Quad Op Amp Diagnosis

[Mad Capper]

I am trying to diagnose a refrigerator motherboard and am at a quad op amp LM324. The fridge has 4 fans, one of them, the condenser fan, works properly, but the three evaporator fans do not turn on but work fine when 12VDC is directly applied. The fans have a 3 pin connector, power and sensing, I read the motherboard senses if the fan is spinning.

According to the schematic, attached, it seems, going backward from the connector CN75 each fan goes to a transistor TIP42C then to the quad op amp LM324 and then to the main chip with all the brains of the organization S3P848A.

I get no power at the connector CN75, I have removed the condenser fan transistor and another one from a non-running fan and they both test the same. I have swapped them and everthing works as it did before, so I assume that means they are good. I take from this that the other two probably are working as well. If a transistor is broken would its heatsink get warm or would it be cooler even though it is in the same environment/ambient temperature?

As you can see from the schematic there are a lot of resistors, caps, diodes in between, all visually look ok but I know this can mean little. I can't see testing them all so I am hoping to see if the problem is with the quad op amp.

Then I went for the op amp.

When idle I get 12V on Vcc and the output pins,1,7,8,14,

on the - and + inputs I get for pins :

2 bounces +/-
3 solid +
5 solid -
6 solid -
9 bounces +/-
10 solid +
12 solid -
13 bounces +/-

When the refrigerator call for cooling and kicks on I get:

1 1.58 14 -1.18
2 -4.85 13 -6.65
3 -4.55 12 -6.5
4 4.55 11 -9.9
5 -4.95 10 -7.1
6 -5.08 9 -7.07
7 -0.13 8 -0.93


As I do not know the proper operation of a quad op amp I do not know how to interpret these values. If someone could let me know if it looks as if it is working properly or malfunctioning I would appreciate it.

Is there anything else that I can test or clarify? If the op amp is working correctly would it then seem that the main logic chip is at fault and thus a new mobo is the solution. Knowing the all the evaporator fans work fine on 12VDC is it even possible that they are the cause?

Many thanks for reading and any assistance you can provide.

[eccerr0r]

Re: Quad Op Amp Diagnosis

Actually the op amp is pretty simple used in this manner - it's being used as a comparator in this circuit. Barring any technical details there's a "+" and "-" input. Any time the "+" input is of higher voltage (closer to V+) than the "-" input the output will "amplify" the difference by a very large number being its intrinsic gain. Being used as a comparator and the intrinsic issues of the LM324 the output will hit V+ minus 1.5 volts or so when + > -. When - > +, it will go to GND. There are for discrete amplifiers in the package.

I just noticed there is that 20K feedback resistor so this sort of limits the gain and is "negative feedback" so this limits the gain (since the transistor "inverts" the signal this negative feedback goes into the "+" terminal making it slightly more confusing), but this information should still be valid to tell if the amplfiers are working or not.

Now I'm sort of confused at what the numbers you posted mean as in how they change over time... hopefully this is enough data to debug the problem.

[budm]

Re: Quad Op Amp Diagnosis

When fan is off, you can verify that the transistor which supplies the voltage to fan is OK or not by connecting a 2K resistor between the Base and ground, this should bias on the PNP Transistor.

[Mad Capper]

Re: Quad Op Amp Diagnosis

First thanks for all the input


The difference is that the first set is when the fridge is at idle, i.e. it is not calling for cooling and therefore not turning the fans. The values were only 0.03 or there about but in my ignorance I found it interesting that the positive or negative nature of their value would be so varied, I thought that each amp would display the same pattern which they do not. I was expecting to see one set, that being for the working condenser fan, be distincly different from the other three. Which, when at idle, it is, i.e. pin 5,6 being both solid -.


the second set, for which the formatting apparently was lost, is the voltages on the specific pins when the fridge does call for cooling and the all the fans should be on.

First budm, Don't know if it makes a difference as to what you wrote on the schematic but 3 of the fans do not work not just the one where I labeled the transistor. All the components between the IC's are pretty much the same.

When fan is off, you can verify that the transistor which supplies the voltage to fan is OK or not by connecting a 2K resistor between the Base and ground, this should bias on the PNP Transistor.

I have removed the condenser fan transistor and another one from a non-running fan and they both test the same. I have swapped them and everthing works as it did before, so I assume that means they are good. I take from this that the other two probably are working as well.

Trying to get my head around this. Based on what I already did, is the test you are suggesting going to conclude something different than what swapping the transistor that goes to the working fan with one that goes to a nonworking fan did, i.e. which showed that the one properly working fan still ran when the transistor connected to it was swapped with one from a nonworking fan?




From your first explanation it would see based on my testing is what is occuring. When idle pins 1,7,8,14 are HI, 12V. when cooling is requested they swing low, -1 to 1.5v and would thus as you say turn on the transistor. Don't know how this relates to what I mentioned above about testing the transistors.


For your second explanation about the Diodes is it possible that the 6 diodes for the 3 fans, + and - inputs, are not working and thus not sending any voltage to the op amp and thus nothing to the transistor and nothing to the connector? And that D705 and D706, which lead to the working fan are ok? This would not seem to be the case as the voltage measurements would not be as they were if they did not work, correct?

I am trying to follow all this but heavy going for me.

[budm]

Re: Quad Op Amp Diagnosis

I could not understand your chart.
I only note one example of one fan drive section, they are duplicated with diff resistors value for diff drive speed ofr other drive sections. Once you understand how one section works, you will know how to check the rest.
Instead of swapping transistors, you can force it on to find out if the transistor is working or not. For the fan to work, you have to have +output from the micro controller on the right side of the resistor, the output of the resistor feeds the Anode side of the Diode.

[Tube_Dude]

Re: Quad Op Amp Diagnosis

Something must be wrong, or you had made a mistake in your measures.

Seeing the schematic.

Pin 4 must have 12 Volts from the power supply. (you measured 4,55)

Pin 11 must have 0 Volts (ground). (You measured - 9,9)

[budm]

Re: Quad Op Amp Diagnosis

I am beginng to wonder how he measures all those voltages, and what he uses for ground ref. point. The charts look confusing also.

[Tube_Dude]

Re: Quad Op Amp Diagnosis

I am beginng to wonder how he measures all those voltages, and what he uses for ground ref. point. The charts look confusing also.

A good way is to put one multimeter probe at pin 11 (ground), and just move the other along all the IC pinout.

[budm]

Re: Quad Op Amp Diagnosis

He must not be using the correct ground point since he is showing pin 11 as '11 -9.9'
He should show the reading in this format at least and tell us what was using for Ground:

PIN1 = XX.X V
PIN2 = XX.X V

[Mad Capper]

Re: Quad Op Amp Diagnosis

Yes I am confused too.

Disregard all previous measurements

As you have surmised I was using an incorrect GND. It is pretty tight in there and I have had many an occasion where my less than stellar hand-eye coordination has slipped and shorted things so I tried to use a heat sink as my ground, now I know that just don't work.

any way I measured again with black probe on GND pin 11 and move the other probe around as you mentioned.

When call for cooling and fans should be running

PIN1 = 9
PIN2 = 2.8
PIN3 = 2.7
PIN4 = 12
PIN5 = 2.5
PIN6 = 2.5
PIN7 = 7.6
PIN8 = 9.08
PIN9 = 2.8
PIN10 = 2.8
PIN11 = did not connect both probes as GND
PIN12 = 3.3
PIN13 = 3.2
PIN14 = 8.8

Instead of swapping transistors, you can force it on to find out if the transistor is working or not. For the fan to work, you have to have +output from the micro controller on the right side of the resistor, the output of the resistor feeds the Anode side of the Diode.

Meaning the test you mentioned in post #3? I take it that is with the board all plugged in. If so I would have to tack wires to the pins as there is no way for me to do in place with the board live as there is just no room for me to feel safe. Attached is scan of the board.

Waiting until its cooling cycle shuts down to take new measurement when at idle.

What is very odd is when I just measured the voltages at the CN75 while the cooling cycle is on everything checks out and the fans now work. Voltages at the connector go to 0 when door is open and return to 7-9V when door closed. Sensing voltages stay constant at around 2. I have no idea why this is.

Regardless, do the new voltages that I measured show that the op amp is working correctly?

[budm]

Re: Quad Op Amp Diagnosis

Output pin of the OP-AMP pin 1, 7, 8, 14 have to go low, the lower the number toward zero will make the fans run faster. At 0.6~1Vat the output pin, the PNP transistors will be fully on.
Right now:
Pin 1 = 9V
Pin 7 = 7.6V
Pin 8 = 9.08V
Pin 14 = 8.8V
So the fan that is driven by pin 7 should be spinning. The other may be spining real slow.
You will have to measure the DC voltage on all the fan to see what they are at this condition. The gain of the transistor will affect how hard the transistor will turn on. That is why I suggest to ground the Base of the transistor to make sure the transistors are good.

[rievax_60]

Re: Quad Op Amp Diagnosis

R717 and R713 form a 3:1 negative feedback divider.
The opamp's output will only swing low enough to turn on the transistor enough so that the voltage fed back to its + input matches the voltage applied to its - input.
With 3.3V on pin 12, 3.3 X 3 = 9.9V at the transistor's collector.

extra: The function of the + and - inputs are reversed because the PNP transistor inverts the opamp's output.
When the same voltage is measured on the + and - inputs of an opamp, it usually indicates all is working properly.

[budm]

Re: Quad Op Amp Diagnosis

That is why he needs to check the DC right at the Fan as suggested or at the Collector, at 9V, the fan should be spinning unless the transistor is bad so it does not turn on.

[Mad Capper]

Re: Quad Op Amp Diagnosis

Based on my measurements, which I now believe I have the correct method on, thanks to you all, it would seem to indicate that they only go low as far as the voltages above show, i.e. around the 9.9v mentioned. When the cooling cycle is off these are the voltages


PIN1 = 11.9
PIN2 = 0.01
PIN3 = 0.01
PIN4 = 12
PIN5 = 0.01
PIN6 = 0.01
PIN7 = 11
PIN8 = 11.9
PIN9 = 0.01
PIN10 = 0.01
PIN11 = did not connect both probes as GND
PIN12 = 0.01
PIN13 = 0.01
PIN14 = 11.9

and the voltages on my the CN75 connectors for output to fan and sensing pins are all 0 when at idle, which I presume is the way it should be since they should not be running.

That is why he needs to check the DC right at the Fan as suggested or at the Collector

Would I be correct to say that instead of at the fan or between the transistors collector and ground, I can check at the CN75 connector, which is where the fans wires connect to the board? If so then as I mentioned I have done so and get this

CN75
Pin 1 GND
Pin 2 9.9 (Freezer fan)
Pin 3 8.1 (Refrigerator fan)
Pin 4 7.57 (Condenser fan)
Pin 5 2.25 (Fan on sensor)
Pin 6 2.2 (Fan on sensor)
Pin 7 2.1 (Fan on sensor)

and on CN76(only two pins on mine as it is a slightly different model than the one the schematic is for)
Pin 1 8.46 (2nd Freezer fan)
Pin 2 2.2 (Fan on sensor)



All the fans run during the cooling cycle. At this point I am happy things are working and grateful for the assitance everyone provided. Hopefully it continues working and if not I will step back to the transistor and somehow get in the confined space to test at the collector for the 9V.

Now I know a little about op amps and thank you for sharing your vast knowledge, as my knowledge is only half vast.

[budm]

Re: Quad Op Amp Diagnosis

CN75 PIN 2, 3, and 4 are the voltage from the Collectors that feed the all 3 fan RED power wires connected to the CN75.