Wednesday, January 09, 2019

Reviving the batteries on the FSP EP1500 UPS

Last year in December I stumbled upon an FSP EP1500 UPS in a thrift store. It was missing its outer cardboard packaging but looked otherwise in pristine condition and supplied with usb and serial cable and plastic wrapping.

There was no price tag on the unit and I asked one of the clerks what they wanted for it. He suggested about 16 dollar which to me looked like a great bargain.

I bought it and got it home. Instructions said that to leave it to charge for 6 h before turning it on which I did.
I left it charging for about 24 h before trying to start it but no cigar, it was completely dead.
Attaching an energy monitor and measured that it pulled a couple of watts but nothing happened.

Next step was to disassemble it to check the insides and access the lead acid batteries. Investigating the date codes inside the machine gave date codes in the 2013 - early 2014 region. Still the unit looked clean and unused inside.

There were two 12V 9Ah Lead-Acid batteries in series (24 V) glued together. Measuring the voltage between the poles was a sad affair both giving about 0.5 V each. This means that the battery has gone into deep sleep and needed replacement.

I tried to connect one of them to a battery charger but it failed to sense that it was being connected to a battery at all.

As a last effort I decided to connect the battery terminals to my programmable power supply. Set the voltage to 14.40 V (2.4 V * 6 cells) and set the current limit very low, initially only 10 mA.

Almost no current was consumed, only about 2 mA. Still I left it on for a couple of minutes. I then measured voltage between the poles and it had risen with a couple of tenths of a Volt.

Given this slight, small spark of life I continued to charge it over multiple days. Slowly, slowly the voltage increased as did the current consumed. After about a week of charging the voltage had inched its way up to about 11 V and the current consumption to about 50 mA. I now connected the battery charger instead and now the battery was detected.
I continued to charge it at the minimum current (0.1 A) the charger allowed.
This took an additional 3 days or so but finally the battery reached the nominal max charge Voltage 14.70.
I still had no idea how much of the 9 Ah capacity that was retained.

The procedure was repeated for the other 12 V battery and it gave a similar charging response.

Now it was finally time to reassemble the unit and power it on. Lo and behold the UPS started as expected.

A power disconnect test without any load connected was conducted with the following graph:

The UPS has, how shall we say it, interesting features such as being able to power itself even though the battery percentage is 0. I also like the voltage raise observed. Runtime was 13492 seconds = 3 hours and 45 minutes.

Recharging gave the following graph.

Next up is to run the UPS with a 25% load. Its nominal capacity is 1500 VA / 900 W. I will hopefully post updates with those charts.