Even after replacing the batteries, the message "PMS18 Battery defected-load test failed"still remains.

you may want to be a bit more specific... there is more than one KRC4 controller type, more than one CCU and more than one battery type. it is unclear what exactly is wrong. no system info like age, history... and what is the exact issue that bothers you? is the issue for you that something is written to a log although everything appear to work, or maybe there is some specific performance issue? if there is a performance issue, what kind? .

so does that mean that you have problem on both of them or just one?

CK stands for "Custom Kinematics" so robots are not KUKA

"KRC4 NA extd" is a KRC4 extended, North American edition which uses same CCU and batteries as KRC4 standard.

batteries are always 12V, connected in series but type and shape id different in compact controller for example.

since batteries hold charge, issue may be with software or firmware or PMB or X305 wiring.

software issue could be remedied by restoring known good backup image.

firmware issue can be detected by running KRCdiag and evaluating logs (send to KUKA). for example stable PMB firmware was 2.0.3.0 for very long time.

X305 is using 4-wire connection. check if contacts are ok. they have low rated life (something like 3-4 disconnections) and should be handles carefully. female contact get easily deformed and after that they don't make good connection. not sure if 4 wires are simply used for load sharing or this is a Kelvin connection. check for abnormal signs like bent, pushed out, oxidized or flared contacts.

not sure.. never needed to play with that...

but at the very least you should be able to change interval to run more frequently. check your PMService.config settings. here is an example:

Quote from panic mode

X305 is using 4-wire connection. check if contacts are ok. they have low rated life (something like 3-4 disconnections)

I took this step to debug the same issue on our KRC4 cabinet - the connectors were corroded, causing the test to fail with new batteries, even though the batteries worked in other cabinets.

Quote from Aleanto

5. I connected set of new batteries (tried to grow up a capacity for experiment) and after that test went ok. After that I changed batteries 12V5Ah to 12V7.2Ah and it has no effect. Test failed.

By this sentence, do you mean that you made the test pass with one set of batteries, then put in a different set and it failed?

I have had a battery test fail in a KRC4 cabinet, even though the batteries were able to hold a charge, had a voltage around 12.7V, and caused other KRC2 cabinets to function normally (shutdown safely, etc). I had to replace the batteries with new ones for the test to pass. Your problem may be similar, based on the test.

I think that your batteries have good voltage without load, but they can't provide the necessary current for the load test - maybe the voltage drops at high power draw?. If this is the case, then a test with two sets of batteries in parallel (your Frankenstein test) would provide the same voltage with a greater available amperage, and the test could succeed (like yours did).

Quote from Aleanto

I connected two multimeters to monitor the voltage and current at the moment when I urgently disconnected the power from the cabinet.

What do you mean by "urgently disconnected the power from the cabinet"? Were you measuring the load on the batteries after shutting off the controller, to ensure that they would let the controller shut down safely?

A set of batteries can keep the cabinet alive after shutdown while failing a load test. I have had battery sets do this. The load test requires higher performance than a typical cabinet shutdown, in my experience.