diyAudio (
-   Chip Amps (
-   -   Shorted input... (

iggy111 23rd January 2013 08:55 AM

Shorted input...
I accidentally soldered the input + to sgnd on my chipamp :(
I was replacing the R2 for a lower value of 10kR to lower the DC offset, but then I soldered the in+ to the wrong end of the resistor.
I turned the amp on and off multiple times and tested with source before I realized what I've done. I was getting nothing but buzzing from the speakers, then I resoldered the in+ and now I hear the amp working but very very quiet and with some noise.

So have I fried the chips, or could it be something else? Don't know where to start testing...

I think I need a :drink:

AndrewT 23rd January 2013 09:20 AM

What schematic did you use.

A fully populated version with all the "optional" components will be quite well protected from and inverted signal form an unbalanced source.

But if you built any of the stripped down versions, because "less is good", then anything may have happened.

iggy111 23rd January 2013 10:48 AM

I have the amp, changed a lot of it as you can see in the pic.
But I looked at my scheme once again and I stand corrected - I actually soldered the in- to the wrong side of my R2, NOT in+ to sgnd!
You can see what I did here, the dashed grey line is where I should have soldered IN- in the first place of course.

Mooly 23rd January 2013 11:07 AM

It could be the speakers or the amp or both...

What you have done will have caused a huge DC offset and that will have been applied to the speaker. Has it survived ? You need to check with another or check that speaker on a known good amp.

The chio may also have fried and been damaged. Check the DC voltage on pin 3 which should be zero. Also check R3 as that is low enough in value to be damaged by any rail voltage flying around. Also check that 2.7 ohm.

iggy111 23rd January 2013 11:22 AM

Thanks Mooly, will check that when I get back home.
The speaker is not damaged, I always use a test speaker and it's still fine.
I'm hoping that the chips are ok cause I hear the music playing, only very very quiet, and with some noise.

Mooly 23rd January 2013 11:24 AM

If the speakers OK then it can only be the chip and/or a low value resistor thats gone up.

johnr66 23rd January 2013 04:00 PM

The way you show it may have caused HF oscillation and can cause Rz to burn. This schematic is terrible. Without a coupling capacitor on the input and FB loop, any DC on the input will cause larger DC on the output and also disturb the bias current set by R2. Then there's that 2.7 ohm resistor returning the input signal to ground. Why? It is as if they are afraid to return the input signal to ground!

iggy111 23rd January 2013 10:28 PM

Ok, so I have to correct myself again - the speaker bought it, not the amp.

@Mooly - I confirmed the resistors are ok and measured DC on pin 3, which is 0.44v L and 0.19v R - same as DC offset at output - I guess this is ok?
I tried the test speaker again and found that it died. No worries, it's a cheapo :)
So the amp is working ok.

@johnr66 - I changed the original scheme after lots of reading here and suggestions made by more experienced members:
the 2.7R is a ground lift resistor, used for isolating signal ground from main ground, and I removed the Cin and Ci because the amp sounds WAY better without these. I just want a cheap tweaked amp, buying Mundorfs is not on the list yet. R2 was lowered from 22k to 10k because of high DC offset and works great.

And why exactly do you think that the scheme is terrible? Peter Daniel's premium kit also has no Cin, Ci, doesn't use R1, and all I hear is praise for it.

johnr66 23rd January 2013 10:53 PM

The input "lift" resistor is nonsense. The input signal still has to flow in the ground circuit. Putting a small resistance there that is a tiny fraction of the input impedance does nothing. Keeping the heavier supply and output currents out of the input ground path via star grounding accomplishes that.

Redshift187 24th January 2013 02:53 AM


Originally Posted by iggy111 (
And why exactly do you think that the scheme is terrible?

Because Cin and Ci probably would have saved your speaker.

You should always check for DC offset after changes and before you hook up a speaker. Even better would be to have a DC detection and speaker protection circuit, since you're running a DC coupled amp.

All times are GMT. The time now is 12:09 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2