Dead tripath? :'(

[Code]

Yesterday i fitted some 2.2uf obbligato capacitors to my Autocostruire 2020 t-amp. I tested it for 10minutes and it was fine, then in transit one of the capacitor legs managed to come unsoldered from the pcb. After repairing this the amplifier does not do anything, no turn on thump..no sound.

There are no tracks crossed with solder or any damage i can see,
also there is a chance the voltage was turned up while i was away because when i returned to the amp it was set to 14v :|

How do i test which components are broken?

[Code]

I'm going to try the old capacitors just in case..

i have access to a occilascope but i do not know what the paterns should look like.

Are there any PCB layouts that i could build with some of the remaining bits i have? and are the AC aircore inductors compatible with the charlize?

[Trebla]

The Autocostruire dedicated supply for this amp is set to 13.7v regulated. I don't know what supply you are using but I wouldn't have thought 14v would be sufficient to kill it, unless its not regulated.
The chip seems quite robust and has survived a few errors on my part, but has always reset itself after a short while.

[Pano]

Sorry to hear of your trouble.
If you have an o'scope, you can look to see if the switching circuits are running.

Place the probe on one leg of the output filter coils, the side toward the chip. With the scope set to 2 or 5 Volts and time set to 0.2 uS, you should see a square wave of about 10 volts.

Should look about like this:


If that is working, then the power section is running. After that we can trace the inputs. Let us know!

[BWRX]

Originally posted by panomaniac
If you have an o'scope, you can look to see if the switching circuits are running.

Just remember that the outputs are bridged and float around half the supply voltage - many forget that and can ruin their amps if they attempt to fiddle around (I know I did but luckily the amp still worked). So don't just connect your ground probe to either output unless you're certain your scope's ground is isolated from other grounds or you'll short out one of the outputs. If you have a dual input scope and it can subtract one signal from the other than you can hook the two ground probes together and a positive probe to each output. That worked for me when I was at school and also using a scope that wasn't grounded - it works but it's not recommended

[Pano]

Brian is correct, of course. I should have mentioned that.

If your amp is grounded (earthed) thru the PSU, inputs or whatever, you can not connect the ground of the o'scope to the outputs. That would effectively ground the output. Not good.

However, you can do this if your amp has NO connection to ground, e.g. battery, SMPS and nothing on the inputs.

In Code's case, all he needs to do is test between ground and 1 output coil. The waveform above is such an example.

SO: Be careful if you are measuring across the outputs. But if measuring from ground to one side, you'll be OK. Or to put it another way; No worries connecting the scope ground to amp ground. But do be careful when connecting scope ground to any output terminal, as that will cause a short.

Make sense?

[Joseph K]

Code,

I would say: Oy.. I think just the same thing happened to me:
I've just kicked off the power on my tube pre feeding the AMP3, which in effect probably gave a similar kick accross the input capacitor [a 1uF jensen pio in my case] - no sign of life afterwards. I have all the instruments necessary and all show a definite death...
Now I'm looking into a new 41Hz order..
It seems to me that these Tripath chips are sensitive in a way.
And they are very sensitive to the grounded output, as well, BWRX is absolutely rigth.

Well, let's hope that this is not the case for You.

Ciao, George

[Code]

thank you all for the help, i will test as soon as i can access the scope

[Joseph K]

By the way, I think it's possible to have some proof also without the scope: You should connect the PS accross a multimeter in current mode, and measure the standby current both in sleep / awake mode. I don't know about the 2020, but in case of the 2021 [and it's almost the same beast] it should be some milliamps in sleep / 30 - 50 mA in awake mode. If it's more than that, like > 100 mA then the chip is gone.

Ciao, George

[Pano]

Oddly, the chips seem to be very robust on the output side. A bit fragile on the input side.

(Don't ask me how I know).