Help with 41Hz AMP4

[mikebarney]

I'm having problems with an AMP4. I posted the following on the 41Hz forum a couple days ago, but no responses. Hoping someone here can help. Thanks,
Mike
---------------------------
I've assembled my AMP4 kit (it went quickly, about 1 day total), but I'm having some problems getting it to work correctly, and am looking for suggestions. I am new to surface mount components, but I have many decades of thru-hole component experience - I build/modify lots of audio equipment.
The symptoms are as follows:
1. The DC offset varies all over the place, and although the two 5K pots affect offset, it is very unstable - it moves from wherever I set it (from tens to hundreds of mV, sometimes wandering to over a full volt).
2. The amp works (a bit): it amplifies a sine wave, but on a scope, the 1Khz test signal output from the amp has a smaller much higher frequency signal superimposed (maybe a couple hundred KHz).
3. The Q10 heatsink (9v regulator) is very hot - I can barely touch it (no other chips are hot).
Other info:
1. The 9V, 5V and main (28VDC) power supplies all measure fine.
2. The relay operation works fine, and no fault LEDs are lit.
3. I've fluxed and re-soldered all of the parts on the board, including both chips, a couple times, to ensure good joints.
4. I've visually inspected both sides of the board with a 10X loupe (a few times now), and see no bad joints or bridges.
5. My technique to mount the surface mount components and chips was as follows: apply flux (with pen), tin one pad, flux again, position part with tweezers and touch iron to the tinned pad/lead, flux other lead(s), touch solder to iron, then iron to joint. Flux and repeat for first (tacked down) side, inspect joint with loupe.
6. In Jan's troubleshooting PDF, he suggests that DC offset may be a damaged TP2050 chip. However, I'm concerned that the hot Q10 and high frequency signal riding on the output may suggest other or more problems.
Any suggestions would be appreciated.
Mike