Wine and Beverage Cooler Control Board Repair Attempt

[KYBOSH]

Resistance does not turn on the ic, you need to check the Voltage on those pins, I suspect pin 3 will be high indicating that driver is turned ON causing JK1 to be activated

Sorry about that... i thought we were looking for an internal short of some sort in IC1.
So when the board is not plugged into mains there is no continuity on any of the 4 pins as I mentioned before.

But when the board is plugged in and activated you get 0V (actually shorted to ground) on pins 1 and 2 then 1.75V on pin3.

[R_J]

when i measure the resistance as you instructed i get ~14.8 kohms on pins 1. 2 and 3 (against pin4 ground)

Resistance does not turn on the ic, you need to check the Voltage on those pins, I suspect pin 3 will be high indicating that driver is turned ON causing JK1 to be activated

[KYBOSH]

Check U3 pins 1,2 & 3 using pin4 ground, that will tell you the state of the outputs from the micro ic1 It may be that when ic3 shorted it damage ic1 port

when i measure the resistance as you instructed i get ~14.8 kohms on pins 1. 2 and 3 (against pin4 ground)

[R_J]

Check U3 pins 1,2 & 3 using pin4 ground, that will tell you the state of the outputs from the micro ic1 It may be that when ic3 shorted it damage ic1 port

[KYBOSH]

This is from post 20, if this is true then you have a broken trace between the +12v on JK1 and where it connects to JK2 then on to JK3. Look at the circuit traces, all 3 relays have one pin connected to +12 volts and the other pin is connected to IC3 which controls the relays. Plug X6 should have +12dc on pin1

Why keep checking the power supply? it is working! The diodes capacitors etc. are all working there is no need to replace them, you will just end up destroying the power supply section.

The so called crack you see is just a plastic molding mark

Wow.. Thank you very much R_J.
Your diagram of my PCB is better than what I am doing.
I have to step my game up!

Regarding checking and rechecking components on the board I wanted to ensure that I am thorough as possible this 2nd time around (recording readings, etc) and that i did not miss anything.
My cursory checks are often with components in circuit. Or if i ddi take them out i dont write down my findings.
Redwire also mentioned that everything can look good on the bench but when the board is put underload it fails.
I have already found a bad cap which i have overlooked today.
But there is truth in what you advice... constant soldering and desoldering runs the risk of ruining the board.
Once measurements are recorded i will not have to take them out of circuit again if they are found to be good.

Regarding the +12V traces and positive leads on the board all have continuity as I just ran down to the bench and checked.
When i checked whether they had voltage or not I checked the + and - leads of each of relay.
The UN2001D driver (IC3), for some reason, is not providing ground for JK2 and JK3. I believe this signal to provide ground should be on by default or come from the main IC (IC1).
If not on by default it must be waiting for a signal/trigger from one of the connected accessories.
But as I mentioned there is 120V going to the internal heater coil as soon as the board is connected to the mains (this is connected to the JK1 relay which is powered on as automatically).
That seems wrong to me...

At no time during my testing have i seen JK2 or JK3 cut on.

[R_J]

I checked for voltage going to the relays.
JK1 has 11.5v going to it, while the others (JK2 and JK3) have none.

This is from post 20, if this is true then you have a broken trace between the +12v on JK1 and where it connects to JK2 then on to JK3. Look at the circuit traces, all 3 relays have one pin connected to +12 volts and the other pin is connected to IC3 which controls the relays. Plug X6 should have +12dc on pin1

Why keep checking the power supply? it is working! The diodes capacitors etc. are all working there is no need to replace them, you will just end up destroying the power supply section.

The so called crack you see is just a plastic molding mark

[KYBOSH]

I'm turning my attention back to the D7 diode next to the transformer.
It caught my eye today bcos it looked like it had a small crack on the surface.
It was always giving weird readings when i tested it with the multimeter but I attribute that to the caps right next door.
When I test it out of circuit I recall it read okay.



Hard to make out the markings but i think its a STPS 3150 which is a 150 V, 3 A power Schottky rectifier.

I dont have a direct replacement so I'm researching which one of the diodes i have maybe a good substitute for it.

[KYBOSH]

Those values look OK but EC6 is a bit rounded top looking. You have +5V and +12V power but the board still acts dead? Did you test the secondary caps as well? The two large ones.
I found it as Zephyr Wine Cooler Control Circuit Board GB-KB3100-MAIN-V3.1
Main Power Control Board 11010114 ripoff pricing

Yes! The board is very expensive for what it is.
This is the reason why these appliances are discarded rather than repaired and that's a pure shame.
But the going price of these models are over $1000US (new) locally so it's worth repairing in my opinion.
At the VERY least its worth the effort to rectify the defective board if possible.

The cap that you say is bulging really isn't bulging much, if at all, and the appearance is really more a product of the camera angle and lighting but to make you happy RW I am replacing it with one that has a completely flat top and a capacitance of 9.76µF and an ESR or 2.1Ω.

I tested the secondary electrolytic caps and this is what I got:


Yes, EC11 was very bad. I changed all 4 of these caps in the group with caps that averaged 105µF and 0.61Ω ESR.
I made EC13 a 100µF 50V bcos i didnt have a stock replacement with long enough legs to bend.

I also checked the last relay JK3 and it tested good.

When I plugged up board it STILL DIDNT fire-up.
Same behavior.

[redwire]

Those values look OK but EC6 is a bit rounded top looking. You have +5V and +12V power but the board still acts dead? Did you test the secondary caps as well? The two large ones.
I found it as Zephyr Wine Cooler Control Circuit Board GB-KB3100-MAIN-V3.1
Main Power Control Board 11010114 ripoff pricing

[KYBOSH]

You might be on a goose chase - I can see two (snubber) capacitors that are likely across the relay contacts, to suppress arcing on them.
So a multimeter will read ACV from the leakage current.

These boards usually collapse due to the electrolytic capacitors failing. They are cheap to begin with and run 24/7. I would test on the primary side, some in the pic look like rounded tops? Like EC6. Carefully discharge them if needed.
The rails might look ok but fizzle out with say the LED loading.

Thanks for the guidance RW. I thought I was all alone on an island for a second!
I followed your lead and tested the caps on the primary side for capacitance and ESR (for the electrolytics).
This is what I found out (ratings on top/measured readings on bottom):



The electrolytic caps are manufactured by LSCON. The datasheet (which was a little hard to find) says that they have a tolerance of ±20% so we are well within spec on these plus the ESRs are pretty low at these capacitances.
I dont have a pair of replacement 10uf/400V caps on hand (just 1) but I will likely replace them with a better brand just to rule them out as the problem. As you say they may look good but underload they are failing.
The "snubber" capacitors are also looking really good too. I see no reason to replace them unless you advise otherwise.

[redwire]

You might be on a goose chase - I can see two (snubber) capacitors that are likely across the relay contacts, to suppress arcing on them.
So a multimeter will read ACV from the leakage current.

These boards usually collapse due to the electrolytic capacitors failing. They are cheap to begin with and run 24/7. I would test on the primary side, some in the pic look like rounded tops? Like EC6. Carefully discharge them if needed.
The rails might look ok but fizzle out with say the LED loading.

[KYBOSH]

I went back to track what the wires were for the AC connectors. The one with 120VAC and 1.8VAC on when the PCB is on my bench seem to be the heaters. Makes sense since they lead into the fridge unit.
This is wrong. The heater should not be on by default but rather the compressor. I tested the relay labelled JK2 out of circuit and it tested good. I will check JK3 tonight to see how it fares.

Looks like my issue is indeed with the board as its not powering what it should be powering in its default state (off the appliance).

[KYBOSH]

Gotta a question...

The scheme here shows a "transformer" being connected to the Control PCB .... Does this mean there is an external transformer somewhere on this fridge? Not the little transformer (T1) in the board....

[KYBOSH]

Thinking this over I realized that the transformer was good as it was the one supplying the lower voltages to the board and a secondary with fewer winding will give a low-Q reading on the Ring Tester.
The next thing I did last night was check the octocoupler to ensure that it was good.
When I tested leads 1 and 2 with my multimeter in diode mode it tested good.
I checked leads 3 and 4 and it was giving a strange reading. Its a Sharp PC817.
I pulled it out of circuit and tested it again but it read good. The readings I was getting was from the lines it was connected to.
I put the OC back in circuit and decided decided to check if the connectors had any voltages across them and I was pleasantly surprised to see I was getting voltages to all of them.




So apparently the board is working for the most part. The 120VAC goes inside the fridge and not directly to the compressor. I am not sure where it goes or what its supposed to be powering though.

The next thing I am going to test is the Start Relay for the compressor. It maybe bad and it may act as a switch that will shut the whole system down if it its.
Im now working off the assumption that something off the board is bad and causing it to shout down. I tried pulling connectors off 1x1 in hopes that disconnecting the offending component would let the rest of the system start up but that didnt seem to add any clues.

Again the issue i am currently having now is the system never powers on. It makes a short beep (solid beep rather than a fading one) when plugged in but no sign of life at all.


Edit to add: The resistance of the starter relay is 5.3Ω which is said to be GOOD so its not the relay.

[KYBOSH]

Thanks for the tip! I studied up on decoupling capacitors last night.
You are correct by the way: 105 is code for 1uF. I think i change out C8 and C10 for fresh ones.
The other ones seem to be rated for 100 nF so I will swap out C2 and C15 for fresh ones as well.

I'll check for voltage after I get those sorted.

Correction: C15 is labelled 102 (1nF) and not 100 nF as i previously mentioned above.

[KYBOSH]

I started taking a look at the big transformer. I measured 162v going into the primary windings but nothing coming out. Grabbed my blue ring tester and checked it. Primary side looks very good. Green lights/ Hi Q. Primary side gives me 2 red lights which is meant to mean bad or low Q. The marking in this transformer is EF20. Not sure what its specifications are...couud be 24v?

[KYBOSH]

C8's value is written under it, observe the same text above C9. "105" so a 1uF decoupling cap I think.
Check if you have DC power 12V (for relays) and 5V for the MCU from Vreg? U1.

So it took me a while to get my book of assorted diodes and caps in but i finally got it and replaced the questionable caps.
Still no joy though. Cant figure out what's wrong with this board. I tested the resistors again. Some where reading low on the board but good when taken out of circuit.

I checked for voltage going to the relays.
JK1 has 11.5v going to it, while the others (JK2 and JK3) have none.
Two things funny about JK1 (JZC-32F). When I looked at it with the FLIR it was markedly warmer than the rest of the board.
I thought it might be shorted but then I took it out of circuit and tested it test fine (although in retrospect i tested it at 30V rather than the nominal 12v... will go back and check it again).
Next I checked U1 for voltage.
At first the readings were dodgy but after touching up the solder joints i was getting a solid 11.52 and 5.02 between pins.

Not sure where to go from here....

Of note, when I was testing the electrolytic capacitors next to the small transformer i mistakenly put one in backwards.
When i installed the board back unto the fridge the control panel inside started lighting up but it was blinking off and on.
This went on for a few seconds until i unplugged the unit to have a look at the board again and I immediately noticed the mistake I made.
So this sign of life is good news. I just have to figure out what I am missing. Doesnt seem to be a short but I cant find the bad component.
It could be a fault IC/Controller but I dont know how to test for that.

edit to add:
just checked that relay and applied 11v to it and it works just fine.

[KYBOSH]

Some of the boards out of china have part values written on them, here and there. Usually resistors but the oddball (bigger) caps standout as well.

Thanks for the tip! I studied up on decoupling capacitors last night.
You are correct by the way: 105 is code for 1uF. I think i change out C8 and C10 for fresh ones.
The other ones seem to be rated for 100 nF so I will swap out C2 and C15 for fresh ones as well.

I'll check for voltage after I get those sorted.

[redwire]

Some of the boards out of china have part values written on them, here and there. Usually resistors but the oddball (bigger) caps standout as well.

[KYBOSH]

C8's value is written under it, observe the same text above C9. "105" so a 1uF decoupling cap I think.
Check if you have DC power 12V (for relays) and 5V for the MCU from Vreg? U1.

What a delightful tip! I did not realize that those numbers were the values for the SMDs!
I cant remember ever seeing values printed on the board like that.