TTi QL355P power supply repair

[giosif]

Hi,

I got the above power supply in a rather battered shape and, of course, it is not working.
I have to admit I paid peanuts for it, so I can't complain.
Plus, it's a chance for me to learn (and pull my hair out).
However, whatever failure scenarios I could come up with, none of them was confirmed with testing & measurements.
So, right now, I am a bit stuck and I'm trying to reset my troubleshooting approach.
As such, could someone please help me understand how the part of the circuit in the diagram below is supposed to provide the controlled voltage output?
In other words, what is controlling how much the MOSFET's open?
I think it's Q14, via whatever voltage gets applied to what is connected to its base via PJ14-J, but I'm looking to confirm.
Q13, if I got it right, is used to enable / disable the power supply output altogether.

Thanks!

[Andy Watson]

Looks like a reasonable assumption. You could confirm it by looking at the circuitry that PJ-14J connects to.

[giosif]

Ok, thanks for that!
If this is the case, though, what kind of voltages should I expect at pin 1 of the MOSFET's: positive or negative?
Am tempted to think negative, due to Q17 and Q21 going to -5V, as well as Q14 itself pulling the emitter to -5V, when turned on.

[Andy Watson]

They are N-channel devices. I would  expect to see something in the range of 2 to 10V positive w.r.t. 0V. Q17 is there to limit the current - when the current through source resistors becomes too high, Q17 pulls the gate of the MOSFET down.

[giosif]

Yes, reading a bit more about N-channel MOSFET's, the voltage on the gate should be positive, indeed.

Sorry for the likely silly questions but, if Q17 is for limiting current by pulling down the gate of the MOSFET, why do that against the -5V rail (via R25) and not GND?
Also, what is the role of R21 and R22 in all this? (sorry, nevermind this brain fart)

Thanks!

[Andy Watson]

Sorry for the likely silly questions but, if Q17 is for limiting current by pulling down the gate of the MOSFET, why do that against the -5V rail (via R25) and not GND?

Good question. I'd missed the -5V bit - so clearly not a current limit.
Try this explanation. Q17 and Q21 form a "long-tailed pair" - a differential amplifier. They are there to balance the current between the two mosfets.

[giosif]

Thank you for that, especially the last sentence (let's say I'd have a challenge understanding your response without the last part)! 

As next step, I decided to chase that -4.2V at the base of Q14, and that comes via PJ14-J which connects to PJ18-J on the control PCB, which then goes between two diodes, as in the picture.
So, the negative voltage comes from IC20-B.
Now, the interesting thing is that, in the schematics, R56 is marked with "DO NOT FIT", but it is actually present on the board. 
At the same time, however, I noticed that IC20-B and IC20-A are somewhat in mirror, and R56 has R95 as its counterpart, which is also present on the board.
So, I would be tempted to say the "DO NOT FIT" marking for R56 is just a typo in the schematic.
Would you agree?

 

[giosif]

Referencing the picture from my previous message above, could someone please help me understand the idea behind that section of the circuit?
The two "face to face" op-amps IC20-A and IC20-B seem to be configured as integrators, but I don't really understand for what purpose.
Also, why are diodes D9 and D10 installed with their cathode towards the output of the integrators?
The common annodes of D9 and D10 is connected to the collector of Q13 and is supposed to drive the base of Q14 which, in turn, drives the gate of MOSFET's Q15 and Q16 (please see picture from the OP).

[giosif]

Anyone that could help me with reading this part of the schematic, please?

[coromonadalix]

do you have an complete schematic use a dropbox or an web site tho host it.

I find this psu overthink-ed for nothing,  using power mosfets for voltage regulation, the only advantages is to have a very low  RDS  (resistance drain source) the lower the better, they will dissipate less compared to power transistors.

Mosfets are generally controlled in current, so don't think in voltages all the times. If you psu is max 40volts out, the gates at pin 1 may go up to 10 volts not 40volts, see it like an amplifier, 10v multiplied by 4 ... gives 40v

I do think Q17 and Q21  are the output current limiter for the mosfets ... not necessarily the main current control section, i don't see where p14 goes

Your question about the two facing op amp with diodes : many power supplies with CV an CC are made like this,  one op amp manage the current section and the other is the voltage

The amp limiter section thru the diode will affect the common point of the diodes and can negate or lower down the voltage etc ... depending of what you selected.
The blue line is the final one going to the power stage transistors

But since i don't have the complete schematics  i do guesses.

My guesses is on P21  with the aci and avc line ...

Have you asked the company for the schematics

[Andy Watson]

Ignore the finer details and assume that the voltage present at the base of Q14 represents that outptu voltage of the power supply. The voltage at the base of Q14 is pulled up via R52.
There are two functions performed by the control section: 1. control the output voltage according to a "demand" signal. 2. control the current. Both of these functions are achieved by limiting the output voltage. This is the purpose of the diodes D9 and D10. Whichever control loop demands lower voltage takes precedence and reduces the voltage at the base of Q14.

The capacitors around the feedback paths of the op-amps will have been chosen to optimise the response of the whole control loop. You want a fast response and accurate output voltage but at the same time it must be stable and presumably free of overshoot.

I would imagine that Mr TTi uses the same control module in many different power supply configurations - hence the missing or "not fitted" components on this particular variant.

[Kevin.D]

Yes one is the voltage cntrl Ic20-b and Ic20-A is the current cntrl amplifier. IC21-A with all the compensation
looks to be the feedback amplifier. I seem to have misplaced my full schematics for this series
if you would be kind enough to share the  service manual containing the full schematics for me I will help you
fix yours.

regards

[giosif]

Many thanks, everyone!

Please allow me to read through your replies and digest the information.

In the meantime, about the schematics, TTi were kind enough to send me the full service manual including schematics.
And I would be more than happy to share it here, but I am not sure it would be ok for me to do that, as I don't know what TTi intended when they sent me the manual in the first place (i.e. if they shared it with me as public info or private).
I'll contact them and ask.

[coromonadalix]

if they ask you to not disclose the documents (NDA  : non disclosure agreement), you can share them

[giosif]

Just wanted to provide an update on this: Kevin.D was so kind as to help me with troubleshooting this issue step by step and he rather quickly found the culprit - resistor R65 is completely open, instead of its expected value of 150KOhm.
The interesting thing is the resistor looks spotless, with no signs of burns or cracks or anything else.
Replacement resistors have been ordered and I'm keen to see this working (or realize the resistor issue was not the only problem with this supply).

So, many thanks to everyone who contributed to the discussion and especially to Kevin for helping me out!

Now I need to replay the sequence of steps during this troubleshooting to understand the thought process.
Also, lesson learned for me to be careful with the assumptions made when starting troubleshooting a problem (i.e. a resistor problem in a modern equipment was the last thing I would have thought of as the possible cause for the supply issue).