Bench PS Problem (Volteq HY3010D)

[orientalsniper]

Re: Bench PS Problem (Volteq HY3010D)

Hello, thanks to you for sticking with me so far.

As I said in post #37, pin 2 (middle) doesn't even go down to 0V, lowest was 51.2V.

EDIT: I meant middle pin.

[eccerr0r]

Re: Bench PS Problem (Volteq HY3010D)

Okay, that's very suspicious, reason why I ask to confirm is that you should be able to adjust it to near 0V or at least drop the voltage by around two thirds at the lowest setting. I'd replace that pot but the numbers you gave me either do not match up with the schematic or indicate there's more to it that I don't know. Photograph of the back of the pots? Or is it just plain potentiometers with wire soldered to them?

How hard is it to remove the pot, is it a lone pot or on a PCB?

(Also double check you're testing the voltage pots, not the current pots.)

What is your spare parts kit like... can you fake a pot with hardwire resistors as a test if you don't have another pot to sub?

The next experiment I really would like to see you try is to remove the LM324 (or disable the relays from clicking without removing their contact connections) and simply short that RP7 wiper pin to GND. This theoretically should force the PSU to drop its voltage down. Before putting a fresh new LM741, ensure all the pins are below ~24V with respect to GND so we don't fry another one.

[orientalsniper]

Re: Bench PS Problem (Volteq HY3010D)

I think the first thing I did was to swap the "current" potentiometers with the "voltage" ones, I'm about 50% sure about that, maybe I dreamed it up.

The pots are just soldered on a single PCB for the potentiometers.

I have plenty of spare parts (dead motherboards, ATX PSU, etc), my IC list is limited though.

Do I need to remove LM324 before shorting RP7 wiper? I shorted it without removing it and the displayed voltage was 18V, however since it's shorted I'm getting 0V with the multimeter.

[eccerr0r]

Re: Bench PS Problem (Volteq HY3010D)

The reasoning for removing the LM324 is to both limit the scope of the test so we know what's going on, and also to reduce the amount of damage that can happen.

Well, since you tested without removing it, okay, damage is done if there was any though it should have been okay.

So we're still dealing with multiple failures here since you can't get the center pin of RP7 to 0V no matter what the setting of it and RP6, and shorting the + input of the LM741 to GND does not reduce the output voltage to 0V... hmm...

---

Now that you understand why I want you to remove the LM324/disable the relays, do so, and measure the voltages for each of the pins around the M1 LM741. If they are all below 24V (meaning, no more 38 to 51 volts anywhere), put in a new one and see if shorting the center RP7 wiper pin to GND will reduce the output to 0V.

Another test if you have a known defective or original LM741 in the circuit is to short its pin 6 to its pin 4, and see if the voltage output of the PSU will drop to 0. This can destroy the LM741 if it isn't already dead.

[orientalsniper]

Re: Bench PS Problem (Volteq HY3010D)

Well, RP7 wiper is technically 0V read with a multimeter when I short the wiper to GND.


Do you think I should wait and replace all the IC's?

[eccerr0r]

Re: Bench PS Problem (Volteq HY3010D)

It should be able to adjust to 0V without shorting -- that's what's bothering me!

What do you mean wait...thought you got the replacements already?

I don't know what your toolbench is like to remove/disable these ICs for testing... that's something that needs to be asked to reduce work on your side to help rootcause the failure. Otherwise need to think about other methods to get to the problem.

[orientalsniper]

Re: Bench PS Problem (Volteq HY3010D)

Nah, it's gonna take 2-3 weeks for the delivery, I will replace as soon as I receive them and will update thread, if it's not the IC's, hopefully you will still be sticking around.


My toolbench: 1 oscilloscope (don't know how to use it yet) - 1 hot air rework station (soldering + hot air) - 1 Fluke multimeter - 1 lab power supply (the thing we are fixing now) - 1 wireless digital microscope (on the way)

[eccerr0r]

Re: Bench PS Problem (Volteq HY3010D)

I'm a bit confused about #33 in which you said you swapped, unless you tried putting the M2 with M1, was that what you meant?

Do you have sockets or did you order these as well? Also, experience with desoldering DIPs from PCBs?

Might be worth scouring your spare parts pile to see if you have any LM741s to steal in the meantime, and hope you don't get fakes from china

[orientalsniper]

Re: Bench PS Problem (Volteq HY3010D)

Yeah, I meant swapping M2 with M1 lol, otherwise I would have said "replaced" but I can see I got you confused


What are sockets? You mean the potentiometers?


I've got experience soldering/desoldering DIP-SMD-BGA (the latter with no much success soldering since I lack the correct equipment)

[eccerr0r]

Re: Bench PS Problem (Volteq HY3010D)

Ah okay... I can't be certain that the swap also killed the other 741 if it had been working, hence asking the 324 be removed in order to protect the 741 - and so that we can do some experiments. The 324 is used to save power.

I was wondering if you had sockets so removing the ICs and plugging them back in would be easier when the new ones arrive. Especially the M1 741: this one I can see it failing often.

Okay that's good that you can fairly easily desolder these ICs without damage to the board. At least this board is a single sided, which makes things easier.

Looking at the data for the LM741 you should be able to short the M1 output (pin6) to one of the rails (say, pin 4 for the test I am curious in) to see if the voltage on the PSU output will drop to zero without damage to the IC. I'd like to see this experiment done. If you can do this with M1 out of circuit, that would be okay too, then again I would hope the voltage output should be 0 with M1 removed. You can also short the anode (the non-banded side) of diode V16 to M1's pin 4 if you'd like instead, both of these theoretically should drop the PSU output voltage to 0.

All things are still pointing at the pots and/or M1 though.

[redwire]

Re: Bench PS Problem (Volteq HY3010D)

Use the power supply (+) output as your common point [multimeter (-)] for making voltage measurements on the control circuit.
Do not use the power supply (-) output as multimeter (-) it will drive you nuts.

The 2N3055 pass transistors and drivers V13, V14 are normally full on.
R4 and zener V11 bias them on, and the voltage M1/current M2 control op-amps turn the transistors OFF.
So both out-of-socket, the transistors will stay full on.

I would pull M2 (disable current control) and leave M1 in and measure voltages on all its 8 pins, WRT output(+).

A problem with these power supplies is the O/P goes straight into the op-amp M1- which leaves it vulnerable. The op-amp gets damaged if the PSU output gets stuck on, with RP7 set to minimum.
So it will continue to eat op-amps.

I have this schematic for a HY-3010D.

[kaboom]

Re: Bench PS Problem (Volteq HY3010D)

check 3 opamp supplies, using out + for DMM negative lead:

+12V at pin 8 of M1 & M2, pin 4 of M4
-15V at pin 11 of M4
- 6.2V at pin 4 of M1 & M2

also check at junction of R2:V6 and R39:V21

you must use out + as the common reference- see schematic

we cannot assume anything about the opamps til it's understood they've got both their pos and neg supplies.

[eccerr0r]

Re: Bench PS Problem (Volteq HY3010D)

O/P has already verified the auxiliary PSU is working by the measurements done on M1 and M2 in #21.

Using PSU out+ as a reference is also reasonable, but really it doesn't matter where your reference is as long as you know what it is. Subtract out the 38V that's on pin 2 as this is connected to out+ from the voltages #21 gives the expected voltages based on the voltages listed...except that pin 2 should be changeable from 0 to ??? volts - indicating either a shorted/destroyed input caused by fast voltage change downwards, or the pot is broken somehow.

---

Another possible mod is to put a resistor in pin 3's path, and use two inverse series 5.1V zener diodes to clamp the input to pin 2. Actually, maybe don't even need zener diodes - two inverse parallel silicon diodes (or even LEDs) should technically do the trick too - just need to make sure that resistor is there to soak up the current that would flow through these diodes. This should be sufficient prevent the slew from damaging the LM741.

[redwire]

Re: Bench PS Problem (Volteq HY3010D)

Yes - the PSU can keep blowing the voltage control op-amp until some protective resistance and clamp diodes added, or the stuck output is fixed.

If OP's LM741 (M1) shorted, it can damage the voltage setpoint potentiometer or voltage divider trimpots. I would check RV7.

I add a pair of 1N4148's across the op-amp +/- inputs and a couple series 10k, one on each op-amp input.
Another protection diode across C-E of the pass transistors is needed but that is another discussion, if you ever charge batteries.

I would re-test the pass transistors and drivers. Shorting E-B on driver V13 should give zero output voltage.

Another weirdness about these power supplies is the voltage control pot/divider string goes from VREF to VOUT(-), it's not intuitive.

[eccerr0r]

Re: Bench PS Problem (Volteq HY3010D)

Staring at the schematic I still don't get how it can maintain regulation... VREF is ~2.5V over whatever the current output voltage is... So it appears to be comparing VOUT with (VOUT + 2.5)*(Frac*RP7+RP6)/(R16+RP2+RP6+RP7) or something like that. whiskey tango foxtrot, how stable can this be or am I missing something here?

Someone needs to give me one of these pieces of crap so I can play with it!

[eccerr0r]

Re: Bench PS Problem (Volteq HY3010D)

Any further words on the experiments while waiting for parts?

[orientalsniper]

Re: Bench PS Problem (Volteq HY3010D)

Hey, thanks for the replies and sorry for the delay, I need to fix some laptops first, need to clear my desk.

So for all my tests I have to use (+) lead of multimeter on (-) output of power supply? Got it! Will do the tests by the end of the day.

[eccerr0r]

Re: Bench PS Problem (Volteq HY3010D)

Well, it was suggested the black input on the POSITIVE output of the power supply. But it really doesn't matter - as long as if you tell us where both probes are connected, then we can figure out the status of the supply.

The current outstanding tests/experiments are:
1. Still not certain about this: with the M1 µA741CN removed or the wiper/center pin disconnected from RP7, and RP6 set to the lowest setting: can you adjust the voltage on that center RP7 potentiometer pin through the whole range by turning RP7? Ideally, volt meter negative pin on PSU negative output as reference.

If result: pin2 of the potentiometer can now go through a large swing of voltages and even hit 0V. Then likely M1 LM741 is indeed shorted and the potentiometers are fine.
If result: pin2 is stuck at a few voltages and cannot swing more than a few volts. Then need to repair/replace potentiometers.

2. The two shorting experiments: either
a) Can you short V13 transistor's B to E (tests pass transistor tree for shut of capability)
or
b) short the anode of V16 to one of these three points: (tests shut off capability in general)
- pin 4 of M1
- pin 4 of M2
- anode of V6

If result: Voltage on the +PSU output will drop to 0 or at least very low.
Provided the diodes are good, this would strongly indicate M1 needing to be replaced if it's the only problem.
If result Voltage on the +PSU output stays the same.
This means one of the pass transistors have shorted or stuck on for whatever reason.

Oh boy just noticed V16 is an unpop too... need to see how the heck this thing is connected. Wow...

[orientalsniper]

Re: Bench PS Problem (Volteq HY3010D)

I'm overwhelmed, I was about to try post #51

Removed M2, put M1 back in and then turned on power supply, then I noticed N3 (L7812CV) giving off a lot of heat, it wasn't smoke, I'm pretty sure it was burning the glue in the heatsink.

[eccerr0r]

Re: Bench PS Problem (Volteq HY3010D)

Has it always been getting warm? Is this a new symptom?

Is it still giving 12V out (measure its output with your black meter lead on the POS output of the PSU)?

This regulator would dissipate the most power if all three relays are on.