Experiment with li- poly batteries in a greenpower racecar

Chat about all things battery in here.
Need to know what type to use or size or capacity then again place your thoughts here
bobc
Posts: 107
Joined: Thu Dec 20, 2007 1:11 pm
Location: Knutsford

Experiment with li- poly batteries in a greenpower racecar

Postby bobc » Tue Aug 17, 2010 8:04 pm

Hi Gents, been a while since I posted. We tried some 20C 2.2Ah model aircraft batteries in our greenpower racecar in place of the lead acid standard batteries. I was particularly interested in the performance at high discharge rates for my current project. Anyway here is a photo of the "installation" (OK they were gaffer taped to the standard battery clips, you can barely see them in the photo...)
Image
The batteries were emptied in 5 minutes - see the data log below
Image
notes
1) at the end the volts dropped incredibly fast, I thought I might have damaged them but they've charged up OK
2) at the end the batteries were hot (~60C)
3) one battery was slightly but noticably swollen until it cooled
4) series resistance goes down as they flatten (unlike lead acid)
5) I got 2.26 Ah out of them at the 20C rate (i.e. no peukert)
In summary the Lipoly's seem to be ideal for my needs. I'm concerned about safety; but I understand the fireballs are caused by 1) overcharging 2) physical damage or 3) short circuit - anyone care to comment?

User avatar
Jeremy
Posts: 472
Joined: Thu Jul 31, 2008 6:35 pm
Location: Salisbury

Re: Experiment with li- poly batteries in a greenpower racecar

Postby Jeremy » Wed Aug 18, 2010 7:05 am

Neat experiment, Bob.

Have you looked at the vast wealth of knowledge and experience from people using these batteries at very high rates over on the Endless Sphere?

I can't recall a LiPo fire from anyone on there, and some of them are running small packs delivering peak power well into the several kW range. There has been a noticeable shift away from LiFePO4 to LiPo RC packs over there in the past few months, as the "early adopters" have shown that running these packs (with a good BMS and charge regime) is relatively safe.

Certainly the fact that LiPo prices have reduced has helped. My next project will almost certainly use a fairly big LiPo pack, around 12S and maybe 6 to 8P, most probably using the fairly cheap Hobby King supplied Turnigy 5Ah 6S packs (http://hobbyking.com/hobbyking/store/uh ... _6S1P_20C_ ).

There are some promising looking BMS designs coming out of collaborative efforts on the ES forum, too, including a pretty neat design that uses the very cheap $15 Cellog unit from Hobby King as the basis for the management system (this unit: http://hobbyking.com/hobbyking/store/uh ... _2-8S_Lipo ).

Jeremy

andylaurence
Posts: 108
Joined: Wed Apr 25, 2007 6:45 pm
Location: Bristol
Contact:

Re: Experiment with li- poly batteries in a greenpower racecar

Postby andylaurence » Wed Aug 18, 2010 8:29 am

I've been looking at the PolyQuest XF LiPos, but it seems the new PolyQuest NanoTech have similar power density for half the money. About £130 buys 20Kw of peak power (12S1P) with the 5Ah Nanotech cells from just 2Kg. I could build a 100hp pack for about £500 and under 7kg, which makes the power source for a sprint car rather viable. Now all I need is a motor/controller that will happily take 100hp peak for <60 seconds and ~50% duty cycle whilst being similarly light and I could add some battery assistance to my sprint car. It seems to be a cheaper solution than a turbo kit and potentially lower running costs (no costly engine rebuilds).
Image
Above figures include track days and the odd competition.

bobc
Posts: 107
Joined: Thu Dec 20, 2007 1:11 pm
Location: Knutsford

Re: Experiment with li- poly batteries in a greenpower racecar

Postby bobc » Wed Aug 18, 2010 12:29 pm

I've not got onto endless sphere, maybe I should take a browse.... I read the "horror stories" on an RC forum, some guy had compiled all the "my lipos just blew up" he could find, and there did seem to be quite a few. Typically there would be an unstoppable fireball 20 minutes after the last interaction, frequently because a wrong charger option was chosen. Next came damage, guys crashing their plane, flinging the remains on the back seat of the car & getting the fireball while on the motorway home.... Or "just nicking" the wrapper of the cells. Or (I love this one) "the dog bit it" - 3 or 4 instances of this! But I now understand that a bit of swelling is not uncommon & some folk "wrap them tightly" to control it. I guess if I look after them OK I should be safe enough - the application will only take power for 40s on a stretch so I'm not expecting thermal problems. Above I was effectively taking a kilowatt from the 2.2Ah 6S cells. Even at that rate I got 2.26Ah out! Impressive.
These batteries were a hobbyking "special offer", the real project will need a LOT more.
For charging I'm just going to have a zillion single cell chargers based on the TI Bq24600. These will all be isolated from each other by coming off separate windings on the mains transformer. I got a 100W transformer kit that I will put my own secondaries on, each transformer will do 6 cells. No BMS, no development, KISS - that's the theory anyway ;^) The first PCB is at the maker's now.
The "electric power boost" system sounds interesting;- alternative to turbocharging etc. Have you been able to get hold of any of the F1 KERS technology? Should be reasonably do-able from an electrical viewpoint.

bobc
Posts: 107
Joined: Thu Dec 20, 2007 1:11 pm
Location: Knutsford

Re: Experiment with li- poly batteries in a greenpower racecar

Postby bobc » Wed Aug 18, 2010 12:39 pm

Jeremy - once again, great spot with that hobbyking battery monitor thing. I was going to build in a stack of LCD DPMs - one for each battery, but this thing is better, cheaper & ready made to do exactly what I want - excellent!
PS I should clearly spend much more time browsing the hobbyking catalog.... ;^)

User avatar
Jeremy
Posts: 472
Joined: Thu Jul 31, 2008 6:35 pm
Location: Salisbury

Re: Experiment with li- poly batteries in a greenpower racecar

Postby Jeremy » Wed Aug 18, 2010 12:47 pm

The cells that seem to get consistently good reports from high-power users on the ES forum are the Zippy and Turnigy cells - they seem to be the best price/performance cells around at the moment - although I've not heard any reports on the Turnigy Nanotech yet. I think the main problem with drawing very high current from small packs is voltage sag, this can be a massive waste of power once you get up around the 20C+ mark. Personally, I take all the manufacturers supposed C ratings with a pinch of salt, as the only rating that really matters is the consistency of total effective Ri (including cell terminations) with load. The gauge of wire and nature of the cell tabs means that there will be significant I²R losses there at high current.

If it were me, then I think that I would very much err on the side of having more capacity, just to lower the total effective battery pack internal resistance. There would be a weight penalty, but I'm sure that the extra power available would easily compensate for it.

One big advantage of your approach is that ultimate hp isn't the primary factor with a sprint car, as you know. Getting the highest possible torque to the wheels, commensurate with maintaining traction, is what counts, as it's almost always acceleration, rather than top speed, that determines the winner. My old sprint/hillclimb Clan Crusader actually got slower when I changed the ~100hp engine for a new Hartwell 120hp+ engine. The reason was the loss of bottom end torque, which seriously reduced performance at some of the short and windy sprint circuits. It was faster down the straight at places like St Eval, but I lost nearly a second on the hill at Tregrehan, despite having invested in a costly Jack Knight, close ratio transaxle and lower final drive.

I'm guessing that just 50hp of electric power might give you performance equivalent to maybe 100hp of ICE power on a circuit or hill where you can't normally get out of second or third gear. Coupled with the reduced gear change time losses you may find that this would be a winning formula, but how would the club/RAC MSA class it?

Jeremy

User avatar
Jeremy
Posts: 472
Joined: Thu Jul 31, 2008 6:35 pm
Location: Salisbury

Re: Experiment with li- poly batteries in a greenpower racecar

Postby Jeremy » Wed Aug 18, 2010 12:56 pm

Bob

My charger is cheap and simple, but is for the lower voltage requirements of LiFePO4. I use some cheap (search eBay) DC DC converters. They have isolated inputs and outputs, so can be connected to a common power supply and still hook up via a multipole connector to the battery pack.

The ones I'm using at the moment are rated at 50 watts, 3.3V. They have a voltage trim pin that when shorted to the negative output increases the terminal voltage to 3.63V, ideal for a fully charged LiFePO4 cell. Mine deliver around 15 amps into a completely flat cell, tapering as the cell voltage rises. I run an array of them from a Chinese 48V switched mode power supply, again bought via eBay.

I think I paid $5.99 each for the 3.3V DC DC converters, pretty cheap for what they are. There are usually heaps of them for sale on eBay, often at very low cost.

If you want to charge LiPo, then you need to explore other options, although I know one chap who managed to tweak some 5V DC DC converters down to 4.2V by changing a resistor or two on each.

Jeremy

andylaurence
Posts: 108
Joined: Wed Apr 25, 2007 6:45 pm
Location: Bristol
Contact:

Re: Experiment with li- poly batteries in a greenpower racecar

Postby andylaurence » Wed Aug 18, 2010 4:50 pm

Jeremy wrote:One big advantage of your approach is that ultimate hp isn't the primary factor with a sprint car, as you know. Getting the highest possible torque to the wheels, commensurate with maintaining traction, is what counts, as it's almost always acceleration, rather than top speed, that determines the winner. My old sprint/hillclimb Clan Crusader actually got slower when I changed the ~100hp engine for a new Hartwell 120hp+ engine. The reason was the loss of bottom end torque, which seriously reduced performance at some of the short and windy sprint circuits. It was faster down the straight at places like St Eval, but I lost nearly a second on the hill at Tregrehan, despite having invested in a costly Jack Knight, close ratio transaxle and lower final drive.

I'm guessing that just 50hp of electric power might give you performance equivalent to maybe 100hp of ICE power on a circuit or hill where you can't normally get out of second or third gear. Coupled with the reduced gear change time losses you may find that this would be a winning formula, but how would the club/RAC MSA class it?

Jeremy

Indeed it is, Jeremy - it's all about the torque at the wheels. Running wide and sticky slicks helps out a lot and at ~475kg, there's not much weight to move. Gear changes are already quick, as it's a sequential dog box but the extra power would be a great bonus. Ideally, I'd like a simple DC motor that I could hook up to an on/off switch, actuated at full throttle. That'd mean I only use the extra power when needed. Regen is not required for the distances I compete over. However, DC motors of that power that are not excessively heavy seem hard to come by, so I guess that means I'm more likely to use a bunch of HXT motors or the Colossus they're building on Endless Sphere, if they can ever find a controller that'll power it!

I have a good amount of space between the front wheels, so there should be room for two driveshafts, each mated to a large spur gear. HXT motors could be mated to the spur gear in a cluster, providing drive. Four on each front wheel would give ~50bhp for around 15kg; 6 would give ~75bhp/~22kg, including controllers and cells. That sounds like a great proposition to me.

As far as MSA regs, there's nothing in the Blue Book about it and when I informally asked about electric power when first pondering the electric Mini, I was told they'd make the rules up when there was a car that called for it. One thing's for certain, I won't be laying more power than a 450bhp turbocharged GSXR1000 could and that still fits in class D1 (<1800cc Sports Libre).
Image

Above figures include track days and the odd competition.

bobc
Posts: 107
Joined: Thu Dec 20, 2007 1:11 pm
Location: Knutsford

Re: Experiment with li- poly batteries in a greenpower racecar

Postby bobc » Wed Aug 18, 2010 8:14 pm

interesting, you've gone the individual cell charger route too. There's a couple of downsides with the fixed voltage charger; allegedly you charge to 4.2 but should then pull the power - the dedicated li charger chips have this state machine built in. Plus they flash indicator LEDs at you to tell you what's happening (any electronics project should have at least several LEDs on it ;^). I should get the first PCB back in a couple of weeks, I'll see what the circuit costs. At the moment the transformer's the big cost but if I'm getting a batch I should be able to negotiate a substantial discount. I might be able to get a stencil made & put the boards down the line at work - that would take all the effort out of the job, and guarantee a quality level I couldn't match at home....

User avatar
Jeremy
Posts: 472
Joined: Thu Jul 31, 2008 6:35 pm
Location: Salisbury

Re: Experiment with li- poly batteries in a greenpower racecar

Postby Jeremy » Wed Aug 18, 2010 9:22 pm

Pulling the power is easy. The Cellogs have an alarm output that will drive a relay, solid state or conventional. You can programme a high voltage limit for either each cell or the whole pack (which is stored in non-volatile memory in the Cellog) and then just use this to detect full charge and turn the charger off.

If you use the more expensive version of the Cellog, the $30 one with the built in datalogger and USB port, then you can download the charge history and keep records of each cell's voltage over time, allowing early intervention if one cell starts to show signs of reaching full charge consistently early, indicating a potential problem.

Jeremy


Return to “All things battery related”

Who is online

Users browsing this forum: No registered users and 23 guests