working on a talkbox amp (3886 amp)

[birt]

I'm working on a talkbox amp. its going really good. I made a nice layout. I use the single power supply schematic that's in the datasheet of the LM3886. with some modifications ofcourse!

it runs on a 24V 120VA trafo with a bridge rectifier. The 2200uF elco specified in the schematic is way to small to kill the hum. after some experimentation i had one 1000uF cap on the board and a seperate 10000uF cap close to the trafo/rectifier. Works great. so i moved on.

The output cap i used was 10uF and i left out the filter circuitry between that cap and the speaker.

If you leave out the volume control at the input you actually take away a 10K pulldown resistor (pot at full volume). this makes the amp distort a little on the lower notes. Since i don't need those lower notes with a talkbox i decided to lower the input cap (i had it at 470nF to start with). I tried 220n, 100n, 47n and so on.

but when i was trying those things the weirdest thing is happening. I change something and play a bit, turn it on and off a couple of times... and suddenly when i turn it on the fuse blows. so i start measuring and notice the bridge rectifier is shorted between the 2 AC connections. it was an old one so i replaced it with an identical one. I started testing again and after some time the same thing happened.

So i got tired of that and put in a really big one (no clue what type but it looked like the other ones and was way bigger in size) and everything worked again. Until... you guessed it, the rectifier goes to semiconductorheaven.



I have NO clue why i keep on losing money with rectifiers and fuses but when i'm finished with this amp it's gonna be a really good clean talkbox amp. I will also share the schematic (i already told you where i got it from and what i'm modifying) and layout.

Help with the rectifierproblem would be nice.

[PerroGrande]

Well, it would help if you posted the actual schematic.

You keep destroying rectifiers because you're changing them without really finding out what the root cause might be.

So -- what can kill a bridge rectifier?

* Too much current (over the rectifier's rating)
* Exceeding Peak Inverse Voltage rating
* Heat
* Large Hammers

In this case, the chances are very good that you are experiencing, for some reason, problem #1.  Check for stuff like loose connections that could end up grounded, solder bridges (or near bridges), shorts to the case, etc.

Oh -- and what fuse blows?  What is its rating?

[birt]

here's the schematic: http://www.datasheetarchive.com/preview/2007587.html
it's on page 5

the weird thing is that it works every time for 10m up to more than an hour. what kills it always happens when i turn it of and on again.

it is connected to a small box with banana plugs that i use to test amps and transformers. it an indicator and a switch that switches both lines so i can safely work with the unit still plugged in. in that box is also a 1A fast fuse. when the rectifier blows, the fuse blows too ofcourse.

i know it's probably a current problem but it has to be a peak because when i'm just playing the LM3886 doesn't get very hot. it would when it's drawing a lot of current, especially without a cooling block. the first 2 rectifiers were BY164, these are supposed to handle 15A 160V... the only thing i can think of is a peak coming from the 11000uF of caps. The rectifier is soldered right to the lugs of these caps.

[PerroGrande]

You may want to check out this article from Elliot Sound:  http://sound.westhost.com/power-supplies.htm

Have you observed any heating in the rectifier diodes?  That would be a sure-fire indicator that your issue is one of current.

I suspect you are be using rectifiers that are insufficient for the application at hand.  Mr. Elliot's "blanket" recommendation for typical amplifiers producing from 50W to about 500W is a 35A, 400PIV bridge.  Probably a bit of overkill, but highly unlikely to fail.  You may also want to review what the optimal capacitance for your application might be.  You already have the big capacitor, so cost isn't really a factor.  However, you might not be getting optimal performance using one giant capacitor.

You could also rig-up a fuse between the cap and the rectifier...

[dxm1]

10000 uF? That's one big, honkin' cap! Computer grade? Low ESR?

Hmm... Wonder how many amps it draws charging when you first throw the switch?

Inrush current sounds like a good cuplrit... Fatter rectifiers or a BFR would be a good place to start...

[birt]

It's not really a 10000uF cap. its 2 4700's and a 2200 in parallel. Togheter they measure 11741uF with an ESR of .04

Inrush current has to be the problem because after some thinking i found that it allways happened after i drained those caps.
i wanted to measure the inrush current but i think it might blow the fuse in my meter, so i won't :p

I'll look into similar power supplies and see how they have solved that problem. the big capitance is pretty standard in Gainclones so they have to have the same problem.

[Valoosj]

any pictures? And what do you mean by a talkbox amp? An amp with which you can use your talkbox?

[birt]

an amp built in the enclosure of the talkbox. i'll post pictures when everything is boxed up. with optional distortion on the input and with a mic preamp to go to the rest of you pedalboard and you normal amp