Testing the big 15 and 20Ah LiFePO4 cells is tough! *Pics*

[cell_man]

Cheers Luke,

I'm wrapping the cells in a plastic sheet to avoid any possibility of the pouches shorting onto anything. I was using sticky tape initially but it's really difficult to to apply it with no air bubbles. The odd air bubble results in the stack being a bit thicker than I had planned. The cells will be under some compression so they shouldn't be going anywhere I reckon.

I'm working on a new idea for the box which could work really well for different pack configurations. I'll get 1 finished before I say too much though :wink:

 

[cell_man]

Hi Chris,

I've got a design engineer friend visiting China at the mo and we made some progress on the box today. I think we've got a workable solution using 2mm aluminium sheet for side panels and 0.3mm between each cell. Most of the first box is made up now and I've got it fairly clear how it's going to be finished off. After that it's just a case of making the parts and then I'll try to get through them ASAP. Hopefully get the first shipment with 2 or 3 packs on the way to you soon

After working through the build a bit today I feel much better about the whole assembly. Maybe I've been worrying a bit too much. It really helps to be able to talk it through with an experienced design engineer, trying to just sit here and work through all the little details has been so difficult. A bit of re assurance and some new ideas is like a breath of fresh air

Great news Paul! It always helps when you have someone who knows what they're on about. I agree that you really need to conduct the heat out of the middle of the pack, and corrugated sheet wouldn't be up to the task. I'd been thinking about the boxes on occasion, but work has had me preoccupied of late. In the meantime, do you have the final dimensions of each pack? I can probably get started on the battery trays next week.

The box design has changed again I'm afraid. I spent all day today and yesterday working on it but there were some things that I wasn't so happy about. The new method is looking very promising though. I've already cut some materials and hope to get something put together tomorrow. It should be fully enclosed in Ally, with width of 175mm, length about 245mm and height 95mm give or take a mil or so.

 

[cell_man]

Hi,

Posted this in the termination thread by mistake.

Be careful of rubbing and vibration!:
http://endless-sphere.com/forums/viewtopic.php?f=14&t=17497&p=256874#p256590

I am pretty sure my Ping died from vibration and shock rather than heat. All bad cell groups were at the ends of the pack, and there were obvious signs of fluid leakage at the bottom of the pack.

Yeah, that's why I put foam liner at the bottom where the battery sits. (Or, if it's in a bag, I don't use anything because the bag material itself has "give" to reduce peak shocks and evenly distribute the forces.)

A layer of cork sheet at about 1/4 inch thick is a good dampening material - which I use in my pack.
Assuming you charge every 20 miles, that's only 250 cycles you had on your Ping. Significantly lower than the claimed 1000+.

I was using that interlocking foam they sell at Costco for playrooms and such. Tried to make it tight enough that the pack wouldn't move around. Even so, the constant jarrring somehow wore down the cells.

Lots of rubbing goes on inside the battery bag or box. Hopefully the shrink wrap used now will help with that vs the duct tape that was on the early packs.

That's why I make the super tight fitting inner boxes custom fit to my pings. There is just no movement at all, and all the rubbing goes on on the outside of the tight box. I have some areas on ping 1 that rubbed quite a bit of the aluminum box away, but remove the battery, and the duct tape still looks cherry.

On some nicad packs I have, it didn't take long to start putting some serious rub marks on the cell cans. I didn't protect them so good, and quickly realized I needed to do more, even with metal canned cells.

I've bought some 1mm PCB type material to put inside the heat shrink for some additional mechanical protection and it shouldn't add much to the weight. Also have some plastic and thicker PCB material which could be used. Can always just send the pack out without the heatshrink and the user could fit whatever they wanted underneath.

 

[Doctorbass]

This PCB idea is great. I'm always using 60 mil pcb thick ( that have been etched first to remve all copper) to protect cells.

For the Konion cylindrical cells, i'm using 6 mil sheet of that pcb between each parallel group.

It's excellent for mecanical vibration protection, flameproof, easy to work with cisor.

I generally use the 60 mil pcb to surround the entire pack and then i fix it with duck tape.. and nylon tape.. and then i put heat shrink.

I had some pack that was droped to the ground accidently and never hav any significant damage.

That idea come from the KILLACYCLE pack.. the entire pack cell sandwich is separated with 60 mil pcb sheet beatwen each layers.

I'm also trying to help Heat transfer between cells in the pack to uniformize the RI. I discovered that often, heat exchange is almost done thru the cell tab (for the cylindrical cells)

Guys from the MIT that are now developing battery for their EV project are using these coroplast sheet to spread air between cells pouches. I also saw someone on the diyelectriccar forum doing something similar.

I remember i heard about someone doing a project that use liquid cooling in clos loop to recirculate and uniformize heat between each cell or battery block along the car. ( the battery are placed at different locations in the car so their temperature must remain the same from one to another all the time. That liquid idea solved the problem.

Doc

 

[maxvdh]

I posted this image of corrugated alu earlier in the thread, but that stuff might be hard to get.

Polycarbonate (lexan) fluted panels would be a way to get cooling between the slabs; it is readily available and less expensive than ally.

Avoid fluted polypropylene (coroplast), it will melt. IIRC, someone mentioned 'honeycomb' panels, that would insulate (bad).

Polycarbonate would be a poor choice, it has nearly identical thermal conductivity to polypropylene. Aluminum would be a good choice but I'm not sure why you'd want to use corrugated material unless you were running a coolant through it. Paul's plan sounds like a good one. I plan to cool my batteries similarly with thin Al sheets between them (possibly just Al foil or flashing) and clamping that to a heat exchanger with coolant running through it. Probably overkill but I want long term reliability.

If you check out my thermal images, even under the worst of situations, just the heat-spreading effects of a thin sheet of aluminum or copper roof-flashing (sometimes can be had for $0.25/sheet) would be all that is needed to spread the heat from the terminal area to the rest of the cell. Another nice effect is as the cell warms, it's Ri drops, so it generates less heat the hotter it gets.

Yeah I thought your thermal images were very cool - thanks for your efforts there. I am thinking I will at least make provision for the coolant but if the temps stay under control I won't actually run it.

 

[peterperkins]

My specs for a set of 50 x 20ah cells are 100A max discharge and 50A max charge rate. The mean current will be much less around 40-50A discharge and 25A charge. I will be using a simple big stack of cells with a very thin plastic seperator between them. No internal cooling provision. Anyone see an issue with heating at these power levels? Of course i will have a some temp sensors in the pack so i can see what's going on in the middle!

 

[jonescg]

The box design has changed again I'm afraid. I spent all day today and yesterday working on it but there were some things that I wasn't so happy about. The new method is looking very promising though. I've already cut some materials and hope to get something put together tomorrow. It should be fully enclosed in Ally, with width of 175mm, length about 245mm and height 95mm give or take a mil or so.

Oh well, whatever works out best and easiest. Lots of pictures OK?

Great news on the dimensions - I can get cracking now

 

[liveforphysics]

My specs for a set of 50 x 20ah cells are 100A max discharge and 50A max charge rate. The mean current will be much less around 40-50A discharge and 25A charge. I will be using a simple big stack of cells with a very thin plastic seperator between them. No internal cooling provision. Anyone see an issue with heating at these power levels? Of course i will have a some temp sensors in the pack so i can see what's going on in the middle!

You won't have to worry about thing. 100A discharge is like taking a nap for these cells.

 

[Mudder]

I didn't read all the posts thus far, so this might be answered already, but here's an explanation for why the FETs fail at lower voltages:

They aren't turning on all the way because we can't get into the saturation region and are therefore the resistance through the FET is quite high. As we apply current through the FET, the Gate to Source voltage will oscillate, which causes a channel to form across the junction, but the channel cyclically collapses at the natural frequency of the oscillator network created by the FET's internal capacitance, etc. Until we can fully bias the gate voltage to open up the channel, our FET is going to absorb a lot of energy at the channel, which will cause the oxide between the channel and the gate to break down. Once this happens, the source is able to leak into the gate, which kills the channel.

 

[liveforphysics]

I didn't read all the posts thus far, so this might be answered already, but here's an explanation for why the FETs fail at lower voltages:

They aren't turning on all the way because we can't get into the saturation region and are therefore the resistance through the FET is quite high. As we apply current through the FET, the Gate to Source voltage will oscillate, which causes a channel to form across the junction, but the channel cyclically collapses at the natural frequency of the oscillator network created by the FET's internal capacitance, etc. Until we can fully bias the gate voltage to open up the channel, our FET is going to absorb a lot of energy at the channel, which will cause the oxide between the channel and the gate to break down. Once this happens, the source is able to leak into the gate, which kills the channel.

Great explaination. The Ri on the FETs was measured at about 5x as high at 2V as at 10v, so I can definately saw the low voltage causing problems in the FETs. The big helper was re-wireing the FET drive setup. I didn't realize it at the time, but my FET driver was using a ground point on a big buss-bar, and this buss-bar was actually having a higher inductive kick-back voltage across it everytime FETs cycled than the V-supply to the FET driver chip, which was causing all types of crazy problems for the poor FETs. Getting this setup correctly enabled the FET group to function and do it's job for me, though I did eventually end up melting the FETs clean off the buss in an extended cell tortue test later on. lol

Best Wishes,
-Luke

 

[olaf-lampe]

You are talking about raising the drain-source voltage from 2 to 10 Volt, Luke? Or did you raise the gate-source voltage only?

-Olaf

 

[jonescg]

I didn't read all the posts thus far, so this might be answered already, but here's an explanation for why the FETs fail at lower voltages:

They aren't turning on all the way because we can't get into the saturation region and are therefore the resistance through the FET is quite high. As we apply current through the FET, the Gate to Source voltage will oscillate, which causes a channel to form across the junction, but the channel cyclically collapses at the natural frequency of the oscillator network created by the FET's internal capacitance, etc. Until we can fully bias the gate voltage to open up the channel, our FET is going to absorb a lot of energy at the channel, which will cause the oxide between the channel and the gate to break down. Once this happens, the source is able to leak into the gate, which kills the channel.

Pfft obviously! I was going to say that but you beat me to it




... I dont know shit

 

[liveforphysics]

You are talking about raising the drain-source voltage from 2 to 10 Volt, Luke? Or did you raise the gate-source voltage only?

-Olaf

My gate-source voltages were always 12v.

I just wanted to see the voltage-drop difference across my bank at a higher voltage, so I setup 10v at 200amps, measured the D-S voltage drop, then measured 2v at 200amps, and the D-S voltage drop was ~5x higher. This helped me to understand why the FET bank I built that should be able to handle >2,000amps was smoking at 500amps.

 

[dozentrio]

Please forgive me for wanting to harvest the bottom-line, without reading my ass off. How do these cells compare to hobbycity lipos in price, discharge capacity, and energy density? And maybe cycles/hardiness... And are they even the same chemistry?

 

[liveforphysics]

Please forgive me for wanting to harvest the bottom-line, without reading my ass off. How do these cells compare to hobbycity lipos in price, discharge capacity, and energy density? And maybe cycles/hardiness... And are they even the same chemistry?

$2.50/Ah $0.78 per watt/hour A123 claims >8,000 cycles to 80% capacity.
This makes it the absolute cheapest battery per cycle/use of anything on the market.

15Ah and 20Ah pouches available.

~15% lower energy density than modern HobbyCity LiPo.

Similar or better power density than good HobbyCity LiPo.

This is the best performance in the world for a safety chemistry. Huge improvement over the A123 round cells. Pouches come in sizes that don't require fitting together groups of 6-10 of them to make useful cell group capacities.


Unfortunately, I think cell_man is currently sold out if I heard correctly. :'(

 

[dozentrio]

Thanks! That saves me a whole lot of time.

Unfortunately, I already paid for some Turnigy's. It would have been worth it to save for a little longer and build a 15Ah pack. Oh well, perhaps in the future. That's the great thing about putting off buying batteries. It seems like you will never regret waiting a little while longer.... hehe

 

[cell_man]

I'm running pretty low on 20Ah cells but there's still plenty of 15Ah cells remaining so get them while you can as I can't seem to get any more.

 

[olaf-lampe]

I'm running pretty low on 20Ah cells but there's still plenty of 15Ah cells remaining so get them while you can as I can't seem to get any more.

That's the baddestestest news for the ES forum. After all this testing and pack-building hassle
Hope it all turns good for you!
Olaf

 

[cell_man]

I'm running pretty low on 20Ah cells but there's still plenty of 15Ah cells remaining so get them while you can as I can't seem to get any more.

That's the baddestestest news for the ES forum. After all this testing and pack-building hassle
Hope it all turns good for you!
Olaf

It'll be ok There's lots of other good products I'm looking at which I'm sure will be of interest. I'll be glad to see the back of the pack building too and maybe I'll have time to get a website up and running when I'm not building packs for about 50USD a day or probably a lot less.

Unfortunately even though these are very well priced it's still not cheap enough for most who'd rather stick with a li po pack which will save you a few $$s in the short term or a 2C ping pack. Unfortunately these cells don't come cheap and just because I charge more than a Ping, it doesn't mean I'm making more money.

 

[John in CR]

Cellman,
Your efforts, great service and fair pricing will be rewarded in the long run, just as Ping's were. Even those of use who haven't bought cells yet appreciate your efforts.
John

 

[cell_man]

Thanks John, I'm trying and hopefully everything works out for the best. I hope to get a few samples soon and get some good quality high C rate LiFePO4, some more run of the mill 2/3C stuff and maybe even some Li Po in the not too distant future. Will probably keep some A123 26650s as well, assembled into packs with BMS etc. I can't beat the price of a cheap power tool pack but it might work for some if they just want a finished solution.

I'm gonna try to steer clear of getting too involved in building stuff in future, it just takes too much time. Unless I can get some local staff it just doesn't make sense. Gonna try to get a good range of ebike motors, controllers etc as well and most important of all, get a proper website up and running to sell it all Gonna be busy...

 

[miro13car]

CELL-MAN,
THANK YOU FOR ALL THIS THREARDS, i LEARN SO MUCH.
sHOW ME ANOTHER supplier so technically knowledge, so forthcoming.

I agree with above poster - all your affort will be rewarded if you keep supplying these cells.
You already have quite hight ratings as supplier.
MC

 

[cell_man]

CELL-MAN,
THANK YOU FOR ALL THIS THREARDS, i LEARN SO MUCH.
sHOW ME ANOTHER supplier so technically knowledge, so forthcoming.

I agree with above poster - all your affort will be rewarded if you keep supplying these cells.
You already have quite hight ratings as supplier.
MC

Thanks MC, but I wouldn't consider myself particularly knowledgeable in this whole area. There's lots of folks on ES that know much much more than me, I'm just a newby However, I have worked in Engineering for quite a few years and did a B.Eng at Uni so hopefully I can pick things up pretty quickly. At least I'm in the right place to learn, here on ES

 

[swbluto]

Thanks John, I'm trying and hopefully everything works out for the best. I hope to get a few samples soon and get some good quality high C rate LiFePO4, some more run of the mill 2/3C stuff and maybe even some Li Po in the not too distant future. Will probably keep some A123 26650s as well, assembled into packs with BMS etc. I can't beat the price of a cheap power tool pack but it might work for some if they just want a finished solution.

Don't worry about it. Customers are diverse and they all prefer different things. Most prefer less price, but many prefer less work and many prefer the best of the best that's going to last a long time in a relatively safe chemistry. Just measure the demand after making your offerings and see if the predicted business volume justifies it.

BTW, you're somewhere in Asia, right? Don't you have the gift of cheap labor readily available to do the drudgery?

 

[cell_man]

Yeah, I'm in Shanghai, China. Of course there is cheaper labour available here but once you have staff it changes everything. They need to be trained, paid every month, kept busy and do a good job. I'm pretty much where I could keep a guy busy full time but I just need to take that jump and find someone that is at least reasonably capable.