APC SMT1500i problem

[nnemanja]

Every few days, and consistently at night when it is at minimum load, a problem arises with the UPS, marked by a clicking noise (resembling an abnormal and rapid relay clicking, as captured on camera). After about a minute, it draws so much power from the wall that the fuse in the fuse box turns off. Following this, it operates on battery for the next 5-6 hours until the battery is depleted. In the morning, after manually resetting the fuse in the fuse box, the UPS charges the battery and continues to work normally. This has happened multiple times, always at night and under minimal load. I disassembled it, but I am unable to diagnose the issue. The only thing I noticed on the table is that a few hours after powering on, relay RY900 shuts off, and transformer T900 on charger board (module located on a separate charge behind the battery compartment) starts abnormal whining. Has anyone experienced a similar problem or does anyone know what steps to take next?

[asis]

Hi,

Apparently at night the mains voltage is exceeded and the UPS, through a relay, tries to switch the output windings of the transformer to autotransformer mode, lowering/increasing this voltage (you need to observe).
If this anomaly is associated with the electrical distribution substation, then, naturally, this will greatly wear out the contacts of the BOOST/STRIME relay, which leads to electrical erosion of contacts and their sticking.
It is possible to change the values (range) of threshold voltages, for which you need to enter the settings via RS232 using the UPS control mnemonic code.
Or.
If you have an APC network card (NMC), you can do the same and monitor all events associated with a dip/rise in the network voltage (daily log).
In your case, it is easier to replace relays with the same characteristics, understanding that their resource is not unlimited.
You can really assess the condition of the relay contacts after opening the case. Alas

[nnemanja]

Hello,

I have already purchased a set of new relays, and I'm waiting for them to arrive because I suspect that the problem may be in one of them. There is one piece of 894H-2AH1-F-S U04 and four pieces of 793-P-1C-F 001 on the main board.

[asis]

Hi,

Please tell me the model of your PCB (640-30xx....) and take a photo of the charger on both sides of the PCB.
Check the spark extinguishing circuit for relay contacts RY900 (R932 - 1.5k and C916 - ~ 85nF).

On your PCB I do not see the presence of ECAP C8 (or C7#) 2700uF x 40V.
Wasn't he there?

[nnemanja]

1. Please tell me the model of your PCB (640-30xx....)
   640-3081A_REV06
2. take a photo of the charger on both sides of the PCB.
   I'm sending all picture I made in google drive link. If you need more just tell me.
3. Check the spark extinguishing circuit for relay contacts RY900 (R932 - 1.5k and C916 - ~ 85nF).
   I took the charging board from another SMT1500 yesterday because when relay RY900 turn off, my charging board's transformer T900 start whining. I can't use that donor board because
   it's not the same as mine (I'm sending pictures of both), but I did find R932 and C916 on it. On mine, I can't find R932 and C916.
4. On your PCB I do not see the presence of ECAP C8 (or C7#) 2700uF x 40V.
   Its back now, I took this pictures while testing that cap.

https://drive.google.com/drive/folders/1g3xhYjSkwfIUknllgreSZJZoyBUh1o3P?usp=sharing

[asis]

So.

Lots of visual and audio information.
Thank you.
I don’t know yet how to build behavioral tactics.

I’ll start with “what I see is what I have”.........
I advise:
 - do not use a charger from another UPS.
   They are different both in the number of pins on the J906 connector and in the controls.
 - you can use (dismantle) the relay, which is installed on a separate small board and on the charger,
    which has the same one. This is a relay for controlling group sockets, an option that has hardly
    ever been used. They must be in perfect condition (two donor pieces!).
 - don’t rush to do something until you are mentally prepared for it and follow the links.

https://www.eevblog.com/forum/repair/apc-smt3000rmi2u-battery-not-charging/ 
https://www.eevblog.com/forum/repair/apc-smt1500rm2u-ups-only-blinks-and-beeps-when-plugged-into-mains/
https://www.eevblog.com/forum/repair/apc-smartups-overload/
https://www.eevblog.com/forum/repair/apc-smt-3000rmi2u/
https://www.eevblog.com/forum/repair/how-to-disable-green-mode-on-apc-smart-ups-smt-series/

[nnemanja]

Thank you, Vladimir.

I need to reinstall that SMPS board in its UPS, so I can't use the relay from it. New relays are on their way, and I will replace them when they arrive. From your experience, is there a point in diagnosing exactly which relay on the main board is faulty, or should I replace all five?

I'm also concerned about the loud sound of the transformer on the SMPS board. From what I gathered from the posts you sent me, it makes sense that it happens every time when relay RY900 activates that board, usually the first time 2-3 hours after powering up because the UPS goes into "green mode". Do you know what could make it whine so loudly? (I can record the sound if you want.)

[asis]

Following this, it operates on battery for the next 5-6 hours until the battery is depleted.

Do you use high-capacity external batteries?

Then, after a deep discharge of the battery, the charger becomes heavily overloaded and, as you said, “begins to whine abnormally.”

[nnemanja]

I've noticed that this type of UPS has an issue with SMPS when an external battery is connected. I've also observed that the SMPS is only active when the UPS is in green mode, and this mode can be deactivated through a PC. Unfortunately, I can't connect it to a PC right now to deactivate it, but I attempted a quick and dirty fix by disconnecting the SMPS power supply from the main board. The UPS seems to work fine this way – it transfers to battery when power is disrupted and charges the battery through the transformer when power is back.

A few hours after the battery is charged, the UPS goes into green mode, and relay RY900 activates the power supply to the SMPS. However, because it is not connected, it doesn't buzz, and the UPS doesn't show any errors.

I'm unsure if this could potentially harm the UPS or the battery in the long run.

[asis]

I've noticed that this type of UPS has an issue with SMPS when an external battery is connected.

Hi,

The internal charger is not designed to connect an external battery.
In addition, the first time the UPS is connected to the network, regardless of whether "Green Mode" is enabled or not, the charger provides power to the internal SMPS (*) assembled on IC12 (TL3843).
It produces all the main voltages: +5V(IC600);+5(IC9);+12V (Catode D44);-9V(- C714 22x50V); +24V to power all relays.
If the battery is new and in a charged state, upon turning on the power, the main UPS test is performed, and if it is successful, the UPS goes into BYPASS.
The load connected to the UPS is powered.

If the battery is severely discharged, the UPS test fails and the charger charges the battery until the charge reaches more than 30 percent.
During this period, if the battery charge current (especially external) is maximum, a forced charge occurs.
The charger may not have enough power to power the internal SMPS (*).
For this reason, the behavior of the RY900 is not predictable (judging by the audio).

(Connector pins J905/906 pin3 and pin5 are charge voltage/current control.
Pin3 control DOWN (increment U), Pin5 control UP (decrease U) - control is performed by MPU IC11 89C51RC (pin24, pin6).

The connector contact J905/906 pin7 is a control voltage that is generated through a zener diode D717 (MMBZ5237) with a voltage threshold of 8.2V.
If the charger is working properly, voltage is supplied to the base of Q43 and creates the conditions for the UPS to start.)
Charger - 5G (5th generation) J906 8pin.
--
Regarding UPS operation in "Green Mode" -
https://www.eevblog.com/forum/repair/how-to-disable-green-mode-on-apc-smart-ups-smt-series/
 
I think that the charging power of the battery when the reverse current passes through the “primary winding” of the transformer that controls MOSFET Q914,R966+R967 and MOSFET's Q13,Q25 (the upper half of the half-bridge) will be significantly less than from a standard charger.

Yes, the charge will occur, but to a lesser extent and for a longer time. (half period)
Taking into account all the vulnerabilities described above, there is a possibility of failure of MOSFET's Q13,Q25.
I don't consider "Green Mode" to be the main one.

BTW, as far as I remember, "Green Mode" can be disabled/enabled from the remote control.

P.S.

In this link I tried to reverse the 4.5G charger circuit but didn't complete it.
There are probably errors in the diagram.
Could you please check the diagram.
I will be very grateful.

Thank you.

[nnemanja]

I will check it. Do you know characteristics of R932 and can it be replaced with normal resistor. It failed on that donor SMPS board from another smt1500.

[asis]

R32 1501 (1.5k) -1206 0.25W, any carbon can be used.

[nnemanja]

After replacing the failed R932, the other 'donor' UPS is now working. Initially, it wouldn't start without a battery, and a few seconds after powering on, a loud buzzing noise emanated from the broken resistor. The error message displayed was 'AVR relay weld' or power error. On this UPS, R932 is an SMD resistor labeled 1501, with a resistance of 1500 ohms. It is in the 2512 imperial or 6332 metric size, measuring 6mm x 3mm. As a preventive measure, I also replaced the neighbouring 100nF capacitor.

Unable to find the exact SMD resistor, I used a 1500 ohm 2W resistor and a slightly larger dimensions 100nF capacitor. Now, the UPS starts normally, and it is currently undergoing testing.

The workspace was quite cramped between the relay, power connector, and slow-blow fuse. You can observe some marks where I touched them with the iron, but fortunately, they are all functioning.

Currently, I am awaiting replacement relays for the first UPS.

Thank you for your valuable guidance, Vladimir.

[nnemanja]

Update: Both SMT1500 units are working fine. However, both also experienced the same problem with the spark extinguish circuit for the SMPS relay (RY900). One unit had an issue with a resistor, while the other had a problem with a capacitor. Additionally, in one unit, buzzing from the SMPS transformer is emanating from the transformer itself, indicating a possible broken ferrite core. Unfortunately, I haven't been able to locate a replacement transformer, and I may need to repair it when I have the time. Furthermore, both orders for replacement relays from AliExpress arrived as used, which I opted not to use. I'm currently in the process of finding suitable replacement models and purchasing them for future use.


PS: Asis, I would like to study that diagram, but it is of low quality. Could you please send me the full diagram in high quality?

[asis]

Additionally, in one unit, buzzing from the SMPS transformer is emanating from the transformer itself, indicating a possible broken ferrite core.

Hi,

I want to repeat again - you have two different chargers, the circuits of which are different.
Due to the fact that you are using (not intended) high-capacity external batteries, the built-in chargers are not able to provide the necessary charging current in the event of a deep discharge of the batteries.
You yourself, consciously or not, took this step.
The pictures you provided show the characteristic blackening of the PCB due to overheating of the transformer.
It is possible that the compound/glue lost its properties and a gap formed in the ferrite magnetic core, which led to a characteristic whistle.
If you continue to use this charger, this will lead to its failure and failure of the components.
-
If you have the opportunity (?) to check the circuit (J906 4-pin connector), the reverse of which I did from a photo from the forum?
Please do this.
I'm not sure there are any mistakes there and I don't want to mislead others.
This will give you the opportunity to make some recommendations.

[nnemanja]

Q1: I want to repeat again - you have two different chargers, the circuits of which are different.
Due to the fact that you are using (not intended) high-capacity external batteries, the built-in chargers are not able to provide the necessary charging current in the event of a deep discharge of the batteries.
You yourself, consciously or not, took this step.
A1: I mentioned that I have read that these units have a problem with SMPS chargers when external batteries are connected, but I don't use external batteries with it. I acquired it used, so it is possible that external batteries were used with it at some point.

Q2: The pictures you provided show the characteristic blackening of the PCB due to overheating of the transformer.
It is possible that the compound/glue lost its properties and a gap formed in the ferrite magnetic core, which led to a characteristic whistle.
A2: I am almost certain that the ferrite core broke. I can hear it rattling inside, and in some positions, the transformer is very quiet. Regarding the heating issue, I overlapped a picture of components that are on the other side of the board with the part that has overheated, and it seems that the blackening is caused by the TNY280GN and neighboring diodes and resistors.


Q3: If you continue to use this charger, this will lead to its failure and failure of the components.
A3: I had the same concern, so I disconnected the SMPS charger. This unit is now working without it.


-
Q4: If you have the opportunity (?) to check the circuit (J906 4-pin connector), the reverse of which I did from a photo from the forum?
Please do this.
I'm not sure there are any mistakes there and I don't want to mislead others.
This will give you the opportunity to make some recommendations.
A4: I don't have that unit with me right now. After repairing the spark extinguishing circuit on it, I assembled it, and it has been working fine on site for a month now. I have another unit with an 8-pin J906 on me, and I can check it. I want to study more how this unit work. Do you have the full schematics of the main board?