APC Back-UPS external battery mod (#P11)

Good thing I’m a doofus. While tidying up cables after some keystone module action, I also moved the bulky power supply of my monitor. And then I noticed that the only thing that absolutely has to be on the battery backup outlets of my UPS unit was in fact…assigned to the surge-protected ones that fail when mains power is gone. D’oh. After some cable identification in that tight space, I moved the UPS around a bit and I discovered a drop of liquid. A very dubious drop at a place where liquid would have to seep through 4 centimeters of a wooden table, or somehow permeate 1.5mm of solid steel against gravity’s pull. Panic mode. What is this, how did it get there, and is there more of it?

After removing the UPS and cleaning off the drop, this was left:

Etched in the powder coating of the 19″ rack shelf. That is friggen battery acid :evil:

Battery compartment: Yep, more of it

Lid: Yep.

Oh, and the non-spillable RBC17 lead-acid battery after heavy cleaning of the terminals (there was a 5mm layer of black gunk on both battery and internal faston connectors, the plastic protective cover of the negative terminal was almost eaten away and broke off during cleaning it):

I wanted to test emoji capabilities of WordPress for a long time, so here it goes: 🖕, APC!

Yeah so after the estimated service life the absolutely baby safe batteries turn into acid-leaking monsters. Good job, APC, good job.

My next UPS will be from a different brand. Until then, I decided to do something about it.

Main issue for leaking out (aside from non-spillable batteries that are capable of spilling) is the fact that when the UPS is used sitting flat, the battery doesn’t stand upright. This might be different for other models, but mine (“APC Power-Saving Back-UPS ES 8 Outlet 700VA 230V CEE 7/7” – BE700G-GR) has the battery compartment integrated, so there’s no way to insert the battery in a different orientation. When it is mounted on a wall in horizontal direction (so cable comes out to the left, I think), the battery should be upright and then the acid would need to wick out, hindered by gravity. That’s not perfect and will not completely stop capillary action, but much better than adding gravity in favour of spilling. So the battery should sit with terminals up, which is only possible with an external battery connector.

So I bought three sets of these puppies:

“Premium XT60E XT60 XT60E-M Einbaustecker Goldstecker Stecker Buchse + Schrauben” from eBay, 7.98€ total including shipping from Lithuania (!). These seem to be common in the RC helicopter world, I personally have never heard of those before doing research. The XT60 is just a standard male-female connector system with little overlap, while the XT60E is the one with mounting holes. You can get these from China, but right now I haven’t found a seller that offers single pairs of these, they all just sell either male or female connectors or they want you to take 50pcs at once. Price should be around 1.50€ per pair, so buying these in the EU wasn’t actually that expensive, given they arrived two days later after ~1500km of travelling and crossing two country borders.

Apparently there’s also XT-30, XT-90 and XT-150 connectors for your different current needs, each roughly indicating the continuous current capabilities with peak currents about 50% higher.

The connectors are really good quality I have to say. Perfect gold finish, smooth action, sturdy housing and contacts. I like them and I will order more for future projects (and to compare them to the trusted eBay seller units). Only thing to add as advice is to mate the connectors prior to soldering, as they tend to move a bit when held on soldering temperature for prolonged time. Which is necessary given the diameter of cabling required to e.g. move 60 amps along…

Next step: Opening up the UPS:

Two 40 amp fuses mounted on the board.

Front is somewhat blocked with data lines, so i will add the connector between lower and upper case.

Unfortunately, only the negative battery lead is soldered to the PCB and easily replaceable. The positive one directly goes to the huge transformer – which seems a bit odd for DC from the battery. Maybe that’s for filtering and current sensing purposes, I don’t know (and I won’t take that apart). Thing is: The company is called APC, not EUPC. So this cable is some crazy gauge, I think it was 10 AWG (5.26mm²). In countries without lunatics that refuse other measures than what their great-great-great-great-grandfathers had (so…in the rest of the world) that’s hard to find. On the other hand, using butt splice connectors to extend the wire doesn’t really work either – the difference in diameter makes for a lousy crimp that will not grip the smaller diameter side tightly enough. 6mm² would almost work, but that doesn’t fit the XT60 connector, 4mm² is far too small to make good contact to 5.3ish mm². There’s no metric 5mm² connectors or wire, at least not as a standard.

As I do want a plug on the device instead of an extended wire going directly to the battery (which could and will be under a certain voltage during operation), I had to stretch the positive wire a bit and exchange the negative entirely. The latter was soaked with battery acid, here’s the connector AFTER cleaning:

(first blog-related photo with my new camera!)
The wire doesn’t move a bit, it’s completely crusty. You don’t want that shit in live equipment…

Now I placed the XT60 in the front center:

Good enough of a fit, now add two mounting holes:

Solder the wires onto it:

Dremel out parts of the bottom case which separates the battery compartment from the insides of the device – don’t remove too much, as one would be able to touch live (230V) parts of the PCB:

Add screws and hot snot:

Close it – looks like a bought one!

And connect the battery with correct polarity :) (connector is of course keyed, so one can only screw up on the battery terminals itself)

Works a treat! Now sitting on a glass plate that was harvested from a scale. You know, just in case.

I’ve tested the battery with my electronic load, and at 1C (7.2Ah, so 7.2A for something like 90W total load) there’s still half an hour of juice left. That’s matching perfectly with the figures at the APC runtime calculator (sorry for the German site link, that page changes available products according to the region setting).

1 hrs 5 mins @ 50W
31.0 min @ 100W
12.8 min @ 200W
12.6 min @ 202W (50% load)
6.7 min @ 300W
3.8 min @ 400W
3.7 min @ 405W (max)

(that’s what you get for a “700VA” device, PFC is mandatory for years now, you get capacitive droppers with better power factor than that…)

So the battery seems to be in decent shape despite being 6 years old and now in the leaky regime. I certainly won’t replace it with some OEM part that costs about the same as a brand new unit bundled with a new battery. And as I said, I won’t buy another APC UPS in the foreseeable future.

There’s also icing on the cake: An identical unit appeared at work and people complained that it had abysmal battery lifetime. Battery’s however not leaking, did the same test at 7.2A, basically same result (unit is a little younger than mine). Well, turns out it originally was used to power two 24 port switches that were barely occupied. Battery runtime might go down a little if you exchange those with fully populated 48 port edge switches because your staff literally doubled in two years – and then add even more crap from the 19″ rack because those 8 outlets are pretty convenient. You bet they are forced to order proper rackmount units next time…


Leave a Reply

Your e-mail address will not be published. Required fields are marked *

:mrgreen: 
:neutral: 
:twisted: 
:arrow: 
:shock: 
:smile: 
:???: 
:cool: 
:evil: 
:grin: 
:idea: 
:oops: 
:razz: 
:roll: 
;-) 
:cry: 
:eek: 
:lol: 
:mad: 
:sad: 
:suspect: 
:!: 
:?: 
:bye: 
:good: 
:negative: 
:scratch: 
:wacko: 
:yahoo: 
:heart: 
B-) 
:rose: 
:whistle: 
:yes: 
:cry2: 
:mail: 
:-(( 
:unsure: 
:wink: