Bi-polar Power Supply Questions....

[Deric]

I have noticed them getting warm but not hot. I suppose it's possible though.

[JazzGuitarGimp]

If the diodes and resistors are causing the caps to heat up by radiating heat through the air, then those four parts would have to be getting hotter than the caps, and by a good amount I would think. I have heard zener diode tend to be noisey, but you've already looked at the cap with a scope - certainly, had there been enough noise to cause the cap to heat up, you would have seen it. I think I might be inclined to replace the zeners with 3-terminal regulators and be done with it. You might try adding a 10nF ceramic cap in parallel with each zener diode. If it is some sort of noise issue, then caps should quelch it. Or perhaps that's what the cap across the AC inputs to the bridge was for?

I just noticed you're using the ultra fast 4007's. If you decide to add back the cap across the bridge, you'll probably want to change these to standard 1N4007's - actually for this circuit, you could use any of the 1N4xxx series as the voltage is very low.

[Deric]

If the diodes and resistors are causing the caps to heat up by radiating heat through the air, then those four parts would have to be getting hotter than the caps...

This was my thought too and why I figured it must be something else. C3 was too hot to touch and the diodes/resistors were no where near as hot.

I'll try a couple caps across the zeners.

Is there a benefit to standard 1N4xxx diodes vs the UF versions? Is it worth swapping them out just to see? I have a pile of the UF ones...could probably scrounge up 4 standard ones if it's worth trying.

I may just finish the amp and test with this power supply to see if the reverb circuit and the rest of the project even work.



Thanks for all your help Lou!

[JazzGuitarGimp]

Is it worth swapping them out just to see?

In a word, yes. In a whole bunch of words, here's what I am thinking: We know removing the zeners fixes the overheating capacitor problem. I believe you've said you've tried replacing the zeners, so those should not be the culprit. The circuit is really simple: there's not really much left to look at. I am thinking maybe the faster switching time of the UF4007s is causing a strange reaction with the zeners. I would start by replacing the four UF4007s with 1N400x's. If that does not cure the problem, I would add the missing cap at the input to the bridge. If that doesn't fix the problem, I am seeing a 7815 and a 7915 in your future!

[Deric]

I'll give it a shot.

[Deric]

So...replaced the diodes with standard 1n4007. No luck. Added cap across bridge. No luck. Added caps across zeners. Still the same.

Going to finish the rest of the amp, work any other bugs out and test my reverb implementation. Once all is good I'll start laying out a new regulated power supply.



Thanks for all the help Lou!

[JazzGuitarGimp]

Man, that's a weird one for sure! Sorry we didn't figure it out. I think in the long run, the regulators will serve you better than the zeners.

[Deric]

Well......it works!!! Went ahead and wired up the rest of the amp and worked most the bugs out. Will start working up a new power supply for the bias and verb but at least I know the rest of the circuit works.

Still some small bugs....B+ too high...

Will start a build thread in the Fender forum.

Thanks again for all the help!!!

Suggestions for regulated PS welcome.

[Deric]

So amp is up and running with a few issues. B+ is ~50v high so going to add a Mosfet B+ reducer. Since I have to order the parts for that it's time to start laying out a new power supply, bias, reverb board.

I plan to use 7815/7915 regulators in most likely a TO220 package. Will I need to add heatsinks to regulators? Is the circuit really as simple as the example attached?

[TUBEDUDE]

By "cap across the bridge", you meant a cap across each diode in the bridge, right? Just checking.

[Deric]

No - the original Super Champ schematic has a small cap across the transformer secondary.

[JazzGuitarGimp]

So amp is up and running with a few issues. B+ is ~50v high so going to add a Mosfet B+ reducer. Since I have to order the parts for that it's time to start laying out a new power supply, bias, reverb board.

I plan to use 7815/7915 regulators in most likely a TO220 package. Will I need to add heatsinks to regulators? Is the circuit really as simple as the example attached?

This design is a good start. I would consider adding 0.1uF ceramic caps in parallel with the 1,000uF reservour caps, which will help with any high frequency noise that might get in, as well as 4.7uF tantalum caps in parallel with the output diodes, which helps to stablize the regulators.

[Deric]

Thanks. I was considering the small caps across the 1000u caps but had not thought about the tants on the output. Was considering leaving the diodes on the output out too - it's my under standing they are just there in case it gets hooked up backwards. Since this will be on the same PCB with the reverb that won't be a potential issue.

[JazzGuitarGimp]

The diodes as wired, actually do nothing. It's more common to see protection diodes wired from the regulator input to the regulator output. For the 7812, the cathode would be at the input. For the 7912, the anode to the input. This is sometimes done because if, for some reason, a voltage is applied to the output with no voltage to the input, the regulator will be distroyed. With the protection diode in place, when power is removed from the input, the output is pulled down by the diode. I've built a lot of power supplies with these ruglators, and I have never opted for the protection diodes.

[Deric]

Cool. I'll leave them out.

Just waiting on parts now...

Thanks for all the input!