APC Symmetra battery module

[Anonymous user]

We have had several SYBT5 modules show as failing in the last couple of months, we'd like to be sure they have indeed failed (they have 10 individual batteries inside) before we recycle them and replace the modules.  Does anyone know a procedure to reset the module status and force the chassis or software to recognize the batteries as fresh and retest them?

[BillP]

Well, I tried with a different chip out of the board and it programmed fine.  I was trying to use the factory chip before, maybe that's what the issue was.

[Anonymous user]

Maybe, but I was totally successful doing it in place with the factory chip. Have you tried your chip-clip setup with just a bare chip to make sure the wiring to the clip is correct and that the clip is making contact? You can put a bare chip into a free-standing extra socket if you have one to simulate the physical setup of a board.

[BillP · Accepted solution]

Yep, I programmed the new chip with the clip. so the wires are correct.

*** EDIT ***

I've found that two of the battery modules I have so far have a DIP socket already in place. I was able to program one in board. the other gave the same issue stuck at verified.  If I pull the chip out it programmed fine.

Module Date was 07 on the one that had a socket and I was able to program inboard

Module Date was 08 on the one that had a socket and I couldn't program in board

Module Date was 14 on the direct solder that I couldn't program in board.

[BillP]

is it possible that on a later hw revision they added a resistor for RO? 

[BillP]

But it was able to erase it (write FF's) to entire memory..  So I doubt it's read only.. Not sure though this is my first time attempting anything like this. haha

[Anonymous user]

Hi 

Could you send me your files, please?

Phil

[Kovalt_apc]

Would it be possible that I can get those Bin files also, starting on this next week and would love to have the head start of having the files.  From any one of you of course it would be great!

[BillP]

you don't need any files. 

you just need to read the old data from your existing chip and then modify it... 

if you read the first few pages of this thread you'd know that. 

[ojfsdi]

What an interesting thread!

I'm running a Symmerta PX with SYBTU1-PLP modules.  We just replaced the SLAs inside 4 years ago and had no idea there might be chips (state information needing reset in the packs.  Some strings are starting to show failed already. 

I wonder if not reprogramming a chip caused them to fail early?

Then again we have had some rough environmentals these last four years so it's probably on us.  I've seen a couple warped SLAs when examining failed packs so far.  One had cells cooking at around 180F when I pulled it out, with gloves!  Took several hours to cool down to the point I'd work on taking it apart.  the two groups of four were stuck 'melted' together.   I felt like I was refurbing fuel rods from Chernobyl with my PPE.

It was interesting to know about the funny spade connectors APC uses with a moving sleeve that actuates a lock mechanism.  I've never seen that before..  We never knew that when we did the original replacement so I should probably replace all of them because they must have been destroyed by pulling on the wire.

[triales77]

I just received my revelprog, in my case the pinout is the opposite.

1. Brown

2. Red

3. Orange

4. Yellow

5. Green

6. Blue

7. Purple

8. Grey

Black and white are not used.

[triales77]

How do you edit the locations you listed using Revelprog?

Thank you

[triales77]

Jeremy

Is it necessary to change the date?

Thanks!

[Tal]

A bit late but i stopped getting notifications on this post once it migrated...

You don't need to edit the date, but it's ones of those things that makes documentation inside the NMC easier.

to make changes in revelprog you select the text you want to zero out. the backspace changes it to zero.

selecting text will display the crc code in the bottom bar of the tool, (which you need for the checksum)

[djb_rh]

Do you need to do anything with the checksum if you don't care about the date and are just editing the battery status? It seems the answer is no from what I've been reading, but I just want to be sure. It seems, though, that the checksum is calculated from data that doesn't, strictly speaking, need to be changed. Correct or not?

[Tal]

I don't think so.

It doesn't hurt to try.

[djb_rh]

I'll give it a try soon. I verified what the instructions say, which is that the checksum comes from the memory area that's got the date but doesn't include the area with the status info. So you *should* be able to do just the status info without it.

I got some sockets and was able to program a chip I removed completely. Just need to solder the socket back in, pop the chip in, and rebuild the pack with new cells (waiting on those) and see what happens. I do have a test clip showing up today, so if this works then I'll see how it works to reprogram in-circuit. I know sometimes people have had trouble and sometimes not. My one theory is that you're putting 5V into the entire circuit via the USB programmer, so it may just be a power source issue. The big thing to try would be to unplug the PCB from wherever it's supposed to be getting its 5V from so you're not backfeeding up the line to whatever that is, too.

Anyway, I've got a ton of empty chassis I had saved over time. Maybe like 10 or more. But I only NEED four working packs at any given time and I try to maintain six working packs at all time (down to five right now). I was willing to pay $350ish shipped for refurb packs, but it seems those have dried up and the cheapest I've seen is $500ish+ now.

[djb_rh]

Well, I plugged my self-refurb pack in and it’s been online a couple minutes and seems happy. Didn’t do the dates or checksum and so far so good. 

[djb_rh]

Does anyone have a part number or link for the same “Faston” connectors that are used on some of the battery wires in these modules? I definitely screwed some up when removing batteries in some of my chassis. I found enough non-messed up ones to get the pack going that I refurbed today, but I’d like to get some more on hand for the ones I’ve screwed up. I have some normal spade lugs which I’m guessing would be fine, but I’m OCD.

[djb_rh]

Faston was bought by TE and I was able to do some digging and find what I’m fairly sure are the right parts. So if you ruined connectors by not pulling by the insulator and instead pulled by the wire, you can replace them. You probably only need the pin and not the insulator, but here are both part numbers in case…

Pin: 790319-3‎

Housing: 154719-1‎

Note that you’ll want a large open barrel crimper capable of doing 10AWG connectors and wire. I have a Pressmaster MCT Mobile Tool Frame for use with 4300 die sets on order along with the 4300-3146 die, which come highly recommended.

For a lot cheaper you can just replace these with any properly sized and insulated spade lug, certainly. I really wanted the above crimper for some other things around my shop, and I really like these style positive lock FASTON connectors, too, or other projects.

[Kovalt_apc]

This was quite some time ago, but as this tread still seem to have activity, I can say that the posts in this tread were perfect, I did 30 battery packs in a weekend without issue.  Quite some time ago....

[djb_rh]

Just a quick update…I broke down five more bad packs today to find all five had socketed memory chips! Super stoked about the time savings that’ll provide.  

I went one at a time popping memory chips out and putting them in the Revel. I’d load the memory, highlight from C9-F4 with the mouse, press 0 (zero), and then write the chip. Literally nothing else. Only tested the one I needed today, but I put ten new batteries in and one of these fresh memory chips, and plugged it in and all is good.

A word of caution: Don’t let your bad battery modules sit around for months and months and years before getting the bad cells out. They’ll swell and leak and make a mess. Go ahead and get them out of there, and be careful with those Faston connectors! I find that even with ones removed properly by pulling the plastic shroud (and never the cable), they’re still “loose” if you don’t also just squeeze them lightly with some pliers to put the tension back in them before re-applying. Could also be that new battery cells have slightly thinner connectors (and I know that in this size battery there are ‘big’ and ‘small’ lug types, I’m using the bigger ones).

Anyway, happy battery refurbing.

[Igli_NYC]

Does anyone know how to print a new asset tag label? Say for a SYBT5 and serial if you need to replace existing damaged label. What type of printer to use and software

[SentinelF]

Very late to the party, but I've figured out that bytes 0x00, 0x02, 0x04, 0x06 of the EEPROM make up the 32-bit floating point value representing the pack's shunt resistance in Ohms, this value is unique to each pack.

This value is visible (only in milliohms) from the front display of the UPS, on Page 4 of the raw status menu (or remotely via an NMCv2/v3) for an SYBT5 pack.

For example:

Page 4/5
000.0 2.00 850;
3500 0 22 32

In this example 2.00 is the shunt resistance in milliohms.

Also agree 100% with what @djb_rh said... don't let old failed packs sit around for months, they will most likely leak battery acid inside the chassis and damage the cables and metalwork, ask me how I know!

[Tal]

Not that it would help with charging control as all packs are in parallel, but the shunt resistance could be changed based on the new packs batteries to be more accurate for load calculations.

[rgrp]

@SentinelF @Tal Im a little new to this, but my values in the revelprog for 0x00, 0x02, 0x04, 0x06 are 00. I am using 9Ah batteries so from my research the shunt should be set to 2.00. 

What would the values need to be set per hex in revelprog? Any pictures of what yours looks like in the revelprog software would be greatly appreciated as im new to this.

[djb_rh]

Just took the time to desolder 6 EEPROMs from boards I had that weren’t socketed. All 6 were able to read fine and write, so I zeroed out 0xC9 to 0xF4 on them. I added sockets to those boards and popped them back in. I then did two more boards that did have sockets, so I have 8 spares. Which is probably enough for the next ten years as I only have 8 total battery modules running at one time. Assuming I have a pre-programmed board available I’m down to only needing about ten minutes to swap in new cells.

I do think it’s worth noting that GreenlightUPS.com is now doing a LOT of stuff for these units, though. They sell a replacement CPU board for 70 bucks for anyone that doesn’t want to program their own EEPROMs, but it should be noted that those boards can NOT be programmed again later when that updated pack dies. You have to just buy a new board. I’m pretty amazed they didn’t just make a board that could be plugged into USB or something to easily reset them, but obviously they want to just sell you another 70 buck board every time.

Interestingly, they currently sell a refurbished APC battery modules for under 500, but have a completely new battery module they designed themselves that are supposed to be shipping soon at just under 800. It’s not clear to me what you get extra for your dough outside of 11aH batteries. 

And they have a new monitoring system you can plug into the intelligence module, but I have it and I’m not sure I can recommend it at the moment. The App seems pretty buggy and crashes easily and often.

Oh, I do also have one of their refurbed power modules, which are 750 bucks and seem like a reasonable deal.