battery replacement for symmetra lx

[Anonymous user]

Hello,

Replaced the individual batteries in our symmetra lx (sybt5 modules). The battery modules are not being seen by the unit. Is there any specific set of steps to have the new batteries check in? Please advise.

Jon

Message was edited by: Jon Reynolds

[BillP]

There is no way for customers to "fix" this. As explained before, the on-board intelligence that flags a battery module as bad won't be cleared when you replace the internal batteries separately. They're not designed to be serviceable in such a sense. You need to purchase an entire new module, SYBT5, in order to assure you receive no additional errors.

[Anonymous user]

Hello,

Replaced the individual batteries in our symmetra lx (sybt5 modules). The battery modules are not being seen by the unit. Is there any specific set of steps to have the new batteries check in? Please advise.

Jon

Message was edited by: Jon Reynolds

[Anonymous user]

We installed new battery cells into the SYBT5 module. Sorry for the confusion. Then I installed the battery modules into the symmetra lx. They show zero runtime. I measured the battery modules and was showing good voltage, approx 135 volts before insertion into the symmetra lx.

[BillP]

Hi Jon,

The modules of a Symmetra UPS are not designed or intended to be user-serviceable. For example, if the UPS flags a battery module as "bad", the module must be replaced with another module. It is not possible to replace the internal batteries in a Symmetra LX battery module (SYBT5). The same goes for any Intelligence (SYMIM5) or Power (SYPM4KP) modules.

The battery modules have on-board "intelligence" which reports detailed status and other information back to the Intelligence Modules. There is no way to clear or otherwise recalibrate this chip in the field.

The UPS is reporting zero minutes of estimated runtime because when on-line and not actively discharging the battery system, the UPS uses a complex algorithm which (among other factors) multiplies the perceived battery capacity in watt-hours by the number of good batteries sensed in the system. Anything times zero is zero. To see a more accurate estimate of available runtime in your situation, put the UPS on battery by opening the Symmetra LX input circuit breaker. Within a minute or two, the UPS will look at the actual rate of discharge from the battery system to give you a better indication of your runtime. This means, as you may have guessed, you may have good runtime with the self-serviced battery modules, but the fault condition of "bad battery" will never clear until real new modules are installed.

To clear the battery faults of your Symmetra LX, please purchase genuine SYBT5 battery modules from any APC Authorized Reseller.

[Erasmus_apc]

Assumed answered due to lack of customer update/response.

[BillP]

Paul,

if a SYBT5 is marked as 'bad' (faulty battery), the information is stored in a EEPROM (8-pin Dual In Line), type 93C66B or 9366-3P. (0,2 $/pc.)
If you have a Programmer (like GALEP), read a working EEPROM, copy it to a faulty tested EEPROM.

Best regards
ZDIS

[Erasmus_apc]

Your statement is a little unclear, sorry but I would just like some clarification. Are you saying you just bought new SYBT5 modules (silver casing with batteries in it) and installed them into the unit? Or, did you install new battery cells into an existing SYBT5 and then insert it into the unit?

[Anonymous user]

On 12/16/2010 5:35 PM, Rob said:

This is not recommended nor supported by APC

That about sums up anything but "give us lots more money" with APC.

When speccing UPSs for customers, other than the small ones that we treat as disposable, in this size range it's more a case of who's least bad when it comes to the built in "screw you" factors like this. Ie, you should know that it's not the tech specs that sells the product, it's the relative lack of options in certain size ranges.

I've had "discussions" with APC technical before - and whilst very helpful on some things, they cannot change limitations built into the system. It's quite clear that the system can measure a lot of stuff that it does not make available to the user. Some SNMP data is lower resolution than what it shows on the web interface, the design of the battery modules is such that the system can monitor current in each battery but this is not available via SNMP (as are a few other measurements) so the only option is to wait until the system says "bad battery".

What's so wrong about allowing the user to monitor battery status (specifically capacity in each pack) ?

And now they are quite happy to say "once it's tripped, you're screwed" so when we fit new cells (I've got 40 sat on the bench behind me) as well as the 10 I've just fitted - well we can no longer even use that very very coarse measure of battery quality and must live with a permanently "faulty" system.

It's so sad when good engineering is crippled by the marketers and beancounters who don't understand the concept of "happy customers". It's exactly the sort of thing car manufacturers used to do before their exemption from the laws on control of secondary markets was removed.

Perhaps it's time to invest in an EEPROM programmer and start decoding the data - I've got a bunch of packs, not all of them have failed (yet).

[voidstar_apc]

For SNMP we try to expose every UPS data item that exists on the web interface.

With precision, it's a little tricky because SNMP only supports whole integers so we have to create new OIDs for each bump in precision. Some of our customers don't notice the presence of HighPrecision versions of the OIDs, but if we're missing some higher precision measurements, then it sounds like we've dropped the ball.

[Anonymous user]

When I enquired a couple of years ago, there were quite a few things in the web interface that weren't in the SNMP data, and a few things that the system almost certainly measures but which aren't reported.

Eg, the Web UI shows output A, output kVA, %W, and %VA. SNMP only reports "%power" and it's not clear if this is W or VA.

Not reported at all :

Input A, VA, W.

Output W, VA, Power Factor.

Any of the battery temperatures or currents - there are OIDs for some of these but no values are returned. The system measures both current and temperature for each battery pack, it can raise alarms/determines condition based on them.

APC support confirmed that the values I've mentioned were not available.

But this is a bit OT for the thread.

[voidstar_apc]

On 4/23/2015 11:57 AM, Simon said:

When I enquired a couple of years ago, there were quite a few things in the web interface that weren't in the SNMP data, and a few things that the system almost certainly measures but which aren't reported.

Eg, the Web UI shows output A, output kVA, %W, and %VA. SNMP only reports "%power" and it's not clear if this is W or VA.

Which UPS is this, a Symmetra? %Power on the latest NMC firmware is the highest of %W and %VA. The latest MIB (you can get it here) has three OIDs upsHighPrecOutputCurrent, upsAdvOutputActivePower and upsAdvOutputApparentPower for A, W and VA respectively from which you can calculate percentages and power factor. While they show up on my UPS, it's possible it isn't working on yours.

On 4/23/2015 11:57 AM, Simon said:

Not reported at all :

Input A, VA, W.

Output W, VA, Power Factor.

It depends on your UPS, but typically current is only measured at the output. Input current and efficiency are usually calculated values. Once I know which UPS we're talking about, I can see if battery temperature is available. Just from a brief look, it doesn't seem like the symmetra UPS provides battery temp measurements to the NMC.

[Anonymous user]

I have OP current in 1/10 amps, but I don't get anything for active power or apparent power in the upsHighPrecOutput group.

I did find OP VA (but not W) in upsPhaseOutputPhaseTable.

I have IP V (in 1/10 volt) and IP Freq (in 1/10 Hz) - that's all that's reported in the upsHighPrecInput group.

I have battery voltage (1/10 volt), capacity, and runtime - but no battery currents at all.

TBH, I'm more or less resigned to having to install external measuring kit for the IP and OP values. But that's not really practical for the battery measurements.

It's a Symmetra LX 16 RM, with AP9619 Network Management Card (on which the firmware was upgraded about 18 months ago).

[Anonymous user]

On 12/16/2010 6:35 PM, Rob said:

This is not recommended nor supported by APC. While this may work in a few cases, not every SYBT5 module is programmed the same, so you may end up copying bad info to the new chip.

i would rather burn down 100 pieces (tries) of wrong programmed EPORMs for 20cents per piece then spend 600-800US$ for 1 working module :))

if you would provide replacement eprome modules for like 40-50$ then people mignth not be interested in testing dozens of eproms programming till they find the right one....but we all know thats not the case and never will be....so i am sure how anyone with at least basic knowledge about UPS eproms and required tool will do this trick for sure, rather then spend a month salary for something that realisticly worth  just few bucks...

[Anonymous user]

awesome comment!!! bulls eye....straight to the bone....

its not about happy customer, but about happy seller while customer is happy untill everything goes well....when it goes down, customer will pay a high price for his happines....

lets be honest....APC is not here to make us happier, so this "technical support" in most cases is; "problem might be that, but we suggest to send UPS to official service...to buy original....etc..etc..."

it is like; we sell you a APC brand car that consumes a regular gasoline....but its designed to consume APC gasoline that can be bought only at APC gas stations and this gasoline is 300% higher priced then regular one that actually is completely the same like any other....but has no APC mark on it :)))))))

battery is battery....there is no APC battery, VARTA, DURACELL, FIAMM battery....Alessandro Volta made a principle od battery 200 years ago and battery concept has not changed since then, now we got some "smart" batteries...they are not smarter at all, they are just smart locked to monopolize the consupmption....so, eprom blasts, throw 10 completely OK batteries and buy a new battery pack so APC can make a profite of 600% - 1000%, while on other hand they are providing some "ECOLOGICAL AWARENES" ....SAVE THE ENERGY, DONT WASTE THE ENERGY, GREEN POWER, bla,bla...while their faulty eproms makes a tons of garbage and toxic waste that can be still used  and energy saved....

what a hipocracy.....almost feel disgusted...but thats not the case with APC only....its world wide like that, even about simple thing like toilet paper...

i have a Symmetra battery packs that went eprom wrong with brand new battery....OK, now 10 months old (but thats considered as "new" still since batteries can hold up to 5 years) and by APC ideology of "SAVE THE PLANET-GREEN POWER" i am about to do a paradox by wasting completely OK batteries and buy a new battery packs again....isnt that hypocritic?!? please tell me that is RIGHT thing to do if anyone dare....yea, right...mhmmm

GREEN POWER and SAVE THE PLANET is just a fency cloacked  marketing byword that menas abolutely NOTHING to corporations and their wallets....well, good luck with it....few more generations and their grandsons will enjoy the "GREY POWER" despite the all the billions they will make meanwhile....it will be like a lottery won of 10 millions of US$ in the middle of Sahara desert, with only a drop of water :))

[Anonymous user]

My plan was to read the EEPROM of an "OK" module and then run it until it went "bad", then compare the contents. I suspect there's other stuff in there - such as calibration for the current and temperature sensing, plus perhaps some history - and of course the serial number and production date.

By comparing a number of EEPROMs, and before/after versions, it should be possible to work out what's what. Worst case, if I've got a copy of the contents of all the units, I can just put the old data back after replacing cells.

I've just hit a bit of a roadblock with the reader. I believe I should have a working setup with a Raspberry Pi, a breadboard, and some resistors - but there's just too many unknowns (I'm fairly new to Pi, not used SPI before, not sure if I'm using the right software, and my electronics knowledge is a bit dusty these days) and I haven't had time to go and talk to some poeple I know who could probably help.

It's clear that APC work on the "razor blade model", in terms of extracting ongoing revenue from users.

If they priced their replacement batteries at something closer to real market prices then there'd be less incentive for people like us to go DIY. It's certainly been an "interesting" learning curve since we got this UPS - and it's lucky we got a good deal on an over-specced second hand one since we've had two power module failures in the last year as well.

[Anonymous user · Accepted solution]

On 8/6/2015 11:47 AM, Vlad said:

APC is not here to make us happier

Well that applies to any business - they are only trying to "make customers happy" to the extent that they need reasonably happy customers for the business to survive. So it's a matter of working out where the sweet spot is between keeping costs down (good customer support is a cost) while keeping it "high enough" to keep customers coming through the door.

To be fair, we did have a problem and APC support were exceeding good. Eventually it was determined that the comms module in the external battery cabinet was faulty and they replced it FOC - I suspect level of service varies by region.

For comparison, when we first bought a UPS we went with British Power Conversion who were half the price for an equivalent unit. The fact that we now have an APC unit (which we could only afford second hand) probably tells you all you need to know about the reliability of the unit and their level of "support".