PINNED - Lithium battery conversion/info

[Gnomatic]

All cells new, purchased in January this year from evassemble. The only crimped-only connectors on the pack are the ring terminals on the ends of the nine 22 AWG balance wires.

I'm pretty certain each bolt on the pack got tightened with my 5Nm torque wrench. But I'll hit them all again just to be sure.

[Burgerman]

Solder the crimps... Check connectors. Check bolts, recharge. Then do a discharge with the PL8 alone to about 50% and recharge. Theres a bad connection somewhere. Looks like a bolt to 1 cell. Nobody believes it till they find about 3. Also theres no heatshrink in the way?

You dont have discharge balance enabled? Hopefully.

[Gnomatic]

1. There are the terminals I used on the balance leads. So I guess I'll need different ones if I'm to solder .

2. Why discharge with the PL8 alone, just curious? I can do it but it will take a long time vs my my load dump.

3. I'll double check the bolts, no heatshrink in the way.

4. I only discharged so far with the load dump, and just used the PL8 to monitor. If I use the PL8 to discharge, I'll be using the preset you posted here for me a couple days ago for my pack. So I assume the discharge balance setting isn't enabled in that preset. But will verify.

5. Should I charge again first once all connections and bolts etc. have been checked and THEN discharge?

[Burgerman]

Yes.

[Gnomatic]

Terminals like this will be easier to add solder to.

What is the advantage of discharging the pack with the PL8(which is limited to 100w) vs. using my 600w load dump to discharge while the PL8 monitors?

[Burgerman]

I want to see if the pl8 shows the same different voltages across cells as it discharges. So I can figure out why. If all is good, no bad connections, no big connector losses or anything then it shouldnt make any difference.

[Gnomatic]

I want to see if the pl8 shows the same different voltages across cells as it discharges. So I can figure out why. If all is good, no bad connections, no big connector losses or anything then it shouldnt make any difference.

Thank you. I will take out ~50Ah with the PL8 starting tomorrow.

[Gnomatic]

BM, would you prefer I take the 50Ah out of the pack in one shot, or am ok to break it up into two discharge sessions over a couple days to take out a total of 50Ah? (I'll still post both PL8 graphs if you're ok w/ me doing the latter)

[Burgerman]

Doesent matter.

[Gnomatic]

Think I found the problem. On the bottom side of my pack, Cell Group 6 the entire row, bolts were loose. I must have skipped that row when I torqued the bolts on initial assembly.

(I checked the rest of the bolts while I was at, a couple of those turned a little bit before my torquer clicked)

[Burgerman]

I lost count of the times that I did mine. And STILL found one or two I had missed! Thats when I started marking as I went, with a blue marker pen, after each one, then a black one 2nd go... And STILL found one I had missed on a final check with no marks! Its almost always the problem.

Charge full. Retest.

Q. How do people with no PL8 or Hyperion and PC software, ever find these issues? Or worse, BMS?
A. They don't. And get battery issues, fires, and major issues after a few months. Every EV forum, is FULL of failed cells, battery balance issues, etc. All give lithium a bad name. When the reality is that the problem is the users. Thats why many places only supply EV companies. Too many problems and returns. And why they all want to sell BMS.

These things MUST be assembled with great care and accuracy. Then problem free unless you either get:

Over discharge (once is enough!)
Over charge (again once for a few secs is no problem, but holding them at 3.65V for hours causes shorter life). Thats why we are charging at 3.550 or 3.600V Max.
Also best not to store long periods full. Or empty. About 50% is best. Non critical, 30 to 70% is fine.
Short circuit (too high current)
Too hot! Cool place, above 0C = longer life.

[expresso]

Think I found the problem. On the bottom side of my pack, Cell Group 6 the entire row, bolts were loose. I must have skipped that row when I torqued the bolts on initial assembly.

(I checked the rest of the bolts while I was at, a couple of those turned a little bit before my torquer clicked)

thats great news - i am sure thats all it was and will be -

[Gnomatic]

not perfect but much improved .......

[Burgerman]

Its as good as you will ever get. It varies more than that cycle by cycle. And will improve with a few cycles as balance gets closer.

[steves1977uk]

Just read over on RCGroups forum that firmware 3.34 is available for the PL8, but no autoupdate from CCS so far for me. Wonder if this is a separate firmware that's not included in CCS?

Steve

[Burgerman]

Its a mistake...

[Gnomatic]

BM if I discharge the pack with the PL8 preset you gave me, will the charger automatically stop discharging once a cell reaches 2.9v?

[Burgerman]

yes. As soon as the first cell reaches that figure.

[Gnomatic]

Picked up a cell checker like this one.



What type of connector plug do I need for this?

[Burgerman]

Stock 8s hobby XH one.
http://www.ebay.co.uk/sch/i.html?_from= ... e&_sacat=0

[Gnomatic]

I discharhed the pack with the load dump most the way(was monitoring w/ PL8), then finished discharging w/ the PL8. I watched, just being sure nothing dropped below 2.9v. Then charged.

Anyway, here's the recharge graph. Only change I made to the preset was upping the CC voltage to 30A, to finish up a bit faster.

[Burgerman]

Strange that they are not almost exactly together, but looks ok. Maybe long charge cables. At least cell 8.

[Gnomatic]

My charge cables are long.

[Burgerman]

Well that can do it.

[shirley_hkg]

IS it normal that it shoot up to 3.561V , right after stop , that set point is 3.555V ?

May be 32A current is high .

[Burgerman]

Theres three voltages.

1. Voltage at charger while 32A flows.
2. Voltage of cells while 32A flows, which is false or rather not the voltage that is important. Only resting voltage without current matters. The charger knows this but doesn't display it.
3. Voltage at charger, and at cells is different depending on cable length and resistance and current.

Charger knows all three voltages! And more importantly charges in pulses with breaks where it measures off load voltages many times per second, so it knows what is really too high. So displayed voltage is actually an average of the voltage at charger (rather than cells) and during charge and short breaks. Rather than the off load/charge voltage. So don't worry.

[Gnomatic]

Well, no way to sugar coat this - I made a real boneheaded mistake. Result? I need to replace 2 cells on my main pack, which means tear pack apart and rebuild.

WTF happened, you say?

My straps came in that I'm using(with my buddy doing the lifting) to lower the pack into the battery area. We cut them to size and they make it much easier lo lift and maneuver the pack, so I decided to do a quick test fit. Above and inside my battery area there are two bolts with nuts on them on them for the shocks. There is room for the pack to clear the bolts sticking inside above the battery area. But not much.

So as my buddy was lowering the pack in place, he didin't keep the pack perfectly level(and by no means do I blame him for my mistake), and a cell on each side of the pack(+ and -) scraped against a shock bolt and compromised the the blue protective coating. Snap, crackle, pop. Sparky sparky. So, I need to replace 2 cells.

A million ways I could have avoided this. Remove bolts before lowering pack in, put a sheet of rubber over bolts, etc. etc. ect. Like I said, boneheaded move on my part. But what's done is done.

I took my chair base to a local shop, and am having those shock bolts welded in place to the outside the battery area, and then the parts that stick out on the inside grinded off. And, once reassembled, the pack will be wrapped in a sheet of rubber before it goes in.

Big thanks to expresso for offering to send me a couple of his spare cells and orange blocks. That will help me get started with the fix much quicker.

[Burgerman]

Oops. Hopefully no other damage. Shorting cells can hurt them.

[expresso]

BM - i found myself reading on your site in the BMS section - you say not to charge to the max. 3.6v - the preset i am using does just that -

is it Ok to use it or should i change it to a lower charge per cell ? -


http://www.wheelchairdriver.com/BMS.htm

[Burgerman]

I actually said the following will harm them:

1. charge any cell to above the 3.60 volt maximum. 3.55V is fine. They will live a long life. 3.65v is going to shorten cells life quite drastically. During charge, the charger or charging / cell balance system should NEVER allow an individual cell to exceed 3.60 volts on a LiFe cell for long service life. A typical generic BMS setup does this repeatedly for hours or days at a time.

Accurately at 3.60V and terminate soon, is fine. A BMS bounces the cell over this, regularly for many hours due to weak balance circuits and no ability to control charger power other than on/off.

YOU CAN STOP READING HERE! Or not...

Lithium likes to be at the middle. 50% charged. Thats where it lives the longest. Empty or below 30% is said to be worse. Stored a long time it needs to be below 70% charged. In the same way that lead likes to be 100% full. The highest or lowest voltages are where the cell damage occurs fastest. Theres no set fixed point.

Charging to the manufacturers recommended 3.650V is a little high. They say this is the best voltage mostly because a BMS has a very weak balance circuit of 50 to 100mA usually. So charging to a voltage higher than the cells 100% full level means a small amount of self balancing takes place to help. Cells dont like being this high voltage. Its like breathing in, really deeply and holding your breath!

So why recommend it? Most dumb users packs die from imbalance due to crappy BMS and uneven self discharge. Charging to too high voltage balances cells, by overcharging all. So they get less back under warranty. Manufacturer sees a lot of cells returned because a pack gets out of balance and eventually one group gets overcharged too much or because of a low group, one set runs below the safe point in use regularly and dies.

Think of it like this. Each cell can only hold so much power. Like a brick soaks up water. Once that cell has soaked up as much electricity as it can possibly hold then all the rest just goes to waste. The waste, just causes damage (actually it causes lithium plating internally and ruins the cell over time).

So you can either use accurate but slightly lower voltages and a really good powerful cell balancer and clever charger that throttles its power to make sure each one is equally full without any overcharge. To 95 to 99.9 percent. -- Or, you can just keep pouring power in power above the full level, until all end up full regardless. Cells all end up full, so they self balance to a small degree... The first way is what happens when you use a good charger. The second way is what 3.65V (+/- 0.05V) does with a typical BMS.

So actual voltages...
Your cells will end up charged to 99.99 percent Even if charged at 3.500 volts for long enough, but balancing would require a very long time. And so a termination point of say 1000thCwould be needed. And holding them at 3.500V for extended periods while the balance happens. So only use this if your charger never terminates, like the Hyperion if you have a lot of time.

3.550V Is faster, and they also end up full and is OK for a good, relatively new pack, and balance happens pretty fast with a PL8 but quite slow with the old Hyperion as its balance current is 3x less than the PL8 (still 3 to 6x more than a BMS). I use this, as its adequate for a very healthy pack. It theoretically should be better than 3.600V for longevity. Its a compromise. (Your lead/add-on pack uses this too. Its set to 3.525 I think. 28.200V. To be nice to the lead, and to not damage the lithium over long periods as the slow termination charges the lead properly).

3.600V is OK too - still lower than the 3.650V advised by the manufacturer. And so the cell isn't overcharged. But balance is faster, and stays balanced better on a less than perfect or older pack and charging is faster to end.

So you take your choice! I would use 3.550 for a new pack and if not in a hurry. 3.600 for most packs and for any user not watching a screen to see what is happening. And even 3.650V on the odd occasion that cells were uneven slightly to be sure every group was full. Then revert back to 3.550.

Its import and to note that balance will be slightly different at each voltage. So the first charge after changing it may look odd as some minor re-balancing will occur...