No V- or AV- and U2 overheating

[Skilldibop]

I built this kit a while ago and it never worked. I recently re-visited it while bored and decided to try and fix it.

I fixed a couple of issues with a burned out 5v regulator and hooked it to a constant current supply to try and find the cause without blowing up more parts.

I tested the 5v regulator at U5 out of circuit and its known good. The 3.3v regulator downstream is also good. I can stop the overheating by disconnecting L3. Though L3 tests good also.
Next to avoid damage to U5 I injected 5v at the high side of C20, when I do this I get 5.06v at +5v +3.29v at +3v, 5.06v at AV+ but 0v at AV- and V- test points.

When I increase the current the top part between pins 1 and 8 of U2 starts to overheat. When I remove U2 still no V- or AV-. I can find no bad shorts to ground on the output pins of U2. there is one from pin 3, which shows in the datasheet for a TL084 as "1IN+" I.E the non-inverting input of the first OPAmp in the package. Though looking at the board layout this looks to be intentionally soldered to the ground plane, although the tiny clearances around the pads and the shiny red solder mask make it very difficult to accurately follow traces without a microscope so cannot be 100% certain of that. This is where it'd be much easier if there was a schematic available to download for troubslshooting.

Is that correct that the +input should go to ground? Is U2 overheating due to a fault on an output or lack of Vcc-? IS there a schematic I could download to better understand what's going on and what U2 should be doing?

With U2 removed applying 9v to the main power input all other voltages look fine. Vcc- on U2 looks to come from U4, so I'm assuming this is where I should see -5v be created. The output pin of U4 is short to ground and I get 0v on the other two pins. I assume the output grounded is correct to make the reference pin (2) -5v relative to the output, or am I confusing how that works. I would say that maybe this means U4 is bad, but I've removed it and tested out of circuit and it works fine.

Following the traces pin 2 of U4 goes to V- (TP25) then to L1, D1 and ground via L2 as well as the collector of Q1. L1 and D1 test ok. D1 has a forward voltage of 200mV, which sounds ok for a schottkey diode, L1 has a DC resistance of 3.6Ohms, L2 has a resistance of 5ohms, so they look fine too.
When the circuit is energized D1 has a voltage drop of 0.2V measured both ways, so it is conducting. C>E voltage of Q1 is 8.1v so Q1 is off, again without schematics I don't know if Q1 should be off or on. My gut feeling is it should probably be ON as without that and D1 forward biased U4 pin2 is going to be pulled down close to ground and it can't give me V- like that.


Any help would be much appreciated as am getting tired of reverse engineering this, as a novice this is quite hard work!

[MNTech]

The schematic is on the last page of the assembly manual.

The overheating of U2 and the short to ground on the output of U4 both say U2 is bad.
A ChipQuik removal kit will help you safely remove ICs without damaging the PCB or the IC.
video: https://www.youtube.com/watch?v=FTQqjggeklo
Here is one place to buy it.
https://www.digikey.com/product-detail/ ... -ND/304148

After removing U2 you can test the negative supply. It's a switching supply that depends on a pulses from the MCU.
Step 2-A tells you to short JP4 if the 3.3V is correct. This powers the MCU.
-1.4 volts at the base of Q2 tells you the pulses are being generated.

Good Luck