TDA20XX AMP power supply question/s.

[Jdansti]

^Thanks for providing a better explanation and correct values than my overly simplistic approach! 

[jogina111]

ok, so I powered both preamp and powerc with the 20v laptop power supply, everything was working fine except the jfet's bias. While biasing the jfet, a loud click was heard and out of question, I pulled out the cord, and checked everything, preamp and speaker but the output cap on the 2030 was hot and the top was bulky. I'm sure its wired the correct way, 25v rating and  negative side to output. Any Ideas?

[gritz]

ok, so I powered both preamp and powerc with the 20v laptop power supply, everything was working fine except the jfet's bias. While biasing the jfet, a loud click was heard and out of question, I pulled out the cord, and checked everything, preamp and speaker but the output cap on the 2030 was hot and the top was bulky. I'm sure its wired the correct way, 25v rating and  negative side to output. Any Ideas?

Do you mean the cap's negative terminal was connected to the 2030's output, or towards the loudspeaker (I'm assuming negative ground here obviously)?

If everything is the right way round then one possibility is that the chip is oscillating ultrasonically and the ripple current fubarred the cap. It's not easy to spot without an oscilloscope, but a hot chip with no input signal is a clue. What schematic are you using? Do you have a Zobel network between the chip output and ground? The 2030 datasheet schemo for a single rail amp shows a 1 ohm resistor in series with a .22uF cap between output pin and ground. See figure 15 here:

www.datasheetcatalog.org/datasheet/stmicroelectronics/1458.pdf

I'd suggest connecting an amp to a psu via a 100R resistor for initial power up (before playing any audio signal) as this will help show up faults like output offset voltage and keep the magic smoke where it should be (the resistor will go pop if there's something wrong, not the speaker or the amp).

You do have a heatsink on the chip, don't you? the chip's tab is connected to supply negative (which is ideal for a single supply amp with a negative ground).

A 25V cap is a bit of a meagre rating imo, I'd be more comfortable with 50v. You mention that the fet bias was off also, so I'd check the open circuit voltage output of the psu too. It's probably alright, but it's worth checking everything. Hope this helps.

[Jdansti]

If he only has 25V caps, could he use two caps of 1/2 the value of his current cap (e.g., 2x22uF instead of 1x47uF) in parallel to divide the current between them? Would this allow him up to 50V through the caps, or do caps not work that way?

[jogina111]

I'll try a 50v 200uf cap tomorrow.. yet its kinda scary making a cap explode and having the top on your eye.   is it ok to use 2 lm317 to regulate voltages going to preamp and power amp  kinda like this?:   
         preamp-----lm317-----PSU Out-----lm317-------power amp

might use a tda7240 instead....

[gritz]

An LM317 between power supply and preamp should be fine as I don't suppose your preamp uses more than a few milliamps. If your preamp has a lot of gain it *might* benefit from an additional small dropper resistor and filter caps between the 317 and preamp to remove the last vestige of hiss, but if the preamp hasn't got a lot of gain then it should be ok. Laptop switched mode power supplies can be pretty noisy.

Given the catastrophic failure of the first coupling cap I'd double check everything as per my previous post before throwing the switch again.