überschnell

The issue has nothing to do with voltage, and everything to do with amperage. Say, hypothetically, you bought one of those "too small" packs. You're driving along, and the alternator has fully charged the battery. You park, and decide to listen to the radio for a few minutes. Since there is very little capacity to begin with, even the small load of the audio system will drain it down relatively quickly. Now, let's say you've taken ~2/3 out of the battery, or 30% state of charge. Most lithium cells that I've used (can't comment on the competition, they generally have higher internal resistance) will still crank the engine fine at that SOC. So, you've got a drained battery and a running engine. The alternator produces all the current that it is capable of supplying (say, 140A alternator - 70A running loads = 70A into the battery) to get the system back up to the proper voltage. Generally LiFePO4 cells are rated at a 1C charge rate, which on a 4Ah pack is 4A.

70A going in, cells are rated for 4A. You can imagine what happens next.
I agree on a race car, but I would never have the inconvenience of a killswitch on a street car. I would disagree on a managed battery needing a larger case.
A properly sized battery is going to be physically larger, though.

Spin9k

Excellent points both! And how nicely the cells ebb pointed out, the 38120HP, fit into those requirements. With 80A (10C) max charge current, 200A (25C) discharge, the safety level of these unmanaged 'just cells' goes up quite a bit. That's a decent match amp-wise for the car's charging/starting system.

As for the kill switch, it has a dual function, and is esp. usefull on the 'streetable' car. Normally the OEM battery is in place and connected, so the LiFePO4 battery is installed but out of service because the kill switch is open. At track, close the switch, remove the L-A battery and you are good to go. Reverse, repeat, don't even ever have to reset the clock.

HiFlite999

iTrader: (2)

The key word is "firewall" ... which is designed to be a wall against fire. The fuel tank is located under the back seat area, but is blocked from the passenger compartment by a steel wall in addition to the tank itself. (Gasoline in liquid or vapor over liquid form is not explosive btw, though I wouldn't want to press the issue.) Certainly a battery in the nose of the car is at significant risk of damage during a crash, but it's also in a rather "safe" place to sustain that damage. Possibly a trunk-mount is the best compromise in that regard.

I would think that the rather high risk of crashing while racing would inspire one to take all possible measures to reduce the consequences of such crashes. Admittedly though, having had friends burned by exploding gasoline fumes, scarred by sulfuric acid, and partially blinded with high-current arcs has probably made me more sensitive to risk in such things than is normal. YMMV.

TeamRX8

iTrader: (25)

That's why it's mounted in the trunk. And you don't put in the rearward trunk corner. You put it in the forward RH corner at the shock tower/rear seat backing plate. That's not really a crumple zone IMO.

This was my old Odyssey setup before replacing it with a Voltphreaks VPR-S40 in the same location. I can understand why you don't want such long battery leads given the much smaller capacity of your setup.

ebb

iTrader: (7)

The SCCA and NASA in fact allow batteries in the passenger compartment assuming it is properly mounted. In the case of the dry cell type no box is required, must be secure and that is all. Positive terminal must be covered.

TeamRX8

iTrader: (25)

Looks like maybe they passed the dry-cell rule in the last year or so, that was new to me but I don't follow the GCR heavily

überschnell

Where are you finding 10C 38120 cells? I thought you were using Headway cells, which are 2C max (16A). I've never seen a LiFePO4 pack rated higher than 4C.

Spin9k

I was, but since last year when I got those 38120P cells, they've released the 38120HP cells http://stores.headway-headquarters.c...EV,/Detail.bok which is exactly the same size, weight, just with higher charge/discharge rates. Nice, huh ?

überschnell

Very interesting. I'd like to test some of those to see if they'll actually handle the numbers they claim.

TeamRX8

iTrader: (25)

they do look impressive, nice find if they spec out

jspink

Hi there,
not an RX8 owner but very interested in this thread. I'm from Australia.
I have an 1380cc A series classic mini engine I built with a BMW K1200RS motorbike head to make for twin-cam and 16v and double the HP from standard 1275cc A series spec.
I purchased a LiFePO4 "EVO2" 12 cell battery from Ballistic motorsports as well as a Balance charger; http://www.ballisticparts.com/produc...ies/12cell.php
Specifications are 20Ah, 410 CCA, weight 1.1Kg

I figured that since I used to be able to run my not-that-often-used track car on a Lead-Acid motorbike battery, then this should work.
Result so far;
Engine in road going mini - installed new LiFePO battery, 1/2 day of driving around town, Alternator died, almost flattened the battery, grabbed old lead-acid battery, limped home almost flattening that too.
Purchased new Alternator.
Purchased DC clamp meter.
Here are the amp readings I get;
ECU = 4A
EFI Pump = 5A
Elec Water Pump = 5A
Radiator Fan = 8A
Total 22A whilst running at 2000RPM

Add lights - plus 3.5A for clearance, 9A for H4 low beam add an extra 2.5A for high Beam.
Total night time 37A whilst running at 2000rpm
Have not measured brake lamps.

So, pulled out Lead Acid battery, re-installed LiFePO4 and measured at the battery;
26A (as opposed to 22A above)

Pulled out LiFePO4 until I figure how to limit alternator charge under 20A, whilst still allowing battery to start engine.
Either that or cut my losses and buy a VoltPhreak or similar product - just can't seem to justify $1400+ USD plus shipping.

I'm interested to hear how any DIY'ers are going with using small LiFePO4 packs.

Cheers

Spin9k

^ Reading your story I'm confused as to what you are trying to say. I get you went for a ride after putting in a EVO2 battery, then your alternator died.

Do you blame that on the battery? They say the battery is "compatible with your vehicle’s charging system". Why are you trying to limit charging to 20A? Is it because they say don't use a charger exceedeing 20A? Call them to clarify.

Then you measure electrical loads in the car w/the LiFePO4 and then the L-A were installed and get different readings. You don't give the L-A amperage per load, only the LiFePO4...are you sure all loads were all included the second time? Which drew different? That's puzzeling. Some info is missing here.

Having said that, why are you worrying about these loads and then connected them to the battery failing? In a race application, batteries only start cars / act as a voltage damper for the system. Alternators supply load aperage during operation, not batteries. After you start your car, the battery is needed (or yes you could blow an alternator), but generally doesn't drive the load.

If a battery starts the race car, and has sufficient dampening to preserve the alternator, it should be sufficient as long as its kept from destruction by overheating.

überschnell

Agreed 100%. The only thing you really need to be measuring is current going into the battery from the alternator--the electrical loads should be constant, other than being dependent on whether they are switched on or off. Current into the battery is also dependent upon it's State of Charge; a near-empty battery will put a larger demand on the alternator, because the alt has to run all of the car's systems as well as supply additional energy to keep the system voltage up (usually a constant ~14V). After the battery is full (remember, every engine start pulls a little energy out), there should be little to no current entering the battery because the alt doesn't have to supply current beyond the demands of the vehicle systems.


...or by, y'know, the complete and utter lack of cell balancing.

d walker

Thats not actually true. As long as the alternator can supply enough amperage your good, but modern cars draw a lot of amps, especially at high-rpm. We tried using the Odysssey batteries a few years ago in the C-5R because they were lighter and etc. but we ran into issues. When you tally up the amp draw from the fuel pumps, dash electronics, ecu, coils, injectors, fans, ABS, electric power steering, etc. it can easily be more amp load than the alternator can produce under racing conditions. The battery then acts a storage cell and if it cannot supply enough amps to bridge the gap you will have issues.
The specific occurrence we noted was that on a long straight at high rpm the engine would lean out and begin to overheat. We traced this in data to the battery voltage dropping below 11volts by the end of the straight. Changing alternators etc. did not cure the issue so we looked deeper. When the amp draw was totaled it was well over the 115 amp capacity of the alternator. Obviously the extreme heat does not help either. We cured the issue by going to an Optima deep cycle battery that had enough reserve to bridge the gap on the longer straightaways.

Spin9k

It seems mosty battery manufacturers are saying LiFePO4 balance charging is nice and encouraged occassionally, but not essential. I understand the pitch from both technical and marketing perspectives, but I don't have enough data/observation as to how much matters. Goldilocks theory likely will work best.

I did incorporate a balance-charging data cable in my current protype battery that connects to an 80W Chargery unit. I used it before initial install a month ago, but not since. The car charged it w/no problems as it should w/a 14.5V alternator output.

Spin9k

Transient load response (short & long term) is certainly an interesting topic. In the case you mention, it's hard to imagine the Odyssey not meeting transient needs down a straight. Unsure what you used... e.g., typical PC680 @17AH specs are @680A 5sec/595A 10sec/525A 20sec. But of course you had a problem. Puzzling. An obvious question perhaps, but can you say for sure there wasn’t other issues in play? Cable length/GA size/connectors are notorious overlooked culprits w/battery issues. Many use Odyssey PC680s with RX-8s few problems other than short life (2 yrs max) and having to watch parasitic draw.

Not discounting the C5-R problem at all, but I can tell you that my 8AH LiFePO4 battery w/200A max discharge /80A max charge seems to be working fine on short straightaways up to about 90mph in 3rd (at recent event I attended). That's significantly less capacity than even a PC680, so guesses are LiFePO4 has superior discharge/charge transient response over AGM type batteries that helps and/or our alternators are not undersized (like a Corvette?), or the load simply is less, and I need to try longer straights(!). More torture testing on track to come this yr .

I have, but haven't installed a 200A voltage bridge w/my battery to record instantaneous running amp load. That should shed light on actual RX-8 racing use draw. Hopefully 200A battery + RX-8 alternator amp output is sufficient!! (Although leaning out at high revs might be an unintended blessing lol).

jspink

Sorry about all the confusion, yes, I think that the failing alternator was a coincidence.
I emailed Ballistic batteries, and they replied;
"Yes, you are probably expecting too much. Our current batteries are not engineered to accept the higher charge of the car alternator.
Thanks,
Ballistic Response Team"

But I'm not totally convinced that this battery is not up to race use in my application.
I'm going to continue testing...

Well once I worked out not to keep pressing the "hold" button on the new clamp meter, I didn't then record any difference in current draw from the Alternator with the two types of batteries. :bang head: I confused you with battery type, but really, it's not that important as you point out.
I guess I was worrying because the 12 cell Ballistic spec suggests no more than 20A charge. In practice this is not an issue, as I can measure that after a start, the Alternator is supplying 4-5A reducing to 0A after about a minute. Also concerned about how much "dampening effect" a battery needs to provide. My Alternator is only 65A, and full loads at 2000rpm appear as around 40A including lighting.

But what I have done is remove the LiFePO4 battery and check the cell voltages, and attempt to rebalance, as it would appear that B1 seems to hold voltage better (3.6v) than 2,3,& 4 (at 3.3v). I've tried a couple of discharge and then balance charges.

Tomorrow I'm at Robroy hill climb, so good chance to continue the test!
I'm finding this thread a very interesting read.
Thanks
J

Spin9k

^ Strange that Ballistic can't say yes or no to your question, but only "probably asking too much...." They must be just sellers of, not engineers of their batteries. Its all about the battery chemistry, and construction, and they seem not very confident. One would surmise that any car alternator should charge an LiFePO4 12V ok, as long as the max battery charge current/V is not exceeded. Given the stated max charge amps is pretty low, that may (but not likely in your case given your testing) be the rub in your case along with the more worrisome unbalancable cell banks (your 12 cell seems to consists of four series wired banks of three cell wired in parallel).

I understand there are a lot of Chinese battery suppliers whose product isn't terribly consistant, and if vendors assembling batteries don't test and match their cells, there can be balancing/charging problems with the end product. If they simply buy batteries to resell, quality control is out of their control. Because it's extra work, extra cost, and some batteries can't be used, its not always done sadly.

Overall, battery chemistry dictates the max charge voltage allowed and typical alternator output @13.8-14.5v is fine (typically LiFePO4 cell max is @3.65Vx4 cells=14.6v). Of course if the alternator's regulator is not doing a good job regulating voltage, or one cell bank charges at a different rate it can cause problems. Full time BMSs are best, albeit expensive. I would ask them about your balancing problem and what to do.

überschnell

So it's 4S3P of the 26650-type cells? I can't imagine that a 65A alternator is going to tax that at ALL. I wouldn't worry about too much current.

FYI, open circuit voltage after charging is only one facet of cell-matching. They need to discharge in balance as well--basically, overall capacity and impedance need to be close in addition to static voltage.

After a proper balancing charge, however, the voltages should be equal. I would investigate exactly what balancing algorithm your charger uses...

paimon.soror

iTrader: (15)

http://www.af1racing.com/store/Scrip...dproduct=38305

curious on if anyone has seen this/ given it a shot?

The price is VERY nice imo.

dannobre

iTrader: (3)

It appears from the specs that the Max charge rate is 20A,,,not sure how that will work in a system that puts out up to 100A. Might be some way to limit the current to it...not sure?

paimon.soror

iTrader: (15)

Not too sure either, I was only comparing to the shorai packs which are slightly higher on the charge rate:

http://www.shoraipower.com/p-210-lfx36l3-bs12.aspx

Spin9k

That's a topic I've not seen discussed, and I've looked, as these batteries are typically used for electric vehicles (electric scooters, auto conversions, or motorcycles/sportcraft that have small generators only) or race cars w/no alternator and then charged on a balancing charger.

Therefore why I choose max 80A charge LiFePO4 cells.

Of course there's one sure fire (opps) way to find out...

paimon.soror

iTrader: (15)

Yea you guys are right, missed that one on the specs,

Spin how has your pack been holding up? I am trying to decide with the Deka 18L, PC680, or to press my luck with LiFePO4 diy ... i just need something DD-able heh.

Spin9k

My LiFePO4 works great for track, even back and forth to the track. I haven't used it for DD as I don't feel it has enough reserve capacity, but I haven't actually tried it more than 3 or 4 days DD to find out either. Why? $$$ Because if you dischage an unregulated LiFePO4 battery to much, it's done for.

On the other hand, I've used a PC680 DD for years, just keeping it on a Battery Tender to keep it topped up if I leave it unused for more than a week. They last ~2 yrs like clockwork, then, for whatever reason, they don't charge well anymore. They don't appreciate an accidental deep discharge either, but it doesn't kill them.