for me was to put the same rated PSU as it was originally.

I appreciate the links, however they are known to me. Problem with PSU is it's spatial design and mostly threads about fixing it are about replacing few bad caps, 2 transistors, diode, one IC and few other  elements, however they will fail again (which I know from my experience, unfortunately).

PSU in the last link is the exact part that's in original monitor, with dreaded JunFu caps in there being heated by power resistors. It's around 100 Euros plus shipping.
Custom transformer 2x30V + 2x14V costs 21 Euros, or 27 with steel cup, I have the rest of the parts at home.

A lot of the time, it's the caps that fail. Check them first and it could be as simple as replacing them to have a working PSU again.

I appreciate the help and you are mostly right. There are 2 power resistors that are too close to electrolytic caps (similar story to Samsung TV PSUs) and drying them out. After 5-6 years they are failing. Unfortunantely I found out about it a little too late, so I have four monitors without PSU already. The problem is - if they are starting to dry out, something else fails as well. I tried to fix two of them and there were different failures, including burned out trace on PCB. New PSU unit costs around 100-120 Euros and from pictures I see that they didn't change the design a bit, so it will work for another few years and will fail.


Ok. Can we just forget the part that this is supposed to power up the audio equipment? I just don't want to waste anyone time on discussion about how PSU will impact my hearing experience. I will rephrase my question:
Is the schematic on the first post of this thread suitable for powering up the device that needs +-36 volts and will draw up to 1A?
From simulations on LTSpice when I put 2200uF in front of the LDOs I got a spike of (Vi-Vo) around 37 V during power up. Datasheet specifies that as maximum operating range for this chip, however absolute maximum is 40V.
 

Offline Halvmand

  • Regular Contributor
  • *
  • Posts: 111
  • Country: dk
Re: 36V symmetric power supply
« Reply #27 on: March 12, 2016, 01:37:28 am »
Ok. Can we just forget the part that this is supposed to power up the audio equipment? I just don't want to waste anyone time on discussion about how PSU will impact my hearing experience. I will rephrase my question:
Is the schematic on the first post of this thread suitable for powering up the device that needs +-36 volts and will draw up to 1A?
From simulations on LTSpice when I put 2200uF in front of the LDOs I got a spike of (Vi-Vo) around 37 V during power up. Datasheet specifies that as maximum operating range for this chip, however absolute maximum is 40V.

No.
You are working very close to it's rails. At loud volumes you may lose regulation if the transformer and caps are not sized properly.
You could wind the transformer to a higher voltage to avoid it, but then your regulators will heat a lot more.
Brings forward another question. Why do you think you need less capacitance when using a regulator?

That extra heat from the regulators plus you will be using less capacitance will result in more ripple. More ripple means more heating in the caps. More heat in the caps plus the warm ambient will make them fail prematurely.
Now you can't party loud all night.
 

Offline obiwanjacobi

  • Super Contributor
  • ***
  • Posts: 1013
  • Country: nl
  • What's this yippee-yayoh pin you talk about!?
    • Marctronix Blog
Re: 36V symmetric power supply
« Reply #28 on: March 12, 2016, 06:14:50 am »
I made the same mistake once - where I had a regulated PSU for a power amp. It would switch off from time to time (hitting current/thermal protection) and distort. I trimmed it down to the basic diode bridge and caps and it worked beautifully.

As for how much capacitance, I just measured the ripple and output and kept adding till I was satisfied, then added some more for good measure.  ;D

[2c]
Arduino Template Library | Zalt Z80 Computer
Wrong code should not compile!
 

Offline Halvmand

  • Regular Contributor
  • *
  • Posts: 111
  • Country: dk
Re: 36V symmetric power supply
« Reply #29 on: March 14, 2016, 02:07:00 am »
I made the same mistake once - where I had a regulated PSU for a power amp. It would switch off from time to time (hitting current/thermal protection) and distort. I trimmed it down to the basic diode bridge and caps and it worked beautifully.

As for how much capacitance, I just measured the ripple and output and kept adding till I was satisfied, then added some more for good measure.  ;D

[2c]

Listen to obiwan. Go unregulated and reap the benefits of lower cost, lower complexity, lower internal heating, better performance and better reliability.
I'd say these component values should do fine since I've seen the monitors mentioned has two lm3886 chips right? 1 for bass and one for treble, plus the fact they're not to big either.
I'd slap a 50-60 VA transformer, rectifier and 4700 uF of capacitance / rail and be done with it..

If it where my personal project, i'd make a 3 years warranty on that PSU.
« Last Edit: March 14, 2016, 02:17:12 am by Halvmand »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf