Color Classic Analog Board Trouble

Hello, I've been working on a color classics for a few nights now. Managed to fix the logic board after a recap, several trace repairs and a new DFAC-II sound chip. However, the analog board simply refuses to work. I've already recapped the analog board and replaced 2x TDA4605, 2xCNY17-3 and 2xIRFBC40.

Before recap: No 5v stand-by, crt part is working if logic board isn't installed (high voltage sound and static on crt) but no 5v/12v on hdd connector
After recap and replacement parts: Same symptom. 60V, 24V and 8V generated by crt power supply are all present, but logic board power supply is completely dead.

Took out the analog board again tonight for more measurements, turns out there's no voltage at all on TDA4605's pin 6, which should be charged by RP23 on power up:

Checked RP23 and it's showing correct resistance matching color rings. CP34 and TDA4605 has already been replaced. On the other end of the trace it goes to a transistor connected to the optocoupler on crt's power supply section. Checked the transistor and it's not shorted, the optocoupler is new and crt section is working fine, so I guess I can rule this part out.

Next I tried to bridge power supply's TDA4605 pin 6 to CRT's TDA4605 pin 6 to jump start it. However this pulled down crt's TDA4605's pin 6 to around 2v and it stopped working as well..

Next I tried to cut the trace that goes to the transistor. This sorta worked, voltage on pin 6 jumped to 10V on power up and power supply tried to start, then it started to make ticking sound at around 3Hz:


Checked pin 1 on TDA4605 which is voltage regulating pin according to datasheet, it jumped to around 4V for a brief second, which I think triggered the power ic's overload/short circuit protection:



Checked resistance reading on all secondary voltage pins and they all look fine, no shorts. Also desoldered 2 schottky diodes and they're not shorted. Actually I get 12V/5V on HDD power plug before power supply enters protection mode. Also tried to connect a hard drive to the plug, same issue. Maybe the voltage feedback part isn't working? Checked all smd components and didn't find anything. Also tried to adjust the potentiometer but power supply keeps ticking through the whole range. Also cleaned up the whole section with alcohol just to make sure there's no leaked electrolyte on them, still no dice. Even swapped mosfet, driver ic and optocoupler between the crt part and power supply part, crt part still works so they're not the issue.

I even desoldered the transformer and measured inductance of each winding, no shorts:


Well at this point I'm out of ideas. Anyone had similar issue with CC analog boards? What to check next?

I'll try to figure out the deal with the trace i cut that connects pin 6 of power supply's TDA4605 to the transistor on crt's part, see why pin 6 is pulled to ground with this trace intact. I'm still leaning towards some voltage feedback issue, but I've already check all smd components... I'd really appreciate your input on this.

Spent some time working on it this morning and noticed voltage on pin 6 drops to ~4v immediately after the power supply tries to start, then charges up slowly to 10v and next cycle continues. In theory the capacitor should be charged up by auxiliary winding through RP34 and DP15. They both tested fine.


Aye screw it, since the power supply section only provides +12V, +5V and -12V (which then generates -5V through 7905 regulator), I have something in mind...

Ok it might seem ridiculous but I got it working!

Bypassed the power supply section (desoldered transformer and switching mosfet) and connected 5v/12v/-12v capacitors to an atx psu, finally something showed up on screen. The image still has some stability issues and has hour glass distortion in the middle. Also need to order a picopsu and figure out how to implement soft power, but at least it's working now

Been working on this color classic on and off for a few days now, also spent some time reverse engineering the analog board, here are my findings:

- Optocoupler near the smaller transformer (which outputs 60V, 24V and 8.3V and powers the CRT section) is for soft power only! It's NOT used for voltage regulating, unlike the optocoupler on primary side.

- 5V standby voltage (which goes to A5 - Trickle+5V on logic board), is generated from 8.4V on CRT section through a 7805 regulator. It's NOT generated by the main power supply section

- The standby 5v powers CUDA, which senses the power button on keyboard and sends A7 - SHUTDOWN_ENABLE signal to the optocoupler. If main power supply doesn't provide 5v/12v within ~2s, SHUTDOWN_ENABLE will become low again.


- The trace I cut previously is indeed for soft power control. If SHUTDOWN_ENABLE is low, the transistor controlled by optocoupler will short the capacitor on TDA4605's power pin to ground and prevent the switching ic from working. If SHUTDOWN_ENABLE is pulled high by the logic board, the capacitor will be charged up through RP23 till it reaches TDA4605's power on threshold, then TDA4605 will take over and charge CP34 with the auxiliary winding from the transformer.

For some reason, the SHUTDOWN_ENABLE signal is kept low on my analog board, which prevents the main power supply from working. By cutting the trace I mentioned on first post, soft power control should be bypassed, but it still doesn't power on and keeps triggering protection mode of the power supply ic (i think)..

Also thanks @ArjenCNX for your input, I took another look at the component you mentioned, it's indeed a voltage reference, I thought it was a standard transistor (tested it in circuit with LCR meter in diode mode and it looked fine). Desoldered it and built a simple test circuit according to datasheet, it tested OK:

Also checked most traces underneath filter caps on secondary, the analog board was soaked in electrolyte when I first got it as both 5600uf capacitors have vented, but luckly no damage to traces or any corrosion. Think I opened it up just in time..

Anyway, whatever the fault is, I've gave up and obliterated most of the power supply's components to make room for a standalone power supply module.

Found this thing inside a Lacie external SCSI drive box, it does 8A on 5V, 3.5A on 12V and conveniently -12V as well, saves me some space for another 12V power supply or an isolated dc-dc module:


Hacked it inside the color classic and it worked!


Well it kinda worked, as only about half the time the CRT turns on when I flip the switch, rest of the time I can hear high voltage sound for 1-2 seconds but then CRT shuts off, but logic board is still on and I can see scsi activity. Guess it's some weird timing issue? Like the CRT power supply needs to be ready before you can switch on the logic board and feed the video signal? In the original color classic you have to wait 3-5 seconds after you connect mains power before you can power it on using keyboard, maybe it's something similar?

Also tried to implement soft power with a relay controlled by SHUTDOWN_ENABLE signal and switches mains supply to the standalone psu, but the relay doesn't click when I press power button on keyboard.. Need to spend more time looking into it..