Cyberpower CP900EPFCLCD - F03 Error

[Shinjinkun]

Re: Cyberpower CP900EPFCLCD - F03 Error

It often happens that together with PWM (Viper, ... Mosfet ...) a pulse transformer burns out (turn-to-turn short circuit of the power winding) and when a new PWM is installed, it burns out again.

But is there a way to check that this is happening? I mean, I only have like 5 or 6 vipers left... I won't like to completely waste them if I can avoid it...
How can I isolate the hot area to confirm that there are no shorts in the cold area?
Is there a way to confirm that the viper is working, even outside the motherboard?

[lotas]

Re: Cyberpower CP900EPFCLCD - F03 Error

It often happens that together with PWM (Viper, ... Mosfet ...) a pulse transformer burns out (turn-to-turn short circuit of the power winding) and when a new PWM is installed, it burns out again.

[Shinjinkun]

Re: Cyberpower CP900EPFCLCD - F03 Error

Yes replace and try again.
Keep one thing in mind: the closer you are to the short the lower the voltage will be.
The shorted component might also heat up more than the others...

I replaced the components, included the viper, and connected everything again. However, I am back to step one, the short is present and I cannot get a way to find which component is causing it.

Do you have any recommendation on how to proceed? I am really stuck at the moment, even after watching loads of SMPS videos and reading several datasheets.

I'd really appreciate any suggestion on what to do.

Thanks again for all your help.

[Per Hansson]

Re: Cyberpower CP900EPFCLCD - F03 Error

Well, that must be the current issue, indeed.


I removed the bridge rectifier and, when I connected everything to mains, the light bulb was off. Then measured the voltage. This time, I was getting 230.8V into both inputs for the DB107 bridge rectifier. I suppose that means that everything before it should be working fine, right?

Yup

I disconnected it and thought an idea. I removed the viper22a from the socket, and got a few more measures. First, the bulb was off (probably because part of the circuit was not accessible). The DC output from the rectifier had 316.2V DC. The C3 capacitor measured 316.2V, and C14 between 40 to 50V, but couldn't write it down cause while using the probes, I must have shorted something, and fried the 550ohm resistor and C14 probably. Probably it wasn't a very good idea removing the viper, but at least isolates a bit where the short is.

Then, from now on, what do you recommend me to do? Should I replace the blown components and try again, or remove some of them to try to find the faulty section?

Thanks

Yes replace and try again.
Keep one thing in mind: the closer you are to the short the lower the voltage will be.
The shorted component might also heat up more than the others...

[Shinjinkun]

Re: Cyberpower CP900EPFCLCD - F03 Error

From the photo it looks like you are soldering with soldering acid, and it conducts electricity, you need to use rosin or acid-free flux.

I am using this flux: https://www.jbctools.com/fl-15-flux-...oduct-150.html

Not sure if it has acid or whatnot.

[lotas]

Re: Cyberpower CP900EPFCLCD - F03 Error

From the photo it looks like you are soldering with soldering acid, and it conducts electricity, you need to use rosin or acid-free flux.

[Shinjinkun]

Re: Cyberpower CP900EPFCLCD - F03 Error

Ok, so with these values it looks to me like your incoming line voltage is 233VAC
Since you measured 227v in series with the bulb this is the voltage it drops.
So therefore your measurements of 5VAC show the remaining AC (not being dropped) i.e. it points to the short.

Well, that must be the current issue, indeed.

In this case you are reading a very low voltage on the bridge rectifier of only 3.96VAC and 5.361VAC in C22 before it.
This points to the bridge rectifier or something after it being shorted.
I would remove the bridge rectifier from the board and test both again:
This will help narrow down where the issue lies.

I removed the bridge rectifier and, when I connected everything to mains, the light bulb was off. Then measured the voltage. This time, I was getting 230.8V into both inputs for the DB107 bridge rectifier. I suppose that means that everything before it should be working fine, right?

As for the voltage difference in C5 and C6 that is because one is connected from phase to neutral (230vac side).
And the other between neutral and ground side (this should in a perfect world read as 0vac but a low voltage like you see here is normal).

Ahh, understood. I didn't know that was the correct value I should get.

As for K1 and the capacitor C7 I don't see how they are connected but I would ignore them for now, the goal is to get power to the Viper chip first.
In any case I'm 99% sure that the voltage generated by the Viper is used to control the relay so there is no point going down that path (yet)

Alright, let's check the right part later on.


It took me some time to respond cause I was about to receive a package with the fuse replacements and some extra DB107. I replaced the DB107 with a new one, and when I connected everything again, the light bulb was on, and the values back like last time.

I disconnected it and thought an idea. I removed the viper22a from the socket, and got a few more measures. First, the bulb was off (probably because part of the circuit was not accessible). The DC output from the rectifier had 316.2V DC. The C3 capacitor measured 316.2V, and C14 between 40 to 50V, but couldn't write it down cause while using the probes, I must have shorted something, and fried the 550ohm resistor and C14 probably. Probably it wasn't a very good idea removing the viper, but at least isolates a bit where the short is.

Then, from now on, what do you recommend me to do? Should I replace the blown components and try again, or remove some of them to try to find the faulty section?

Thanks

[Per Hansson]

Re: Cyberpower CP900EPFCLCD - F03 Error

Ok, so with these values it looks to me like your incoming line voltage is 233VAC
Since you measured 227v in series with the bulb this is the voltage it drops.
So therefore your measurements of 5VAC show the remaining AC (not being dropped) i.e. it points to the short.

In this case you are reading a very low voltage on the bridge rectifier of only 3.96VAC and 5.361VAC in C22 before it.
This points to the bridge rectifier or something after it being shorted.
I would remove the bridge rectifier from the board and test both again:
This will help narrow down where the issue lies.

As for the voltage difference in C5 and C6 that is because one is connected from phase to neutral (230vac side).
And the other between neutral and ground side (this should in a perfect world read as 0vac but a low voltage like you see here is normal).

As for K1 and the capacitor C7 I don't see how they are connected but I would ignore them for now, the goal is to get power to the Viper chip first.
In any case I'm 99% sure that the voltage generated by the Viper is used to control the relay so there is no point going down that path (yet)

[Shinjinkun]

Re: Cyberpower CP900EPFCLCD - F03 Error

Yes, just plug it in now, the lamp is allowed to flash once bright as the caps charge.
After that it should be quite dim.
If it is bright then it is expected something is still shorted.

Sorry it took me some time to make the proper tests.

Well, I finally connected the lamp and, as expected, is continuously bright. At least, this way, is not burning the components again.

As I wasn't sure about what to do, I took several measures of voltage in different points. Please check the attached pictures.

I would like to point out a few things:

First, I found only 0.7V in the ceramic capacitor C6 (AC472M), while in C5 that is exactly the same type, found 232V.

Then, the film capacitor C22 measures only 5.37V. Shouldn't it be also 230V?

In the end of this branch, the AC/DC converter inputs 3.96V AC and outputs 2.019V DC.

Going the other way, I am not sure how to test the voltages for the relay K1.

But I found that C7, the 1.0uF X1 CTX film capacitor measures as a short. Could it be that this component is broken, even that when I measure it disconnect I don't get the short?

Please keep helping me to try to fix this!

Thanks a lot

[Per Hansson]

Re: Cyberpower CP900EPFCLCD - F03 Error

Yes, just plug it in now, the lamp is allowed to flash once bright as the caps charge.
After that it should be quite dim.
If it is bright then it is expected something is still shorted.

[Shinjinkun]

Re: Cyberpower CP900EPFCLCD - F03 Error

Hi guys!

I replaced all the components again just to be sure. When I removed the 550Ω, I checked it and it was in short, so I replaced it.

Regarding the lamp holder, I attached the contacts to the fuse, as it is open anyway. Attached the pictures for some reference.

So now, should I connect everything again and test all the components? Which steps should I follow?

Thanks

[Shinjinkun]

Re: Cyberpower CP900EPFCLCD - F03 Error

The "T" in "TAPc" could stand for "Time Delay" i.e. a slow blow fuse.
But none of those letters really make sense so could just be production numbers...
If you try with a quick blow fuse you don't risk anything.
I.e. you can always switch to a slow blow instead if it blows.
But of course start with the incandescent lamp: that is the most important safety test really...

Yes, I will use the incandescent lamp first, as it will take several days or weeks to get the fuse replacement. Problems of living in a small village. I will order the replacement as soon as possible.

Correct, you can also search Youtube or Google for some bridge rectifier test videos, in total there are 4 diodes to test...

I watched two or three of these videos, and it looks like the BD1 rectifier is in order.

As for the optocoupler if you can't read anything it is probably broken.
So desolder it and the 550ohm resistor from the board and test again.

Actually, I tested U3 again, and I get 1.082V between anode to cathode. It seems that a small circuit and a battery are needed to fully test the optocoupler, so I may have to trust it for the moment.
In the end, I will desold the resistor and put a replacement.

I will let you know once I get the lamp holder solded.

Thanks

[Per Hansson]

Re: Cyberpower CP900EPFCLCD - F03 Error

The "T" in "TAPc" could stand for "Time Delay" i.e. a slow blow fuse.
But none of those letters really make sense so could just be production numbers...
If you try with a quick blow fuse you don't risk anything.
I.e. you can always switch to a slow blow instead if it blows.
But of course start with the incandescent lamp: that is the most important safety test really...

The bridge rectifier is the BD1 component (BD107G), right? I measured it, without removing from motherboard, with the negative in the AC inputs, and was getting 0.595V in diode mode. The same putting the negative in the positive DC, so I am not sure if this means it works.
However, the U3 component doesn't give me any readings. I didn't remove it from the board either.

Correct, you can also search Youtube or Google for some bridge rectifier test videos, in total there are 4 diodes to test...
As for the optocoupler if you can't read anything it is probably broken.
So desolder it and the 550ohm resistor from the board and test again.

[lotas]

Re: Cyberpower CP900EPFCLCD - F03 Error

Yes, put any size on 2A.
https://aliexpress.ru/item/400006248...00000159518588

[Shinjinkun]

Re: Cyberpower CP900EPFCLCD - F03 Error

I removed the fuse from the board, but it doesn't clarify much. Attached some pictures for you to see.

The markings on the fuse are like this:

RSFR-H on the self-shrinking tube.

2A 250V | in one side

TAP c UL US | on the other side

I have been looking for hours to see if I can find the same kind of fuse, but no luck so far. Do you have any idea about which one should work?

[Shinjinkun]

Re: Cyberpower CP900EPFCLCD - F03 Error

Yes exactly, just connect it where the fuse was before.

Perfect. I will get a lamp holder tomorrow. I remember I have one in the garage.

Cut away the heatshrink tubing from the old fuse and read what it says on the ends: there are fast and slow fuses and it is important to pick the right type...

Alrighty. I hope the fuse has some print on it. Otherwise, is going to be difficult to pick the fuse.

Yes some parts can be tested, like diode check the bridge rectifier for example.
I would also check the 550ohm resistor and optocoupler U3, this can be done with diode check too.
As for the rest as long as you use the lamp as a current limiter there is not that much that can go wrong...

The bridge rectifier is the BD1 component (BD107G), right? I measured it, without removing from motherboard, with the negative in the AC inputs, and was getting 0.595V in diode mode. The same putting the negative in the positive DC, so I am not sure if this means it works.
However, the U3 component doesn't give me any readings. I didn't remove it from the board either.

[Per Hansson]

Re: Cyberpower CP900EPFCLCD - F03 Error

Ohh, I heard about the light bulb thing before, but I am not sure how to do it.
So, if I understand correctly, I will need a lamp holder (or bulb socket, not sure the correct name) with two cables. Where should I connect these two cables? To the same positions the broken fuse are connected? Or somewhere else?

Yes exactly, just connect it where the fuse was before.

Regarding the replacement fuse, I understand that it must be a 250V 2A one, fuse with leads through holes, but I have seen different kinds of fuses and I am not sure which one to pick. Would you mind sharing a link to buy them here in Europe, let's say Amazon, Aliexpress or Ebay? Or at least telling me the keywords I should look for?

Cut away the heatshrink tubing from the old fuse and read what it says on the ends: there are fast and slow fuses and it is important to pick the right type...

Finally, it is probably that other components in the chain are also affected before arriving the Viper22a. Is it possible to test these components individually without having to connect the whole system? Or at least, a way to isolate parts of the circuit to avoid blowing anything else?

Thanks again for all your help.

Yes some parts can be tested, like diode check the bridge rectifier for example.
I would also check the 550ohm resistor and optocoupler U3, this can be done with diode check too.
As for the rest as long as you use the lamp as a current limiter there is not that much that can go wrong...

[Shinjinkun]

Re: Cyberpower CP900EPFCLCD - F03 Error

Great, I would replace this fuse with an old incandescent 40w or 60w light bulb.
If this bulb turns on bright when you turn the UPS on you still have a big problem somewhere.
You can of course skip this test and just replace the fuse but you risk blowing everything up again...

Ohh, I heard about the light bulb thing before, but I am not sure how to do it.
So, if I understand correctly, I will need a lamp holder (or bulb socket, not sure the correct name) with two cables. Where should I connect these two cables? To the same positions the broken fuse are connected? Or somewhere else?

Regarding the replacement fuse, I understand that it must be a 250V 2A one, fuse with leads through holes, but I have seen different kinds of fuses and I am not sure which one to pick. Would you mind sharing a link to buy them here in Europe, let's say Amazon, Aliexpress or Ebay? Or at least telling me the keywords I should look for?

Finally, it is probably that other components in the chain are also affected before arriving the Viper22a. Is it possible to test these components individually without having to connect the whole system? Or at least, a way to isolate parts of the circuit to avoid blowing anything else?

Thanks again for all your help.

[Per Hansson]

Re: Cyberpower CP900EPFCLCD - F03 Error

Done! I checked with the multimeter in continuity, and is showing OL, and it is not beeping. That means it doesn't have continuity, right?

Great, I would replace this fuse with an old incandescent 40w or 60w light bulb.
If this bulb turns on bright when you turn the UPS on you still have a big problem somewhere.
You can of course skip this test and just replace the fuse but you risk blowing everything up again...

[lotas]

Re: Cyberpower CP900EPFCLCD - F03 Error

So he burned out and broke the chain.