Cyberpower CP900EPFCLCD - F03 Error

[Shinjinkun]

Hello guys.
Please help me fix this UPS!

I got this UPS from a guy on Ebay. It stated that it wasn't working, but didn't specify about the F03 error.
Anyway, when I received it, I supposed that just the battery was dead, a quick fix. Changed the battery, connected the device, and everything worked without issues. However, three days later, the UPS started beeping, showed the F03 error, and had to switch it off, as I didn't know what was going on.

Going forward, I researched about the problem, and found that it was an issue with the charging circuit. It seems that, while the battery was at full charge, the ups was working normally. It was when the battery lost part of the charge, and the circuit tried to recharge it, that the problems arised.

I opened the box, and found that a cap and the vyper22a were blown. And also, it is possible that other parts are bad, as the infamous glue was around several critical spots. I removed all of the glue, but obviously the parts need to be swapped.

I removed the blown cap and the vyper22a, but in the process, I also broke a SMD capacitor (C15). Does any of you know the capacity of the white capacitor I removed by mistake? Or if there is a way to calculate it?
Furthermore, I think the zener diode and the s1g diode may be also dead. Any idea of the values of the piece were I to buy it? So far, I ordered the capacitor, the vyper22a, and some smd capacitors, but I don't know about which values I need.

So, long story short, I need the specs for the white capacitor (C15) and the zener diode (Z1) in order to buy them to try to replace them.
Could you give me some ideas, please? I am really in the dark right now What other components should I test? Is there anything else I could do?

Thanks everyone!

[Per Hansson]

The datasheet for those IC's contains implementation examples.
Usually the manufacturers follow those quite closely so that is a good starting point.

[Shinjinkun]

Originally Posted by Per Hansson

The datasheet for those IC's contains implementation examples.
Usually the manufacturers follow those quite closely so that is a good starting point.

Do you mean the datasheet for the Viper22a or the board? I am pretty inexperienced in electronics, so I'd appreciate if you could share a link to follow, please.

[momaka]

Originally Posted by Shinjinkun

Do you mean the datasheet for the Viper22a or the board? I am pretty inexperienced in electronics, so I'd appreciate if you could share a link to follow, please.

Well, if you can find a schematic for the UPS, that would be great. But if not, the VIPer22a datasheet does contain a lot of useful information.

VIPer22a datasheet:
https://www.st.com/resource/en/datasheet/viper22a-e.pdf

Originally Posted by Shinjinkun

So, long story short, I need the specs for the white capacitor (C15) and the zener diode (Z1) in order to buy them to try to replace them.

Z1 is not that critical for the operation of the VIPer22a - it's there for protection, in case V_dd goes too high. Since the VIPer22a IC is rated for up to 50V on the V_dd pin (though operating, it is specified for 9-38V), my guess would be that a Zener anywhere in the range of 30-40V should be safe to use without possibly clamping too early (if the transformer's primary auxiliary winding - the winding that supplies V_dd to the IC via diode D5 - has a really high output, for example.)

Just make sure diode D5 is good or replace it. Here, a fast-recovery 40-100V rated diode should work fine.

Also, don't forget to replace the small 50V, 47 uF cap! That's your startup/run cap.

As for ceramic cap C15... this one might be a little more challenging to guess. I don't see anything in the VIPer22a datasheet that specifies how to calculate it. That said, I looked up images of random VIPer22a schematics, and the value for the capacitor in this spot varies anywhere from 220 nF to 10 nF. My guess would be that 47-100 nanoFarads will be OK. Just know that the higher the capacitance, the slower the response of the feedback to the IC will get. This could be a good thing or a bad thigh, of course. From what I'm seeing from the online picture schematics, 47 nF is the most common value, so perhaps go with that. As for voltage rating... since pin 3 of the VIPer IC gets is current from the V_dd supply through an optocoupler's output transistor, the voltage will probably be lower. Nevertheless, 35V-50V rating of the cap should be sufficient.

[Shinjinkun]

Originally Posted by momaka

Well, if you can find a schematic for the UPS, that would be great. But if not, the VIPer22a datasheet does contain a lot of useful information.

VIPer22a datasheet:
https://www.st.com/resource/en/datasheet/viper22a-e.pdf


Z1 is not that critical for the operation of the VIPer22a - it's there for protection, in case V_dd goes too high. Since the VIPer22a IC is rated for up to 50V on the V_dd pin (though operating, it is specified for 9-38V), my guess would be that a Zener anywhere in the range of 30-40V should be safe to use without possibly clamping too early (if the transformer's primary auxiliary winding - the winding that supplies V_dd to the IC via diode D5 - has a really high output, for example.)

Just make sure diode D5 is good or replace it. Here, a fast-recovery 40-100V rated diode should work fine.

Also, don't forget to replace the small 50V, 47 uF cap! That's your startup/run cap.

As for ceramic cap C15... this one might be a little more challenging to guess. I don't see anything in the VIPer22a datasheet that specifies how to calculate it. That said, I looked up images of random VIPer22a schematics, and the value for the capacitor in this spot varies anywhere from 220 nF to 10 nF. My guess would be that 47-100 nanoFarads will be OK. Just know that the higher the capacitance, the slower the response of the feedback to the IC will get. This could be a good thing or a bad thigh, of course. From what I'm seeing from the online picture schematics, 47 nF is the most common value, so perhaps go with that. As for voltage rating... since pin 3 of the VIPer IC gets is current from the V_dd supply through an optocoupler's output transistor, the voltage will probably be lower. Nevertheless, 35V-50V rating of the cap should be sufficient.

Wow, that's a thorough explanation! Let me review all these values and see if I could get all these components. But at least, it looks promising
Thanks a lot for the suggestions!

[Shinjinkun]

Originally Posted by Shinjinkun

Wow, that's a thorough explanation! Let me review all these values and see if I could get all these components. But at least, it looks promising
Thanks a lot for the suggestions!

FYI, I already ordered all the missing components.

I got a zener diode kit for Z1, with several values:
ZMM30V (30 V), ZMM33V (V), ZMM36V (36 V),ZMM39V (V).
Which one of them should I use? You commented it should be between 30 to 40V, but which one should do better?

Bought the S1G cause I think is in bad shape.
Got an extra 50V 47uF in case it doesn't work.

For the C15, also got a kit with loads of values, 47nF included. I hope that one should work ok.

One last comment, in case the SMD resistor 551 break or something, also got a resistor kit but, the kit only contains a 560 OHM, instead of the 550 OHM that the 551 should have. Because they have a 5% tolerance, I suppose it should work too...

I will let you know once I receive all the components. Thanks again!

[sam_sam_sam]

Originally Posted by Shinjinkun

One last comment, in case the SMD resistor 551 break or something, also got a resistor kit but, the kit only contains a 560 OHM, instead of the 550 OHM that the 551 should have. Because they have a 5% tolerance, I suppose it should work too...

I will let you know once I receive all the components. Thanks again!

It should work

[Shinjinkun]

Hi guys!
Finally, I got all the components needed to reattach.
As you can see in the attached picture, I got:
Viper22a integrated
Capacitor 400V 10uF
Capacitor 50V 47uF
Diode S1G
Capacitor SMD 470pF
Zener Diode MELF 39V

I already removed all the components.
I checked the 50V cap, but I found that it was fine... anyway, I will put a new one.
About the 551 resistor, I don't think I need to change it.
Do you have any suggestion to resold everything back?
Is there any test I should perform after having everything in place?
Thanks again!

[sam_sam_sam]

So have you tried to power up this switching power supply yet to see if you fix it or not

[Shinjinkun]

Originally Posted by sam_sam_sam

So have you tried to power up this switching power supply yet to see if you fix it or not

No, I haven't yet, as I didn't complete all the soldering. I am waiting to get a socket for the integrated, so once I receive it, I will complete the final piece. I attached the picture of the current motherboard after soldering most of the components.

I will show you the final pictures and tell you if it works or not once it's complete.

Thanks!

[pete_c]

@Shinjinkun

Very nice soldering on your board.

Have you tested the UPS yet?

[Shinjinkun]

Originally Posted by pete_c

@Shinjinkun

Very nice soldering on your board.

Have you tested the UPS yet?

@pete_c I finished with the fix a couple of days ago, but I was using it in case I got the F03 error again. It doesn't look like it is the case though. So far is working great.

Big thanks to @momaka for the suggestions of components. It seems that they were spot on.

In the end I swapped both caps (the 50V and the 400V). About the Viper22a, I added an extra socket in case I have to change it in the future, but I hope it won't be needed.

You can see the pictures with the final board components. I was using the multimeter to check that the voltage of the battery was within normal values. It is curious that this UPS doesn't overcharge the battery as some off-line ones that I own. The others charge the battery at 13.5V, but this one seems to only use 12.5V instead. The reason is probably the better design and components, but who knows...

Again, thanks to all for your help. It is really appreciated.

Regards

[Shinjinkun]

Guys, forget about what I said previously. I just got the dreaded F03 error again. I checked the battery, and it was around 10.29V, so I have the impression that for the latest 3 days, the battery has been discharging without realizing it....
this only means that even that I swapped all the broken components, there is still something wrong with the board. Now I have no idea how to track the error and what can I do to fix it.... T_T
Any suggestions will be really appreciated.

[Per Hansson]

Yes proper standby charge voltage for a lead acid battery is around 13.7v and charge voltage is around 14.7v
Now that you replaced the Viper have you checked what voltages you have on the capacitors you replaced, so we are sure that it is working at all?

I notice in your first picture some soot around component U7.
I think it is a ST 324 op-amp that is probably used in the charge control circuit.
Hard to know without a circuit diagram of course.

I attached a picture where I highlighted a possibly bad solder joint for a through hole component.

I also would not exclude the 550Ω resistor or the optocpupler U3 since they are in the path of destruction too...

[lotas]

Yes, it is clear that the 550 Om resistor has a burnout.

[Shinjinkun]

Originally Posted by Per Hansson

Yes proper standby charge voltage for a lead acid battery is around 13.7v and charge voltage is around 14.7v
Now that you replaced the Viper have you checked what voltages you have on the capacitors you replaced, so we are sure that it is working at all?

Can I use the multimeter directly over these caps while working? I mean, should I get the 50 or 400V on them? I never did this, so it is difficult for me.

Originally Posted by Per Hansson

I notice in your first picture some soot around component U7.
I think it is a ST 324 op-amp that is probably used in the charge control circuit.
Hard to know without a circuit diagram of course.

Yes, I also noticed the burnt marks, but I thought it was from the cap or the viper when it poofed. I checked continuity on the components on top of the U7, but I am not sure how to test the U7 itself.

Originally Posted by Per Hansson

I attached a picture where I highlighted a possibly bad solder joint for a through hole component.

I can desold and resold it just in case.

Originally Posted by Per Hansson

I also would not exclude the 550Ω resistor or the optocpupler U3 since they are in the path of destruction too...

I thought about replacing the 550Ω resistor, but I thought it was ok. I can replace it as I have some spares. Regarding the optocupler, how can I check it? I will try to find this part somewhere.

[Shinjinkun]

Originally Posted by lotas

Yes, it is clear that the 550 Om resistor has a burnout.

I will try replacing it.

[Per Hansson]

Originally Posted by Shinjinkun

Can I use the multimeter directly over these caps while working? I mean, should I get the 50 or 400V on them? I never did this, so it is difficult for me.

Yes exactly, start there and then we can look at the rest.

[lotas]

If the switching power supply works, then voltage should appear on this capacitor.

[Shinjinkun]

Originally Posted by Per Hansson

Yes exactly, start there and then we can look at the rest.

I took the measures for the three capacitors. These are the values I obtained:

C3 - 400V 10uF ---> 0.403V
C14 - 50V 47uF ---> 0.184V
C17 - 25V 470uF ---> 0.939V

I haven't replaced the 550Ω resistor yet, though. Not sure if this point can affect the taken measures.

I attached a couple more pictures in case they help.