Problems with used TPS 4000 linear PSU (Model 4303)

[EE54]

I've recently bought a used TPS 4000 linear PSU (Model 4303) for 1.200.000VND(around 52 USD).
I tried testing the load regulation using a simple CC load(LM358 driving a MOSFET), PSU set to 15V, current limit pot turned to max  sweep from 0-3A, measure voltage at binding post. Channel 1 works perfectly, fluctuating within a few mV of 15V. But Channel 2 acts really weird. At 0A, output measure 15V, as I increase the load current, the voltage jumps to more than 30V. When I increase load current pass about 1A, output voltage starts to lower, close to the set 15V as load current is increased to 3A, and during all this, a hissing sound is heard, which disappears when I remove the load from Channel 2.
So I tried connecting a 3R9 power resistor to Channel 2, set to 5V, current limit turned to max:
 -No R: 5.057V
 -With R: 4.552
 -Delta V: 0.505V(9.92% of 5V)
This is way worse than Channel 1, which holds 5V perfectly when the same test is done.
So I try following the troubleshooting section of the manual:http://www.dudleylab.com/Power%20Supply%20TPS-4000.pdf
Everything was right till the Voltage Control Circuit test, step B, in that step, I measure:
 -Channel 1: +1.57V
 -Channel 2: -3.57V
Channel 1 is close to the +2V +-20% while Channel 2 is way off.
So what should I do now?

[aargee]

What happens if you flip into the slave configuration, does the output behave the same?
I assume the voltmeter/ammeter on the PS reflects the same readings?

Hissing sounds like a capacitor venting. Can you localise the noise?

[EE54]

What happens if you flip into the slave configuration, does the output behave the same?
I assume the voltmeter/ammeter on the PS reflects the same readings?

Hissing sounds like a capacitor venting. Can you localise the noise?

Yes, the voltmeter/ammeter on the PS reflects the same readings.
The 2nd output behaves the same when in slave configuration, voltage shoots to over 30V, hissing sounds... I don't think the sound is a cap venting, it's more like some kind of high frequency switching, coming somewhere from Channel 2 mainboard.
Also, the Channel 2 doesn't tract very well. I took some measurement:
 .Channel 1       .Channel 2
 +10.063V        +11.004V(0.941V higher)
 +20.002V        +21.658V(1.656V higher)
 +30.002V        +32.377V(2.375V higher)

[floobydust]

It's a Topward TPS-4000? I have repaired them and find they do have some weaknesses.
They roast the driver transistors. They are on the main board Q02 2SC2270C. I remember adding a small heatsink and 47R 1W resistor from emitter Q02 to Q07 2N3772 base.
I also add a 1N4004 diode across C-E of the pass transistors so the PSU does not die charging batteries.

Check Q02 whichever is Ch. 2 it is likely damaged.

[EE54]

It's a Topward TPS-4000? I have repaired them and find they do have some weaknesses.
They roast the driver transistors. They are on the main board Q02 2SC2270C. I remember adding a small heatsink and 47R 1W resistor from emitter Q02 to Q07 2N3772 base.
I also add a 1N4004 diode across C-E of the pass transistors so the PSU does not die charging batteries.

Check Q02 whichever is Ch. 2 it is likely damaged.

I checked Q2 and all other transistors on the Ch.2 mainboard, all of them were OK. During this, I also notice that R204(which can't be found in the manual's part list for some reasons) is missing. Comparing to the Ch.1 board, I found it to be a 470R resistor, then I installed a 470R resistor into Ch.2 board. Disappointingly, this didn't fix anything. I'm gonna check all the zener voltage then report back.

[shakalnokturn]

This isn't really troubleshooting...
I have a Toward PS of similar design (dual 60V 3A) that had a regulation problem, I can't remember the details but it turned out to be a small electrolytic near an op-amp that was shorted.

[floobydust]

If a part seems missing, was someone else in there trying to fix it? Which resistor did you put in?

The trick to working with these power supplies is to realize the control circuit common is the PSU's (+) output. So all multimeter voltage measurements (when troubleshooting the control circuit) are done with multimeter (-) connected to the PSU (+) output. I know it seems weird but the circuit floats on this.

First check the +9V and -11V rails are OK.
Next, with the PSU defective channel sagging under a small load, measure the voltages on the two IC's U02 (current control) and U03 (voltage control) and post here.

[EE54]

If a part seems missing, was someone else in there trying to fix it? Which resistor did you put in?

The trick to working with these power supplies is to realize the control circuit common is the PSU's (+) output. So all multimeter voltage measurements (when troubleshooting the control circuit) are done with multimeter (-) connected to the PSU (+) output. I know it seems weird but the circuit floats on this.

First check the +9V and -11V rails are OK.
Next, with the PSU defective channel sagging under a small load, measure the voltages on the two IC's U02 (current control) and U03 (voltage control) and post here.

I checked the 2 rails by measuring at Q5 and Q6 emitter respectively, no load on PSU:
 +9V: 9.67V
 -11V:  -11.66V
U2 and U3 measure about the same as well. The measurements were varying in the 100mV range because of the zener diodes tempco and because of my warm room.
Then I set Ch.2 to 10V and connect a 3R9 power resistor and repeated the measurements on Q5 and Q6 emitter and U2 and 3 as well.
+9V rail dropped by about 20mV.
-11V rail dropped by about 2-5mV.

[floobydust]

That looks good, I still need the voltage readings on the IC pins of U03 and U02 when the supply has a light load on it.

These are actually a good vintage linear power supply. I pulled some out of the garbage at work, fixed them and like the aluminium front panel and analog meters.

[EE54]

That looks good, I still need the voltage readings on the IC pins of U03 and U02 when the supply has a light load on it.

These are actually a good vintage linear power supply. I pulled some out of the garbage at work, fixed them and like the aluminium front panel and analog meters.

I'm gonna assume you mean the voltage at pin 6( output pin) of U2 and U3.
I repeated the measurements, Ch.2 set to 10V, current limit maxed out, load is a 15R resistor, which should draw about 6.6W.
 -No load:
  +U2: -3.75V
  +U3: +8.89V
 -With load(15R resistor):   
  +U2: +0.39V
  +U3: +9.15V