Multistar lipo 10ah + 50a draw = damaged pack(resolved)

[brumbrum]

Edited to say, the pack is fine, it was a controller issue. I have a modified lyens controller that must use a special equation for setting the lvc. I had the lvc too high and it activating regen at 59v, this was causing the pulsing/pulling back effect

Been using a 16s 67v 10ah pack over the past few weeks on my commuter with a max draw of 25a, more like 15a when coasting.
Today i got my magic pie back on the road its a bit of a beast and dialed in 50a which it took continuously on an uphill trail.
After 10 minutes with a fair amount of voltage sag at 50a the pack was down to about 60v and things started going wrong. The max draw became about 35-40a and the motor was pulsing/oscillating. Continued riding for another 20mins like this. Came home recharged the pack and they would not charge to 4.2, but about 4.15 per cell. Went out again and once again 50a draw, all good for 10mins. Then the pulsing started again. Came home recharged pack and each cell would only charge to 3.9 / 4.0 volts before my hobbyking quattro balance charger was saying they were at full capacity.

I know i have used them at 50% of their discharge rating( i think) and i expected a lot of voltage sag, but i did not expect a damaged pack.

Anyone have any thoughts? Have i been really stooopid?

Would like to add that all the cells charged and discharged equally during both rides as i checked before and after with a cell alarm.

 

[Punx0r]

IIRC the bench testing Icecube performed on the Multistar cells indicated a maximum continuous discharge capability of 3C. However, it's surprising you encountered significant permanent damage from a partial discharge at 5C. The pulsing of the motor is also strange, I can only guess it was caused by the controller, perhaps rolling back output due to low input voltage, which then recovered and formed an unstable cycle.

 

[brumbrum]

I didn't even run the pack down to a low voltage. 59v being the nominal voltage for the pack which is as far as i took them.

 

[amberwolf]

recharged the pack and they would not charge to 4.2, but about 4.15 per cell. Went out again and once again 50a draw, all good for 10mins. Then the pulsing started again. Came home recharged pack and each cell would only charge to 3.9 / 4.0 volts before my hobbyking quattro balance charger was saying they were at full capacity.

https://endless-sphere.com/forums/viewtopic.php?f=3&t=41572&p=1102733#p1102733

It's been 5 years since I started this quest of building

<snip>

...along the way I discovered:

<snip>

-No matter how I hooked up R/C LiCo (also known as LiPo) batteries, they almost universally loose capacity within the first few weeks of operation, stabalize and maintain that capacity during all but the final few months of their life-cycles. Detailed operation boils down to: under charging, under discharging, capacity exceeding C-discharge requirements by a minimum factor of 2.

I don't have any personal experience with *new* RC Lipo to say if this is common or not, but I recall reading similar things in other's threads over the years here on ES.

Not as bad as your problem, usually, but I suspect something else at play in your pack (bad batch, some other "global" pack issue that isn't easily seen in the info so far posted here about it, etc)

On the pulsing issue, maybe there is a setting in the controller that is limiting current below a certain voltage--this has been documented recently in another Magic Pie (MP4, I think) thread.

 

[brumbrum]

Thanks for the thoughts, any clues will help me get my head around this. The controller used is a lyens 12 fet. I have already emailed Edward Lyen about the pulsing to see if it could be an lvc issue as the controller has a jumper lead for 54v and 84v, but i was using the controller at 67v. I had reprogrammed the controller to suit the pack and was told by Edward it would be fine to use, but as said may be Edward could shine some light on things. The only thing i can think is that the pulsing was the pack trying to deliver the amps in bursts.

 

[TheBeastie]

Came home recharged pack and each cell would only charge to 3.9 / 4.0 volts before my hobbyking quattro balance charger was saying they were at full capacity.

I know i have used them at 50% of their discharge rating( i think) and i expected a lot of voltage sag, but i did not expect a damaged pack.

Anyone have any thoughts? Have i been really stooopid?

Would like to add that all the cells charged and discharged equally during both rides as i checked before and after with a cell alarm.

You could consider a parallel balance charger to charge up your cells. http://www.banggood.com/RadioLink-CB86-Plus-6A-210W-Professional-Lipo-Balance-Charger-Discharger-p-972541.html

 

[brumbrum]

Ok folks, time for me to stop scare mongering about these multistars.

To answer my own question in the opening post........ I've been really really stooopid, lol

When i charged the pack (both times) and lost capacity the batteries were straight off the bike and were really really warm. This must have caused an internal resistance or something( i am no wizard). Today the pack balanced up fine(charged from cold).

I went out again at 50a discharge and the bike was awsomeness in every way until i hit 58-59v when the oscillating/pulling began again. I presume this MUST be a controller issue.

The multistars from 67v (4 packs of 4s 10ah) came down from full charge to around 63v quite quickly on the 50a '5c' discharge and then settled soon after to 60v, then after it went to the nominal voltage of 59. Then when hitting 58.5 or there abouts the 'pulling back' started. I am still waiting for Edward Lyen to get back to me about this happening.

 

[brumbrum]

Came home recharged pack and each cell would only charge to 3.9 / 4.0 volts before my hobbyking quattro balance charger was saying they were at full capacity.

I know i have used them at 50% of their discharge rating( i think) and i expected a lot of voltage sag, but i did not expect a damaged pack.

Anyone have any thoughts? Have i been really stooopid?

Would like to add that all the cells charged and discharged equally during both rides as i checked before and after with a cell alarm.

You could consider a parallel balance charger to charge up your cells. http://www.banggood.com/RadioLink-CB86-Plus-6A-210W-Professional-Lipo-Balance-Charger-Discharger-p-972541.html

Nice charger btw.

 

[Punx0r]

The battery getting hot isn't good. I think you're just running them too hard.

 

[TypeThree]

Do you run lipo alarms on your balance leads whilst you're doing your trouble shooting or have you got some other diagnostic device to tell you whats going wrong with your cells at the time you're getting the pulsing?

I was running a monster RC setup the other day which I built 4-5 years ago as a 10s2p. The batteries (4x 5s hyperion 35c G3 lipos) had been on storage charge for nearly 4 years which is no good for lipos even of that quality and had the exact same pulsing thing happened when I discovered two of the cells had gone bad. They seemed to come off the charger OK but under load they died and wouldn't recover. The pulsing was just the LVC in the castle controller kicking in. Put a couple of fresh packs in that I knew were good and it was fine, no pulsing.


(If anyone is interested to know what I'm talking about when I say monster RC. https://www.youtube.com/watch?v=xY81U_Y5XX8 and this is under the skin https://www.youtube.com/watch?v=tBNSjmyPp_o

 

[brumbrum]

Do you run lipo alarms on your balance leads whilst you're doing your trouble shooting or have you got some other diagnostic device to tell you whats going wrong with your cells at the time you're getting the pulsing?

I was running a monster RC setup the other day which I built 4-5 years ago as a 10s2p. The batteries (4x 5s hyperion 35c G3 lipos) had been on storage charge for nearly 4 years which is no good for lipos even of that quality and had the exact same pulsing thing happened when I discovered two of the cells had gone bad. They seemed to come off the charger OK but under load they died and wouldn't recover. The pulsing was just the LVC in the castle controller kicking in. Put a couple of fresh packs in that I knew were good and it was fine, no pulsing.


(If anyone is interested to know what I'm talking about when I say monster RC. https://www.youtube.com/watch?v=xY81U_Y5XX8 and this is under the skin https://www.youtube.com/watch?v=tBNSjmyPp_o

The batteries a pretty much new. I have used them lightly for a couple of weeks ruuning them at 25-15amp loads only down to the nominal voltage. I have a hobbyking balance charger which of course balances each cell and is monitored. I then always check the cells with a lipo alram before and after using to check the state of each cell which all stay in balance with each other. I may get some more alarms to put on each of the 4 packs whilst riding, just incase there is a major sag going on which could be causing the pulsing. This may be the case, but i think due to my controller being configured for either 54v or 84v and i am now using it at 67v which sits right in the middle, i feel it must be to do with the controller lvc. But as you say, i think spending a few british pounds on some lipo alarms is a good idea. I use a cycle analyst whilst riding, although this only tells me how the pack as a whole is running. I have unplugged the CA to see if that is the cause of the pulsing, but it still then happens.

 

[neptronix]

What's the LVC of your controller?
How many volts are they dropping under load?

You cannot say your pack is damaged without this info..

 

[brumbrum]

What's the LVC of your controller?
How many volts are they dropping under load?

You cannot say your pack is damaged without this info..

the pack is not damaged, as far as I can tell, it just would not charge to capacity while it was warm - straight from use, after pulling 50a continuously with the magic pie. Once cold it charged and then discharge fine.

The lvc of the controller is a bit complicated. I have programmed it myself for a 45v cut- off (which is low, I know), but the CA is programmed for 56v lvc (3.5v per cell). I tried lowering the cut-off on the CA to see if that would make a difference, and it didn't. As said above, the pack goes from 67v - 63v within the first 5-10mins, then stabilises at 60-59v (which is the nominal value for the 16s pack). but then at round 58.5 the pulsing starts. The Lyens controller has a jumper lead to change the LVC for use at 54v or 84v. The bike will not power up with the jumper plugged in(54v), it works with it unplugged at 67v(though it is set up for 84v), but I assume that Lyen must have put in an LVC for this higher voltage which is being hit as the 67v battery drains.(just an assumption), Edward has not answered the email I sent on Sunday as yet.

I have order some lipo alarms which should be here in the next week to use in situ, to see if any cells are sagging badly under load. at rest they are all balanced (before and after use) as I have one lipo alarm. I can try tonight swapping the alrm between the packs 4 balance leads to see if major sagging is happening.

The only concern I have at present is how warm the pack becomes at this 50% discharge rate. Its very 'warm' all over , not hot.

Is it normal for a high discharging pack to get warm? this is the first time I have discharged at this high rate.

 

[neptronix]

Becoming warm and seeing lots of voltage sag at 50% the stated C rate is pretty much the norm for any lithium battery.
That voltage sag will cause you to hit LVC sooner, and thus you will never get full capacity at high loads, unless you like, just kinda tickle the throttle and/or pedal a lot

The actual point where you start seeing no temperature increase, super low voltage sag, and full capacity is around 1/4th of the stated maximum continuous C rate; unless we're talking about newer NCA 18650 cells, where it's a bit lower than that.

Your battery probably does not last very long at 50A continuous anyway.. if you added another 10AH in parallel, it'd perform really well.

 

[brumbrum]

thanks for the reply Nep.
Yep, I have another 10ah pack of the same multistar / same pack configuration, but find carrying that extra load just too much of a ball ache, plus I do not need the extra range at present, but I do understand the extra ah equals lass sag and better performance as well as better range. This is a commencal downhill mtb with the 10ah pack in the frame made for quick blasts up the local trails, along ridges and then back down again without overheating. Just for adrenalines sake in my old 42 years old bones. If this pack last 100 cycles by discharging in this way, and I get 8 good amp hours out of it each ride, I will be more than happy. As you predicted to me the other week when I was asking about how much power to put through the pie, you were right in saying that the pie will just eat up the amps. It certainly does.

My only real issue now is the pulsing, which I hope lies in the LVC cut- off that would have been pre-programmed by E.Lyen for the 84v jumper lead that I originally asked him to configure when buying the controller about 2 or more years ago. As this was probably a circuit board mod, reprogramming with the cable alone won't be enough. I am hoping that the jumper can be removed if nothing else.

Cheers
Andrew

 

[neptronix]

I see. So, you probably have very limited battery space anyway, huh?

What helps most on the kind of terrain like yours is not a light battery, but a mid drive, by the way.
I used to have a bafang 750w on this bad boy and 25ah 48v lipo in the triangle of this:

It handled like a boss offroad.. until i switched from mid drive to hub, lol.

I'd say maybe you might want to switch to the 16ah multistars at the very least. or try some kind of newfangled way of mounting more battery on a bike like that..

or.. and this will be helpful only by about 10%.. Go replace your 'pie with a leafmotor.. same weight, but 90% efficient instead of 80% efficient. This will result in lower battery draw and help out your multistars.

But really, the "1/4th rule" would mean that your cherry point for that 10ah 10C battery is only 25A. You still need more amp hours..

On that note, if you added 5ah of 20C Lipo to that multistar pack... It will mean that you effectively have a 15C 15AH pack, which would give you a 'rated' amp output of 225, whereas 1/4th of that would be 56.25v.

You'd have higher capacity, and the voltage sag would be about almost 1/3rd of what you are seeing. The pack voltage would be very stiff across the discharge at 50A.

..which is awesome, honestly.. when your pack is stiff because of lower IR, you really get better speed and range because when you're *not* sagging it, you're delivering 5%-10% more of your pack's actuall watt hours into your motor instead of turning that energy into heat.

thanks for the reply Nep.
Yep, I have another 10ah pack of the same multistar / same pack configuration, but find carrying that extra load just too much of a ball ache, plus I do not need the extra range at present, but I do understand the extra ah equals lass sag and better performance as well as better range. This is a commencal downhill mtb with the 10ah pack in the frame made for quick blasts up the local trails, along ridges and then back down again without overheating. Just for adrenalines sake in my old 42 years old bones. If this pack last 100 cycles by discharging in this way, and I get 8 good amp hours out of it each ride, I will be more than happy. As you predicted to me the other week when I was asking about how much power to put through the pie, you were right in saying that the pie will just eat up the amps. It certainly does.

My only real issue now is the pulsing, which I hope lies in the LVC cut- off that would have been pre-programmed by E.Lyen for the 84v jumper lead that I originally asked him to configure when buying the controller about 2 or more years ago. As this was probably a circuit board mod, reprogramming with the cable alone won't be enough. I am hoping that the jumper can be removed if nothing else.

Cheers
Andrew

 

[brumbrum]

the extra few amp hours is definitely food for thought. I actually have some small bricks of 20c zippys here but they are 3ah a piece. hmmm., that's got me a thinking for sure.

I am currently saving the pennies for possible mid drive purchase in the near future, it will probably be the new bbs hd 1000 that bafang are bringing out. Though I want to see it tried and tested first by other consumers, and also find out if it has good fets on board, and what extras will be got with the package, i.e a plug n play reed switch etc. I will just have to wait to see what em3v and others start offering when the time comes. I have had the pie for 5 years, its heavier than I would like and more of a motocross ride than a peddler, but it refuses to break, it was what I cut on teeth on in the electric bike world and due to that it has sentimental value.

 

[neptronix]

Hey man, i loved the pie too.. i was hoping that the wider stator version of it would have been produced ( those extra poles really make it blast off from the line ), but it's real dated by now.

The 1500w leaf kicks it's butt, pound for pound.

The 1000w bafang will probably kick ass; maybe not as much as the pie, but for offroad it will be awesome.

We've got an ~88% efficient mxus 3kW motor now that fits in a 135mm drop..

Time to move on for sure

 

[brumbrum]

Yeah, maybe time to move on. I want torque these days more than speed, when oh when will there be a bullet proof geared motor. I fancy a 12t mac too, almost clicked the buy button plenty of times, but would want to run it at 67v , which will probably kill the clutch or keyway quite quickly. I wish i could try before i buy, there's just too much choice.

 

[neptronix]

It's the offroading that would kill the clutch before the power.

On flat road, the motor will hold up perfectly as long as your phase amp to battery amp ratio is much lower than stock ( IE 2.25 to 1, versus the stock 2.75-3 to 1 that i suspect these infineon type controllers are initially programmed to )

 

[brumbrum]

Flat roads? What are they, lol

Thanks for the heads up on the mxus, just read spinningmagnets article about them on electricbike.com. The V2 sounds like well sorted motor.