MNats PSU overheating problems- SOLVED : )

[Purusha]

[quote author="Samuel Groner"]Re-reading this thread I realised that you're regulating 2x 24 VAC to 2x 12 VDC, right?[/quote]

Yes.

So then use a pair of series resistors for the transformer secondaries to get things down to about 18 VDC at the input.

I really suck at theory in DIY. I found some formula but don't know how to meassure the current
from the secondaries. I also tried a few resistors in front of my ammeter and there were no changes
on the toroid secondaries. What am I missing here?

Alternatively you could set the regulators to +/-24 V and not use the negative supply. Makes about half the heat.

Well isn't this what I am already doing? I connected + wire to +12V and - wire to -12 output of my PSU to get 24V.

 

[Purusha]

OK, I measured the current and it's 6,7mA on one side of the main secondaries out.
I did calculations for 6V drop and got around 900R. So I just put 900R on both secondaries
in series before the PSU?

 

[Samuel Groner]

6.7 mA? And it's getting hot? Impossible, that's not even half a watt dissipation... Did you measure DC current? At the transformer secondary it's still AC. And please measure the current draw (now DC) at the outputs as well.

Samuel

 

[Purusha]

OK, I measured again:

one side of the mains secondary out was 189mA and the the other was 181mA (AC)

Current draw measured at the PSU output was 2.2A (DC)

 

[Samuel Groner]

Ok, now that looks more like a hot regulator! Of course the input- and output current should match more or less, but depending on the meter type etc. the AC current might be a bit tricky to measure, so let's continue with the 2.2 A figure. Is that figure about equal on both + and -12 V and does it sound about right for the audio circuit you use (2.2 A is a hell lot for audio, but that Neve stuff is supposed to draw that much, right)? If not, you got something wrong elsewhere.

So at the regulator input we have 30 V and at the output 12 V. For best regulation and some security we want at least 18 V at the regulator input, so the required voltage drop is 12 V. Makes a 5.1 ohm resistor. Dissipation is a hefty 26 W, so take at least a 50 W part. All in all, your regulators will have to dissipate 52 W less with the resistors if my thinking is not completely wrong... Could someone confirm before Purusha orders wrong resistor values because of me? Alternatively you can temporarily insert a 1 ohm (0.5 W might survive for a few minutes to make the measurement) and check if that really drops the input by 2.2 V--if so, the 5.1 ohm should work well.

The alternative is--as I said above and the preferred way IMO--to ignore the negative part of the PSU, change the according resistor to make the positive output spite out +24 V. Wire the secondaries in parallel. This will reduce total dissipation from 80 W to less than 15 W.

Samuel

 

[Purusha]

Ok, now that looks more like a hot regulator! Of course the input- and output current should match more or less, but depending on the meter type etc. the AC current might be a bit tricky to measure, so let's continue with the 2.2 A figure. Is that figure about equal on both + and -12 V and does it sound about right for the audio circuit you use (2.2 A is a hell lot for audio, but that Neve stuff is supposed to draw that much, right)? If not, you got something wrong elsewhere.

I just measured it in series after the Neve boards, 2 extra LEDs and 4 Relays.
The unit sounds great. Nothing else is hot except those two transistors.

Alternatively you can temporarily insert a 1 ohm (0.5 W might survive for a few minutes to make the measurement) and check if that really drops the input by 2.2 V--if so, the 5.1 ohm should work well.

I did try a 680 resistor just for fun on one secondary but didn't notice any
drop of V compared to before. What am I doing wrong?

The alternative is--as I said above and the preferred way IMO--to ignore the negative part of the PSU, change the according resistor to make the positive output spite out +24 V. Wire the secondaries in parallel. This will reduce total dissipation from 80 W to less than 15 W.

OK, will try this now.

 

[Samuel Groner]

I did try a 680 resistor just for fun on one secondary but didn't notice any drop of V compared to before. What am I doing wrong?

Hm... :?: Can you give a link to the PSU schematic?

Samuel

 

[Purusha]

The only info I have is here http://www.geocities.com/m_natsume/psu.html

BTW, did you said to wire two secondaries together to the positive side of the PSU for 24V?
Why do I have to wire them together? This will pop the fuse I think. And I can use only one
side anyway without rewiring the secondaries cause I have a trimmer for adjusting the output voltage.

 

[Samuel Groner]

And I can use only one side anyway without rewiring the secondaries cause I have a trimmer for adjusting the output voltage.

OK, then wire the input transformer as for a standard +/-24 V supply (AC-CT-AC as shown on the board, with CT the place where both secondaries join), trim the positive output voltage to +24 V and use ground/+24 V to get the supply voltage. You'll still need some heatsinking on the positive regulator, but live should be much easier for it...

If you have some relays or LEDs to power, you might be able to run them from the -24 V rail to share the current between the two regulators.

Samuel

 

[Purusha]

OK, then wire the input transformer as for a standard +/-24 V supply (AC-CT-AC as shown on the board, with CT the place where both secondaries join),

Are you sure I should join those two? I tried and the fuse poped as I said it would.

I just tried the only positive side as 24V for the whole unit and amazingly the positive LM317 regulator doesn't get so hot anymore. But the other LM337 is still very hot even if I am not using that side.

 

[Samuel Groner]

Are you sure I should join those two? I tried and the fuse poped as I said it would.

And you got the polarity right?

But the other LM337 is still very hot even if I am not using that side.

That sounds very wrong. Are you sure that you don't have a hidden short on the PSU?

Samuel

 

[Purusha]

I was looking wrong. The LM337 is fine and the LM317 is hot but it's not such a big deal I think :grin:

I will not try to wire the secondaries in parallel as I don't hink it's needed. Will just use the negative side for the relays and LEDs and positive for thje Neve boards.

 

[Samuel Groner]

I will not try to wire the secondaries in parallel as I don't hink it's needed. Will just use the negative side for the relays and LEDs and positive for thje Neve boards.

Just to clarify this: If you use both PSU outputs you'll need to wire the secondaries in series (wire the input transformer as for a standard +/-24 V supply (AC-CT-AC as shown on the board, with CT the place where both secondaries join) I said).

Wireing the secondaries in parallel is possible for one supply voltage only and advantageous as it reduces transformer heating and regulation--but you'll need to observe the polarity of the windings, otherwise you'll short the transformer.

Samuel

 

[Purusha]

Wireing the secondaries in parallel is possible for one supply voltage only and advantageous as it reduces transformer heating and regulation--but you'll need to observe the polarity of the windings, otherwise you'll short the transformer.

I measured both secondaries vs middle CT/gnd wire and both show +V only. What am i missing?

 

[Samuel Groner]

Didn't understand--you're measuring AC (AC never get's negative on the meter) or DC, with the transformer connected how?

Samuel

 

[Purusha]

I have transformer connected as shown on the PSU board. Sorry I kinda lost it now.

 

[Harpo]

Hi Purusha,

if I get you right, you have a center tapped transformer, meaning 3 wires coming out. You can't wire these in parallel. Think of it as a single coil with a tap in the middle. Joining any wire with another will give a short.

If you've got a dual secondary, meaning 4 wires comming out, you have 2 separate coils that you can set in series, giving you the center tap at this junction, or as samuel wrote in parallel. (a multitap transformer may have 4 terminals as well, so take care if transformers are not labled) Paralleling them is an adventure, as both secondary windings should have the same voltage and polarity.

 

[Purusha]

This is weird, I moved the + wire of my relays and LEDs to the negative PSU 24V side but it doesn't work in this way. :?:

 

[Harpo]

[quote author="Purusha"]This is weird, I moved the + wire of my relays and LEDs to the negative PSU 24V side but it doesn't work in this way. :?:[/quote]
Turn the LED's and diodes across the relay-coils. You changed the polarity by feeding them from the negative rail.

 

[Purusha]

Ah, I see. Thanks :thumb: