SI 5066 problem (supply present at output)

[DcibeL]

Hi,
I have finally found the time and energy to make a project of a SI 5066 board. Thought it would make a nice afternoon project. I should have really tested it to make sure it worked fully when I bought it, but I just checked to make sure it would turn on. Here's a list of what I have done:

-Install SI 5066 board into small hammond chassis
-use 14V 3A LCD monitor supply for power (switching supply)
-Banana plug outputs
-RCA inputs
-polypropelyne coupling capacitors offboard
-shorted C3, C4
-replace volume pot and power switch
-replace R1,R2,R4,R5 with 20K 1% resistors
-remove R01, R02
-replace 330uF on-board power supply cap with 680uF
-add 18,000uF supply cap offboard (I know it's huge, but it was free)
-add solder lug to bottom of Tripath chip (where there is holes in the pcb)

So I powered it up, and there is 14V present on the speaker output. I searched through the forum, and read that it is usually a problem with the 2.5V bias not being present on the input. I checked this by measuring voltage from ground to the amp side of the coupling caps. Only about 30mV was present.

I have checked for shorts everywhere, and stared at the board for the last hour, but I cannot see any problems with it. What am I to do?

Here's a high resolution shot of the input section of the board:
Top
Bottom

[BWRX]

Pretty decent soldering job there. Are you sure there aren't any solder bridges between pins on the chip? The board looks fine so that leads me to believe that there's either a solder bridge somewhere or you applied too much heat to the chip while soldering the slug and nuked the chip.

[DcibeL]

I am thinking that possibly the chip is nuked. I have worked as an electronics assembler soldering parts as small as a grain of salt, and have applied large amounts of heat to components (rework) without damaging them at all. Nevertheless, it is a possibility.

I have looked at the pins of the chip under 5x magnification, and they are all good.

[Edit: Make that 0mV from amp side of coupling cap to ground. Also 0 Ohms measured from amp side of coupling cap to ground. Somewhere there is a short]

[BWRX]

Originally posted by DcibeL
I have worked as an electronics assembler soldering parts as small as a grain of salt

I figured that might have been the case after seeing the nice shiny joints and clean solder bridge

Hook your amp up to power again and check to see if the internal 5V regulator is working. You should see 5V on pins 1, 4, and 9 of the chip. Pin 1 is the pin closest to the notched corner of the chip.

Yep, a short on the input would drive the output to the positive rail because the amplifier is inverting.

[DcibeL]

Actually, my soldering on this board could be better. I don't have a very small tip on my own soldering iron, so it is very easy to apply too much solder. With the right tools, soldering done by hand can be nearly indistinguishable from soldering done by a machine.

5V is present at the amp side of C5 (easier to measure there without shorting anything). I guess that means the chip is working. I don't know, I guess I'll poke around at it some more tomorrow. I have midterms to study for.

Thanks for your help.

[Pano]

I think you're on the right path, must be something sorted at the input. Even 50K to ground at the wrong place will cause big DC offset.

Check the connestions, check the schematic, keep your fingers crossed.

BTW, cool logo!

[DcibeL]

Obsessive compulsive behaviour has made me look at the amp once more before I go to bed tonight.

Ok, so it turns out that the coupling caps have to be on the amp side of the volume pot. I really don't know why I didn't figure this out earlier. Since the caps are blocking DC from the amp, not from the source, the pot with it's volume at the lowest position was shorting the DC bias of the amp to ground.

All is working now, thanks to all that helped.

[Pano]

Yep, that will do it! Don't ask how I know....

Glad you got it sorted!