the Poor Man 660 support thread

[Kingston]

I am using the Avel lindberg toroid for my 660's the rating for the heater psu is 6.0A, is that enough? would that be the cause of excessive heat on the LM337?

regulator IC getting hot is a completely different thing to a hot PSU transformer.

the heat in the regulator is caused purely because it's dropping about 8-10VDC to 6,3VDC. The difference voltage has nowhere else to go but as heat to your cooling block.

 

[Kingston]

about PSU transformer heat.

I was recently given a bit of insight into how to specify a PSU transformer for projects like these. Turns out the heater part of the PSU specs that are floating around here are severely underrated. They run far too close to specified maximum (max 6A winding running at some 5,2A). It means the transformer is going to get HOT and the core will start to saturate.

If you don't want to cook eggs with your PSU trafo in a situation like this I was told you should rate the required current to about 60% of the maximum for the winding. In this case, the 5,2A should be pulled from something like 8A max rated windings. Otherwise you're in for core saturation effects, heat, and transformer "singing".

wrong. this will be different from manufacture to manufacture. a good one will calculate the needed headroom.

of course it will be, but now were playing the wheel of fortune: will they or won't they? Over-specifying the transformer is a negligible cost increase, and then we'll always be in the safe margin.

 

[Knarleybass]

so, aside from heat sync and possibly adding a fan there is no other way to keep that LM337's heat down?

 

[[silent:arts]]

so, aside from heat sync and possibly adding a fan there is no other way to keep that LM337's heat down?

mounting them [isolated] to a big enough heatsink and chassis should work fine

 

[Knarleybass]

you can see the heat sink that I made, it's mounted on a little maple block to keep it isolated from the case, and it's still getting so hot that it's shutting down. I guess I could add more pieces of aluminum to dissipate the heat.

 

[radiance]

(max 6A winding running at some 5,2A)

Where is this current draw meassured? Before or after the rectefier?

Also, when using the regulated heater supply you'll definitely need more than 6A to get the 5,2A needed by the heaters..
With a full wave bridge + capacitive input filter you'll get the following:

V d.c. = 1,41 X V a.c.
I d.c. = 0,62 X I a.c.


EDIT: I'm talking about a stereo unit here...

 

[Moby]

you can see the heat sink that I made, it's mounted on a little maple block to keep it isolated from the case, and it's still getting so hot that it's shutting down. I guess I could add more pieces of aluminum to dissipate the heat.

make it bigger and black.This is too small  :-\

 

[Knarleybass]

i measured the current draw after the slow start heater psu

 

[[silent:arts]]

Knarleybass, mounting the regulator [isolated] to the chassis will give you much more cooling than the heatsink you created. should be enough.

 

[[silent:arts]]

... Also, when using the regulated heater supply you'll definitely need more than 6A to get the 5,2A needed by the heaters...

mhm, you mean all the working units out don't work correctly? ???

 

[khstudio]

Thank you EVERYONE for your help.  ;D

Doug,
The heat is with the lid off!

The IRF820 gets about 200F (this is directly shooting it with a laser thermometer - NOT the heat sink)
Not sure if this is a problem.

I'm not sure the "Current draw" is the source of all my problems but I will address the TX at some point.
I spent a LONG time last night making sure every single connection & measurement was correct... & it is.

Studying the schematic, your right Doug, the "Draw" is what makes them move... Why the opposing channels threshold effects the others meter, etc... I can't figure out.

It's time to change my heater resistor & hook up ONE channel at a time to troubleshoot.
I'll post back soon, wish me luck.

& Happy 4th to everyone!

I'll be spending the day eating McDonalds & trying to fix this beast.

 

[tommypiper]

Kevin, I would put your Edcor on the back plate, outside the box like Doug suggests (I thought of that the moment I saw your pics, even before I read his comment).  And as others have suggested, maybe get a cheap toroid to replace the heater supply coming from the Edcor. 

Otherwise your build looks fabulous! 

I can't comment further as my build is still in parts...

How, where, did you make your front panel!?

 

[radiance]

... Also, when using the regulated heater supply you'll definitely need more than 6A to get the 5,2A needed by the heaters...

mhm, you mean all the working units out don't work correctly? ???

No, I'm not saying THAT.

All I'm saying is that according to the math, you should have a transformer capable of delivering 8,38A, to get the 5,2A for the heaters.
I sure hope these "other" group buy transformers (I was not refering to the edcors in paticular) are over specced.
I've got one myself so....

 

[khstudio]

OK, I removed the:
Gain, Threshold, Attack & Meter connectors from each PCB (one at a time!!!)
to start narrowing down the problems but the distortion still occurs when compressing, on each channels.

What's strange is... the disconnected meter STILL moves even though it's not even connected.
Moving the wires around also makes the unconnected meter move... weird.
The meters are Sifam:
http://www.sifammeters.com/Audio_Level/AL19_Retro.htm

 

[Kingston]

(max 6A winding running at some 5,2A)

Where is this current draw meassured? Before or after the rectefier?

5,2A is from the tube datasheets. A stereo poorman, meaning 4X 6BC8 and 4X 5687 will eat about that much.

A brave man does not rectify, and feeds AC for the heaters. Did anyone try this with poorman? noise?

 

[[silent:arts]]

What's strange is... the disconnected meter STILL moves even though it's not even connected.
Moving the wires around also makes the unconnected meter move... weird.

Kevin, I know you are not new here, but you haven't read all posts in this thread, didn't you ???

 

[khstudio]

What's strange is... the disconnected meter STILL moves even though it's not even connected.
Moving the wires around also makes the unconnected meter move... weird.

Kevin, I know you are not new here, but you haven't read all posts in this thread, didn't you ???

Yes I did... I was just stating that I'm having "Similar" problems but haven't found a working solution, YET.
I'm going to stop talking about the meter because it's NOT my main problem here!!!

DISTORTION, DISTORTION, DISTORTION!

I'd like to see a voltage chart with test points from a working unit.
Including:
Negative bias voltages @ different levels & their corresponding gain reduction.
& Cathode voltages  (Plates are noted in schematic)

 

[khstudio]

Kevin, I would put your Edcor on the back plate, outside the box like Doug suggests (I thought of that the moment I saw your pics, even before I read his comment).  And as others have suggested, maybe get a cheap toroid to replace the heater supply coming from the Edcor. 

Otherwise your build looks fabulous! 

I can't comment further as my build is still in parts...

How, where, did you make your front panel!?

Thanks man,
I Do have a big Steel Plate that goes in between the PT & Circuit!!!
And as of now, any hum that may show up from the PT being in the case is not my main problem nor do I think it has anything to do with the Distortion I'm getting while compressing.

ALL my Faceplate are Vinyl stickers. I design them then send them to my brother who works at a sign company and he prints me what I need... pretty cool huh?  ;D  They aren't the easiest thing to work with but once on, they look great & are very stable.

 

[khstudio]

OK... I removed my time constant switches & wired it TOTALLY STOCK!

good

Good for trouble shooting YES but it's STILL slow... Way too slow of a release for my liking.

 

[emrr]

I meant on the back for heat purposes.

A filament transformer run terribly much under the rated current draw will always give a higher than rated voltage.  You can't really have it both ways, at least not that I've seen.  I can actually point to the opposite design strategy in the 1946-56 Collins 212A console, which runs a 7V filament transformer that's right at or slightly under current spec, and I've yet to see one burnt out.  You just can't touch it.  And no, they never turned them off.  I had one example that had been on nearly continuously from 1950 to 2002.