Should I sell my LifePO4 and get AGM?

[bquick]

I have a Ninja 250 that I converted this summer. Everything is working well, but I'm struggling with the batteries. I ordered batteries directly from Headway, 80 cells to make a 40 ah pack, 20s4p. I'm charging the pack with two FMAdirect 10S balancing chargers. The charging works well. I was working on a microcontroller based LVC. It works, but as soon as the speed controller kicks in, the microcontrollers go crazy (random code execution). My controller has a LVC, but I don't have cell level LVC at the moment.

After a number of cycles, I had one group of 4p cells go to zero volts. Headway replaced those for me free of charge. A while later, I had another set of 4 go to zero. Headway didn't answer my email the second time, so I replaced those 4 cells with 19 A123 cells. This works surprisingly well, but it's expensive. I went for a ride today and in less than 10 miles, another set of 4 cells went bad. When I first started, I could get 24 miles out of a charge.

The second set to fail were on the bottom of the pack. They may have gotten too much pounding. This last set to fail are either on the top or the bottom, I'm not sure yet, I just know it's the last group in the chain from the charger.

I don't want to keep wrecking LiFePO4 cells. I'm seriously thinking of selling the remaining good LiFePO4 cells and switching to AGM batteries. I figure the AGM can take more physical abuse. I know they are heavier. I have three 24V chargers that are just waiting to charge a 72V pack.

Am I nuts? I value your opinions.

- Brad

 

[michaelplogue]

You ought to switch over to the new BMS kit being sold on Tppacks.com devloped by GGoodrum & Fletcher

http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=5416&st=0&sk=t&sd=a

http://www.tppacks.com/products.asp?cat=26

 

[pgt400]

Brad, these are the agms I'm using. They were $384 delivered for (6)...24lbs each. I have over 600 miles with no problems other them reduced capacity in the cold (not today though, 70+ in NJ). I hope to go to PSI cells in the spring.. but in the mean time these are fine....and if one goes...big deal at $52. I am using 3 cheap, 24 volt 4 amp on board scooter charges. How many miles did you put on so far? I here the newer Headways are supposed to be better....we will see, as the big group buy is about to get their cells.

Phil

http://www.powerstarsla.com/orth_lifestyles_wheelchair_battery.html

 

[swbluto]

The high switching of frequency of the controller might be causing ripple voltage to be passed down the line to the BMS, which could cause such randomness as you've described. You might want to add a low-pass filter onto the BMS's output to eliminate that potential problem.

Also, make sure to wire all of the chip's inputs otherwise that might also cause noise.

 

[Lessss]

How many amps do you run at?

 

[bquick]

When cruising, I'm drawing 70 amps or so. I can draw over 200 amps when accelerating hard. I think my Kelly 400 amp controller is the limiting factor. I don't think it really puts out 400 amps.

Yes, a working BMS would have saved some of my cells, but won't solve my problem. The BMS doesn't keep the first cell from failing, it just keeps it from taking others with it.

Phil, we should get together sometime. I'm in NY, about 2 hours north of the city. It would be nice to see how heavy your bike feels and how the acceleration feels compared to mine.

- Brad

 

[swbluto]

The BMS doesn't keep the first cell from failing, it just keeps it from taking others with it.

There are BMS's that actually prevent that from happening. Hate to break it to you.

 

[bquick]

I'm sure you sincerely hate to break it to me.

Tell me how a BMS will save a failed cell. I have a 10AH cell that is fully charged. It loses it's charge after 5AH are drained from it. How does your BMS turn this back into a 10AH cell?

 

[swbluto]

My sincerity knows no bounds. As does my insincerity.

A BMS won't save a failed cell, that you're right. But it should prevent the electrical conditions that induce cell failure which it seems like the majority of your failed cells have undergone(The same row going out multiple times? There's some specific cause to that and it might be thermal in nature. And, also, the cells reading 0 v? a BMS is supposed to trip its LVC before it gets to that point if it's wired correctly.). Now if you're working with crappy cells to begin with, that's another story...

 

[dogman dan]

Maybe replace that cell group with an agm? and keep the discharge depth lighter? And do try a better bms.

 

[pgt400]

Brad, My batteries are about 140 lbs. My bike weighs just over 300 lbs (same as stock ex250). Almost 100lbs lighter then my Yammy FZ6. My acceleration is excellent, I'm at 72 volts now but am adding another 6v 35 amp pack this winter and going down to a 12 tooth drive gear. I'm a firm believer in getting my power from volts and not amps.... Volt up, gear down. 80v is about all the Kelly and Mars will stand. My 50 mph cruise at 72v is 50 amps.

I almost bought the headway cells after seeing your bike on youtube....decided to use AGM's first to see how the Lifepo4's early adopters fared. I am happy with the AGM's so far...relatively cheap, powerful, bulletproof, gives me 18-20 mile range when geared for 50mph. I do lose range on cold mornings....but so do Lifepo4's. Fortunately I lose interest in bike riding under 55 degrees.

From what I've read, your cells seem to fail around 50% DOD, and you balanced charged. Were the other cells well above 2.1? I don't think the Kelly 400a can supply more then 300 amps.....maybe the headways don't like 75a bursts? I read were some headway cells varied 30% in capacity, this alone would explain your situation with 4p cells. True, single cell LVC monitoring may have helped to catch and save cells....pita to wire up 80 cells though. Probably best to cycle, grade and match cells that are to be used in parallel.

I would like to give the PSI cells a shot in the spring if the price is right, they seem very robust. The Thundersky 60ah pack may be good too...also much easier to monitor 24 cells instead of 96! I may even try the Silicon AGM's, the newer Greensavers sound pretty good.

Yeah, would be great if we live closer....lots of charging required between RI & NJ !

Phil

 

[pgt400]

Hey bluto, nice job on the ev calc program! Any chance you could include the Mars Motors is there? I guess if I can get the specs from them I could just punch them in?

Phil

 

[fechter]

I guess it would be good to know why the cells are failing.

If the present LVC is not doing it's job and allowing cells to fall below the allowable limit, then the cell can fail. Damage could be done in the previous cycle and show up in the next cycle.

It could also be that the discharge current is too high for the cells and that causes them to fail.
Over current on discharge typically causes cells to expand or vent, which should be visible.

 

[bquick]

There is no visible indication of cell failure. I will start with all cells fully charged. My balancing charger makes sure they are all at 3.65 volts. I'll go for a ride and notice I'm starting to lose power when I'm at 50% discharge. I'll put it back on the balancing charger and all cells will read 3.2xx volts and the failed one will read zero, or .5 volts.

I don't really want the extra weight, but I would like to have bulletproof.

- Brad

 

[pgt400]

A 72v 40ah headway pack would weigh about 90lbs, thats 55lbs lighter then my agms. The weigh is held low though so the bike feels very light (at 308lbs it is light really). The gas version would have 30lbs of fuel held high.

 

[pgt400]

So I guess there are 3 choices...

1). Keep going and likey lose and replace cells..
2). Break apart the pack, test cells and make 4, 20 cell strings and LVC each cell at 2.1 volts
3). Sell off the headways and replace with ??

 

[fechter]

There is no visible indication of cell failure. I will start with all cells fully charged. My balancing charger makes sure they are all at 3.65 volts. I'll go for a ride and notice I'm starting to lose power when I'm at 50% discharge. I'll put it back on the balancing charger and all cells will read 3.2xx volts and the failed one will read zero, or .5 volts.

I don't really want the extra weight, but I would like to have bulletproof.

- Brad

The BMS should never allow any cell to get below 2.1v or so.
If you're reading zero or .5v, that's low enough to destroy a cell.

 

[pgt400]

I notice the A123 & PSI guys have no trouble running 1 BMS channel to each parallel cell bank (instead of 1 monitoring each P cell). Makes sense that P cells would stay balanced.

 

[bquick]

Yes, that's the way I set up my BMS. Unfortunately, I was never able to keep it running.

- Brad

 

[Kurt]

I run the same brown cardboard warped headway cells in a 20ah 36v configuration with headways supply's bms/vms. All though I only run the pack with a ebike and x5 motor not a emotorbike. Even so I often pull a steady 50A from the pack when climbing steep hills and have seen some higher peaks when taking off from a stand till.

At 20ah its 1/2 the AH capacity of your pack so its like pulling 100A from your 40Ah pack. I have a couple of hundred Km on the pack and I am monitoring and logging all trip and charge/discharge data. Everything is running very and the pack gives me almost the full 20ah before lvc pulls the pin.

I was a little concerned about my headway purchase when I read some of the feedback on them. story's about cell heating,voltage drop,dead cells.I can say my battery has never got any hotter than room temp and its the same with the bms. As for voltage drop yes there is some but I think its about on par with the specs of the cells.

The only negative thing I can say about my pack was the spot welding on the tabs wasn't the best.I think they could have cranked up the amps a little on the spot welder to get better penetration and because of this I had one tab that wasn't weld on every spot weld.I had to solder the tab to back it secure and all is ok now.

I am not sure why you have had such bad luck.I would advise using a good bms.I am sure if my pack was double the size I could pull a steady 100ah without any hiccups.
kurt.

 

[monster]

is it a problem with the cell insulation or short circuiting? if its only happening on the botton thats where all the whight is focused and it can grind the insulation and cause shorts.

 

[pgt400]

Option 4) Build a Konion pack (don't need to be balanced) at 72 Volt and 50 amps (500 amps peak)
Bikeraider

Hey BK, isn't this the sme as option 3?

Were can you get Konion cells? Aren't they only 2ah ? I would need 480 if so!

 

[pwbset]

Were can you get Konion cells?

Therein lies the problem. I'm having a heck of a time trying to get just 140 cells for a pack I want to build. DoctorBass sells reject Makita packs with 8 good cells in them for $20 sometimes when he has extras. Unfortunately even that's too steep for my broke ass (140 cells = 18 reject packs @ $20 = $360.. sure that's cheap all things considered, but that's still a chunk of change for a working class stiff living paycheck to paycheck).

 

[pgt400]

Are they the same cells drewjet used? If so, he claims to be down 50% in is range after only 1000 miles? They do sound like good cells, but I still like PSI for their larger form factor, very robust cell with very high cycle...I don't want to deal with 500 cells! Unless I got them for free like Docbass, I would definately try them! Anyway, where can you buy them?

Phil