PSU for tube amp from PC ATX 200WPSU

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello,

last time I got a few old 200WPSUs and my intention was to rebuild it to get 6.3V 6A instead of +5V, +12 >2A (existing one) for cooling fans or tubes (or 12.6V for filaments?) and HT voltage from +150 up to 250V 200mA (or more?). All of these voltages should be termined by passive non global nfb easy "zener + E follower" circuits for the best sound. So in fact "input" voltages should be slightly higher because "zener + EF" will lower output voltage.

Unfortunately after opening of the cover I did not find no +5V stabilisers nor + 12V to make the modifications.:confused:

Could anybody advice and indicate on some schematic (I do not have any) which modification has to be done?

Please give the response if my idea is clear, because I feel that I have made a mess during writing? :)

PS. One primitive use of the fan from PSU as above is to cut one of the wires of fan-engine, insert 100 Ohm resistor in series and then usin fan for cooling aluminium radiator of any device. Then RPM lowers and the noise disappears.:clown:
 
You are not finding a 5v regulator as the 5v is used as a reference for generating all other voltages. It will probably be circuit specific but to raise the 5v output you really just need to change one resistor in a divider which feeds the output to a comparator. Last time i tried doing this was almost 10 years ago so my memories are a bit sketchy but it was very easy.

Is it worth it? Not if you are one bit interested in the resultant sound. Apparently filtering the output to be acceptable for heater duties is either very difficult or completely impossible.
 
The 5V/12V relationship comes from the main transformer and from the buck inductor, only negative low current outputs usually come from linear regulators. The only way to change this relationship is to rewind these magnetic components

The main transformer usually has two primaries of 20 turns each [sandwitching the secondaries and connected in series], two 3 turn symmetric secondaries for +5V and 4 additional turns in series with these to get +12V

However, both values may be increased proportionally by tweaking feedback resistors [feedback is taken averaged from both +5V and +12V outputs]. Adjusting 6.3V output on the +5V line would produce 15V on the +12V line

Getting +250V from the SMPS is a bit more complex since magnetic components would have to be fully rewound and some transformer isolation issues arise when using such high voltages. Note also that +250V requires a lot of secondary turns and these transformers usually have little or no free winding space

Emitter followers at the output of a SMPS are mostly useless, they would only attenuate LF ripple since high frequencies travel happily through C-E capacitance. Use pi filters with high Q inductors and low ESR capacitors instead and common-mode filters to prevent wires from radiating
 
Dear Eva,

that was what I have been waiting for!!!

I undrestand that you have good skill in switching PSes,
could you send an exemplary schematic of mentioned PS? or indicate a webpage with swirching PSes? like PC ATX 200W?

Getting +6,3V will cause getting +15V: that is very good!!! Even some higher voltage would be nedeed...

+250V is high value and not easy to get, but ~+-200V DC is almost at the start of sw. PS so that will be not a big problem to work on it.

Regarding of the fact that EF will not filter high frequencies I should not agree. Please look at http://www.tnt-audio.com/clinica/regulators_noise3_e.html where is shown that such regulators are better than TL431.
 
Just have a look inside Horowitz. A nice description of an AT ps with explanation of how +5v regulation works. In any case just look for a resistive divider from the +5v output feeding a fraction of the output to a comparator. Tweak one of the resistors by soldering a larger value in parallel to see how the output varies.

I still maintain that to get an output of usable quality to power heaters (even indirect) will take an incredible amount of effort, lots of chokes, multiple filtering stages. As ATX supplies cost next to nothing, all the cheap-skate high end manufacturers would be using them if it was easy.
 
Hi there,

I found an interesting article in the online magazine Silicon Chip num.190, July 2004, which tells how to tweak a PC AT SMPS to feed B+ and filaments, so I decided to buy the whole number although I was just interested in this article (for 8.80$). My surprise was that the whole article is based on a previous published article in October 2003 (num.181) so the one I own is insufficient. I think it is unfair and they had to warn about it. If someone could send me this article I would appreciate it.
Thanks in advance.
Miquel
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.