Cyberpower cps1500avr not charging batteries

[carrzkiss]

Hello, all.

Cyberpower Battery Backup Unit.

I bought two new ML18-12 batteries through Amazon (They have already refunded my money for the two batteries as we thought the batteries might have been bad.)
When I got them in, they were not fully charged. I connected them to the UPS and waited 24 hours before connecting any networking equipment.
In parallel, they read 24v, but within 24 hours, the meter on the front of the unit dropped from 4 to 3 bars on the battery charge indicator.
Checking the voltage, it was down to 22v.
I quickly removed everything from the unit and powered it off.

I put a Trickle Charge on each battery and brought each battery up to 13.6v.

I connected them to the UPS, and nothing else connected to the system.
The batteries have gone from 24.9v to 24.8, down to 19.8v within 48 hours.

I checked the voltage from the UPS output wires, and they each read 24v.

I am assuming one of two things.
Either the unit is no longer charging the batteries
OR
The batteries are, in fact, not good.

I had batteries on this unit about a year ago, but the batteries died, so that unit has been off for a year. (The batteries that were connected to the unit lasted for 4 years with constant runtime except during long power outages.)

When the two new batteries are connected to the unit, they do not seem to be charging at all.
And if I do a [test] on it, the system just shuts down.

Everything looks good on the board; I do not see anything unusual, like glue issues or burnt components.

I bought this unit new in 2001 and would really love to continue using it.
So, any suggestions would be wonderful.

I am not the best at electrical, but can do some things.
Thanks All.
Wayne

Pictures of the board.

[sam_sam_sam]

The first thing I would do is look at the charging circuit and this going to probably be near the small switching transformer this is the one that has the yellow tape around it what you are looking for is small value capacitor anything lower than 100uf change it even if looks good the other thing I would look at is for a switching power regulator ic chip ( this is a device that has 5 or more pins ) and check and see what is your DC Voltage at main filtering capacitor and post the results of the voltage reading

Also check the three automotive type 40 amp fuse and report there status open or close circuit

[carrzkiss]

Hey, Sam.
Thanks for your reply.
I have never done anything like this before, so I am at a complete loss.

(I honestly watched a video on how to test for open and close connections on a fuse, as the last time I tested a circuit for an open was about 20+ years ago.)
With the system unplugged and fuses still on the board, all three showed 0.00 closed.


charging circuit
I am unsure what I am looking for on this part.
I have taken high-resolution pictures that might help spot what I need to check. Also, how do I check to see if it is "100uf or lower"?

[sam_sam_sam]

Show some pictures of the back side of the board so I can locate the switching regulator if it has one or if you need to look for different type of regulator

[carrzkiss]

Here is the back.
I took three images, close-up, to make it easier to see.
Plus, a completed full board image.
Thank you, Sam.

[sam_sam_sam]

Can you please post the part number of “V1” and “V2” or a in focus picture of the part number

Look at the picture what is the black stuff around the soldering joint where the red circle that is marked if it is flux reheat and remove it from the soldering joint and then reheat it again until you have the majority of the flux from that soldering joint

There are some other soldering joints that are questionable at best from the picture of the hole board but it is possible that it might be the way the light is hitting them but I would look the hole board over for dull cracks around the component pins and the soldering joints to the board or missing solder around the component pins to the board or a lot of flux around the soldering joint and the component pins and if you have a magnifying glass would even be better for accuracy of the soldering joint and the component pins to the board especially the transformer pins look at them very carefully for bad soldering joint

[carrzkiss]

It seems the black might have been ink. It would not come off with the hot end of a solder, but it would scrape off with my fingernail. It is just a solder joint in that area with some kind of black ink or pigment on it.
I scrapped it off with my fingernail.



Attached is a better picture of that area before I scrapped it off and a picture after.

The V1 and V2.
Where will this be?

[harp]

Check nuclear between relay...

[redwire]

The smoke outline on the relays might be normal. After many years air currents cause dust and silt/smoke to deposit on things. But wait I see a fan, so that is odd there. Relay contacts might have arced.
I would give it a vacuum cleaning/compressed air blast and toothbrush rub on the back side, clean it up.

It has mains iron-core transformers- no SMPS here.
There are many fuses to check. Three of 40A automotive (red) blade fuses.
F1 not sure where it is.
F2 by the MOV's.
F3, F4 maybe for AVR taps.
F6, F7 by CN3 the serial port or something.

Usually it's relays that fail, fractured solder joints. Some MCU pins look dodgy soldering and those sockets are shit, they oxidize with humidity. Have you tried tapping (not too hard) the board with a plastic pen, relays included- to flush out poor connections.

If more complicated, the battery charge controller would have to figure out. The big heatsink at the board edge seems my guess for that. The other two big heatsinks are the (double) H-bridge to drive the (two big) transformers.
U2 might be the 5V regulator for the MCU power.
U3 looks a little baked? TO-220 next to HS5 and the red fuses. What is the IC on heatsink HS5 or next to it. One is a 5V regulator to power the MCU.

I can see it has a small power transformer making around 12VDC for the relays, and then 5V for the MCU. The two big (parallel-connected) transformers are for the UPS and connected backwards (by relays) they also charge the battery.
The medium size transformer is the AVR voltage boost, relays to switch it up or down taps.

[carrzkiss]

Update.
Once I cleaned that black ink or whatever it was from around the solder joint, I put everything back together, and it seems to be charging the batteries now.
When I connected the batteries, they were at 22.2v
They are now at 23.4v.

I will update you all tomorrow on the status.
That is going to be weird if that is the issue.

And I also cleaned the back and as much as possible on the front.
I vacuumed, but I didn't have any air, unfortunately.

[redwire]

I think you've got a dud relay or fractured solder joint, bad connection. Fingers crossed.
Battery pack float voltage should be around 27.6-28.0V.
Each at <11.7V is 0% charge, 12.5V is 75%.

[carrzkiss]

Yep, that is what it seems like.
It has been at 23.4 for over two hours now.
Can you mark the images where I need to check? I will take it back out tomorrow.
I just have the wires connected up, but not put together.

[redwire]

The microcontroller could have turned off the charger, or the charger circuit/IC is damaged, or the charger fuse is blown etc.

I'm guessing the battery charge control is an LM317 IC on a heatsink. This is the usual Cyberpower circuit.
Can you read any part number on that part, I know it's in a sandwich. I can't see that portion (top) of the board either to see the parts (where I put green dots).
Can you measure the DC voltages on the pins and post. Use BATT(-) for multimeter common(-). Careful the board has hazardous live AC mains on it too.

[carrzkiss]

It seems harp might have been correct.
After taking everything apart, I found this.
It is burnt, which would explain the black from the back.
The black was soot from the front resistor burning up.
The resistor is R42.

Is it just the resistor that needs replacing, or is the power relay beside it also?
Could someone please provide a link to purchase this from?

I've marked up the last image to show where everything goes.
I believe I got it marked upright.

The only thing I have to work with is a small soldering iron.
I also have some solder, but I am unsure what type of solder I have. (I remember my father telling me years ago that there are different types of solder for different jobs.)
If you all could let me know what I will need to remove the resistor
And to put one back on, I will get them.
And any other information or advice.

==================Memories==================

I'm looking forward to doing this as it reminds me of watching my dad in the 80s when I was a teenager working on his home stereo receivers.
I remember sitting there and talking with him about what he was doing. I was very interested in doing it, and about a year later, I was able to make my combo stereo louder.
When my dad came down to pick me up for the weekend, I showed him what I had done and played the radio.
He was amazed at what I had done and told me to be careful as electricity was nothing to play around with.
When I was in the 5th grade, I connected two motors on my RC race car and made it jump 30 feet across the sunroom.
The things we did as kids amaze me that we are still alive.

[harp]

If you all could let me know what I will need to remove the resistor

NHF, but I am not sure that you can handle this job, it envolves some experiance with soldering iron, and ofcourse troubleshooting methods and equipment.

Yes, you can learn it, but dont learn on board what you intend to repair, but use some scrap board, ewaste... for knowledge is easy, only spent time and effort surrounded by willing to do so.

Your board is double or multilayers which is probably burned and lead to generating heat in it. Proper way to fix this is to desolder both relays and resistor, inspect them separately to be sure it is ok contact, or repace them... then you must fix board, inspect layers, fix it with dremel, epoxy... and ensure that there is no resistance between it... and this fix is no garantee that device will work again, but it is ok to try.

When resistor is overheated, it change solder tin properties, so it is harder to melt. When you desolder bottom side, the top side is still soldered, and you may end up with ripped tracks, which is already weak due to burn treatmant.

I just want to warn you before you start, maybe have some friends who can do that...

Ok, to remove that resistor, you need, soldering iron, fresh tin 63% to lower melting point and ensure good contact, and some tweezer to grab a leg, lift each simmetrical in small steps.

[sam_sam_sam]

Your board is double or multilayers which is probably burned and lead to generating heat in it. Proper way to fix this is to desolder both relays and resistor, inspect them separately to be sure it is ok contact, or repace them... then you must fix board, inspect layers, fix it with dremel, epoxy... and ensure that there is no resistance between it... and this fix is no garantee that device will work again, but it is ok to try.

When resistor is overheated, it change solder tin properties, so it is harder to melt. When you desolder bottom side, the top side is still soldered, and you may end up with ripped tracks, which is already weak due to burn treatmant.

I just want to warn you before you start, maybe have some friends who can do that...

Ok, to remove that resistor, you need, soldering iron, fresh tin 63% to lower melting point and ensure good contact, and some tweezer to grab a leg, lift each simmetrical in small steps.

I would add this if you decide to remove the relay from the board I would suggest that you get a desoldering sucker at a minimum or some soldering wick and some flux and put on the soldering joint and heat the soldering joint and when you have the solder flowing then remove it from the soldering joint if it takes longer than 10 seconds for the solder to flow you do not have it hot enough it should flow with in less than 5 seconds for best results do not get inpatient and try to remove the relay until you have all of soldering joint clear of solder and the replay pins are in the center of the hole and there is no solder touching the pins before you attempt to remove the relay it should take no effort to remove it from the board if it does then you do not have all of the solder remove from the top soldering joint put a little bit of solder on the joint and try removing it again until you have all pins free

I hope this helps you successfully remove the relay without tearing the top traces of the board where the relays are

[redwire]

OP I thought you checked that out already.
The resistor one end connects to mains Neutral, other end to C16 so I would guess it's a snubber resistor and cap C16 has shorted.
Putting in just a new resistor will smoke it. Resistors only burn up when overloaded, so there is another problem- hopefully shorted C16.
This still doesn't explain the no charging problem, unless the relays melted and don't work. You need a desoldering tool to get the parts out.

[carrzkiss]

Thanks guys.
And Harp, you are probably right,
I don't know how to troubleshoot this type of equipment. I can narrow down issues with my Dell servers and determine what is wrong with them.
But with this UPS and no workbench or equipment to run tests and do what needs to be done. I'm at a loss.
I have looked to see if there is a repair shop in my area, but there does not seem to be any that deals with this type of equipment. Most are computer and cellphone repairs, not UPS Repairs.

So, at the moment, I am stuck.
But I do appreciate the help you all have given me. Taking this unit apart and at least troubleshooting a portion of the issue was enjoyable. I love learning new things.

redwire, I only scraped the soot from the back of the board and had not done anything else to it. I did not look between the relays until I took it apart today.
Not having a place to put the board safely has prevented me from getting the readings that you requested.
I looked at it this evening and thought about it, but that is just too much power to mess with.

Once again, thank you all so much, and thank you, Harp, for being honest with me and telling the truth. It is very much appreciated.

If I had someone to send the board to without the transformers and case (Lower weight for shipping), I would do that.

Too bad my old friend from the 80s is not still around; I would call him up.

Thanks, everyone.
Wayne

[carrzkiss]

redwire I think this is what you had asked me for earlier. I think.

[harp]

I see something suspicious... what is model of your multimeter, picture?