Prius Batteries - How to make them work

powermed

10 W
Joined
Jun 22, 2008
Messages
80
I'm posting this separately since it might be too far buried in other threads - mostly about Prius failures. There's been quite a bit of mystery and misunderstanding about these amazing cells. It's no coincidence that Toyota has been using them successfully for almost a decade, and decided to stick with them in the new high end Lexus sports car. My information is gleaned from the original published paper (you have to buy it and it's copyrighted of course) and now approximately 3 months and 200 cycles. I commute 10 miles each day, 5 miles round trip. My standard pack is 44.4v (6 modules) on either a 600w Wilderness brushless hub motor or a 1000w mid-engine (through the gears) brushless currie. They should be strongly considered for commutes of 10 miles each way or less.

Each module is 11 x 4.75 x .75 and contains 6 cells in series for a total nominal voltage of 7.2v and 6.5 amps.
These are not typical NiMH chemistry in any way. They can deliver 10C + discharge rate, have very low internal resistance, deliver rated output to over 80% discharge (performance starts to fade), can be safely charged in about an hour, and have a proven life span of thousands of cycles and at least 10 years. They do not go out of balance when used this way. The performance is quite comparable to A123 cells. The pack shown is 6 modules in series with chargers - weight is around 14#.

1. Do not buy anything except recent Gen III cells that have not been sitting around. The cells must have a resting voltage of at least 7.2 when you buy them. Use matched cells - you should probably consider buying an entire battery and taking it apart rather than scooping up individual cells.

2. Do not even bother trying to charge them with a standard commercial NiMH charger. I have found no NiMH charger that can charge them safely - they all miss the delta V.

3. Charge them to 8.3v CV / module (CC if at all possible, but I've found no ill effects from regulated chargers that can deliver up to a rated current of 4.5a). I have not tested higher amps, but the cells by specs should handle 1C charge rates. Measure the charge time to get to 8.3v for your use. I tend to add about 15 minutes after 8.3v - the charger output tends to fall to 1a or less.

Other checks worth doing to make sure your chargers are working properly are listening for subtle crackling sounds at the vent ports (gassing indicating end of charge) and infrared temp checks down the thermal wells looking for a fast delta T.

Notebook chargers that are rated at 16v output will often have 16.65 - 16.68v outputs. If you can find matched ones go for it and charge them in strings of 2 cells - make sure the charger voltages are within .02v of each other. There is a degree of self balancing. Chargers start up warm at initial charge, then cool towards the end of charge.

CRITICAL USE INFORMATION
Treat them like A123's - DO NOT CONTINUE TO USE THEM ONCE PERFORMANCE DROP IS NOTED - recharge before any continued drain on the cells. I have had no trouble doing a full ride, charging, and using them again. Also common has been riding for 30 minutes, charging for a short time, and using them again. I've noticed no heat nor swelling of the packs, I do not believe compression is required for healthy cells that are well treated.

I've made some of these packs for family and local colleagues that are commuting, and have a few extra packs I made from the last Prius battery I bought (I'm selling them on ebay - do a search for 44.4v*). I can make about 3 more if folks are interested: $255 for 29.6v pack, $375 for a 44.4v pack, and $480 for a 59.2v pack (all wired up includes shipping and 4.5a chargers).

I assume the technical types here might want to buy a full prius battery, carve it up etc rather than pay for ready made packs. I'm personally not into the group buy thing but someone here might have the inclination.
 

Attachments

  • Prius_pack.JPG
    221.6 KB · Views: 18,784
Thanks Bill. I stickified your thread.

Any long term test results would be appreciated.

Do you have any capacity measurements? I recall they were rated 6.5Ah or so.
 
honored for the stickyfication...

6.5AH is the tech spec, but regrets - no wattsmeter. Having the same ride day in and day out for several years, and owning and using Dewalt and Tenergy's, I can make the following empiric statements;

1. vs Tenergy sub C NiMH (3.8 AH) packs:
Those packs would routinely be done just making it 5 miles one way with light to moderate pedaling. They'd be pretty hot at the end of the ride.

2. vs dual Dewalt A123 "36v" (2.4 AH x 2) packs:
Those packs would never make it round trip, I'd always charge at work. Used them for about a year.

With the Prius packs I can make the 10 mile round trip with light pedaling and still not have experienced noticeable sag. They are not even warm at the end of the ride. The cells are typically around 7.2v when I check them before going on the chargers.

I would discount the AH rating of the Tenergy's, I think most of us know / feel they are generously rated. Given the A123's have a similar discharge profile, perhaps a little closer to the 90% before noticeably sagging, with Prius cells giving it up between 80 - 90%, I think the stated AH in the technical paper is spot on. Using the A123 experience as a benchmark would suggest between 5.2 and maybe 5.75 usable AH in the Prius cells.

Select quotes from Taniguchi et al. / Journal of Power Sources 100 (2001) 117-124 so I don't get sued...
Most of this refers to Gen II cells. Couldn't find anything specific about Gen III which I use.

"Fig. 14 shows the power characteristics of the prismatic
module at different SOC. This battery has a very high power,
1000 W/kg"

"the durability was found to be equivalent to
more than 100,000 km driving and neither changes nor
deterioration in battery characteristics were observed"
 
Powermed, very interesting, thanks for sharing.

I'm thinking the 29V pack could be interesting to use on my 24V Walmart Ezip. Some folks are using 36V on Ezips but require extra techniques for cooling the motor, 29V might give a nice performance boost without this hassle.

But I wonder how the 6.5aH would compare to my current SLA 10aH? My commute is 6.5mi one way, rolling terrain but nothing steep; my SLAs are pretty pooped afterwards.

Also, how much do these weigh? I assume the dimensions you gave are for a 44.4v pack?

Alan
 
Hi Alan, I think it's fair to say a good SLA will deliver up to 50% of it's capacity at good power before fade time. Figure 80 - 90% for Prius NiMH and A123, so I think you'd probably see close to the same range all other things being equal - if those are good SLA's. However given the 20% bump in voltage I suspect you'd see a pretty significant jump in performance all around. Correct on the dimensions. The pack would weigh about 9 pounds.
 
I've had my MY1018 motor (same as Ezip)up to 42 volts NiMh and 36 LiFePo4 without incident in some really hot weather. Those motors can take some punishment without dying! I would ventilate like Tyler and Reid did if you hit a lot of hills in heat at higher voltage. Those 29 volt NiMh sound great! Mine are the Batteryspace ones that I am about to bump up to 48 volts.
otherDoc
 
These prius cells are interesting, seems like a reliable challenge to the low current lifepo4 Chinese kitchen table manufacturing runs that are all over ebay. They only get 40 amps max out of a 20AH pack.

Do these need forced air cooling for high current usage? I know the Prius has a blower as part of the battery module.
My application is a trail bike with Mars motor powering it. I'd like to run near 48 volts ( about 6 in series ?) , and limit the controller to 150 amps. Maybe 4 strings in parallel would be best. To charge I would have to separate the strings and peak charge like that, I assume.

I talked to someone who tried to charge a Prius pack at about 8 amps, and the cells were bulging out on him. Maybe his charger missed the peak, like you mentioned can happen with run of the mill peak chargers?
 
Prius cell talking points, in my opinion (and in no particular order), are charging speed / convenience, deliverable power, form factor, known quality, life span, and durability. The charging method is just "different" than folks are familiar with. I'm really trying to emulate (not duplicate) the 2 principles adhered to by Toyota. In plain talk;

don't drain em' and don't stuff em'

I don't believe they need cooling at a 5C rate. The tech specs suggest 10C is no big deal, I'd think that's conservative since they have very little internal resistance and big bolt connectors, but I haven't tested them continously at that mark.

My standard ride is a 1000w BMC motor on a midengine - through - the - gears running through a PowerPack controller. From previously using a watt meter on a lesser ride I know the motor easily pulls 1000w at 44v* (20+amps - around 3C) without the batteries getting warm (I don't drain them - end of discharge is where you see garden variety cells heating up).

I think you'd be fine in parallel - good luck if you ever hit the peak 10C rate of >200A! They'd be pretty easy to hookup to a bus bar, you'd want a big fat cable to the controller. I do something similar for my electric boat which can hit > 100 amps when I'm showing off. One thing to keep in mind is finding matched cells - you'd want a complete Prius battery. A Gen III prius pack has 28 cells - you'd get your 4 x 6 pack that way (send me the two left over ;)

Charging should be no big deal, at least with the charger setup I use which is 1 charger for every 2 cells (all matched for output to within .01v). Check your AC circuit to make sure you can pull the total wattage though. Natch, without a LV cut off you'd need a comfortable way to do health checks on the cells from time to time.

An 8 amp charge rate (>1C) seems aggressive to me if they were charging a big string. Although they have little internal resistance I'm much more comfortable pushing charging limits when I'm hitting individual cells rather than feeding top amps through a battery string. If your friend was relying a commercial charger he was in trouble as soon as he plugged them in, regardless of the amps IMHO.

By the way, thanks for all the personal emails and interest - since there's quite a bit of duplication in questions I just posted the answers / comments here.
 
That was a mix up. The original dimensions was for all 6 of the Prius modules as shown - I edited the post, thanks. The width of each cell is .75" The form factor is actually perfect for virtually all the nice rear rack bags such as Topeak. I could easily fit an 8 module pack in the medium Topeak bag - with chargers.
 
Thanks,

The application I was thinking about would a 40AH, 50 volt pack, or thereabouts, for an electric scooter. Still a lot of volume compared to LiFePO4. But if you don't need such large capacity this is the way to go. Toyota quality cannot be compared to Chinese Thundersky...

These would quickly gobble up space if one was thinking of a long-range plug-in Prius too.

Than again, bit off-topic, but the only reason for so many cells is because it is the only way around that pesky Chevron/Texaco/Cobasys patent and their refusal to license the sale in the US, or use in an automotive application, any cell larger than a the capacity of a D-cell. Supposedly this restriction expires in 2010, and all NiMH patents expire by 2014. It can't come soon enough. So much for patent law spurring innovation.
 
"The application I was thinking about would a 40AH, 50 volt pack, or thereabouts, for an electric scooter. Still a lot of volume compared to LiFePO4. But if you don't need such large capacity this is the way to go. "

Yeah, I think once you get over the 28 modules of 1 complete Prius battery you've got problems. That size pack would require more than 1 Gen III Prius battery - 28 modules in each. That translates to: 7.2v x 7 modules = 51.4v, x 7 "packs" of modules to get to 45.5AH = 49 modules total. I'd want to have my hands on those complete batteries to make sure they were really closely matched before I paid serious money for them...

The volume is problematic if you're going over the typical commuter thing say 10 miles each way, but I will say the form factor softens the blow an awful lot. The cells fit together perfectly while the ridges on the sides maintain an air gap for cooling between them. There are 2 threaded bolt holes on each end on the bottom of each module so you could get pretty creative with mounting. As they are, the typical battery "pack" of 6 - 8 cells and chargers fits pretty darn good in the slide in Topeak, Crystalyte, or other standard rack bag and the #10 ring bolt terminals and copper bus bars take all the "fun" out of assembling gazillion cell packs...
 
Well, I learned something last night, just not entirely sure what it was...

For the first time since I've been doing this - some several hundred cycles - I left a pack charging over night. I saw the plug first, and with that gut wrenching feeling placed my hand into the bag to check the pack - no swelling. I checked the cell voltage, each cell 8.3v, slightly warm. Chargers were cool.

I'll cycle the pack today, but I expect normal performance. I always thought there was a tapering of the current since the chargers cool towards the end of cycle, so I think this confirms the chargers deliver a constant voltage, with up to 4.5A delivered but tapering as the voltage is maintained by the packs. I think this probably approximates the part of the charge profile designed for 10 year Toyota / EV use where the packs are only charged to around 80% capacity. It also confirms the need to not exceed 8.3v or so, otherwise the chargers would have continued to pump in amps until cell destruction.

While I'm not going to go passive on the time, it does reinforce my feeling that recent Gen III cells are kick ass and far more resilient that older cells - whether it's age or not.
 
That is very interesting and I am glade your didn't damage a pack. I still cant bring my self to throw away my cells so I still have a 5cell 36V pack made up that I have been playing with.I tyred charging using my rc charger on lipo settings where you could select a charging voltage of 8.4v couldn't go any lower but it was much lower than I had before. The charge cycle Tapered off as time went on I was able to get about 2amp hrs into the cells or about 45min at 2.5amps before I got a very faint gassing sound and stopped the charge perhaps if the voltage was lower they would have excepted charge for longer. My wife used them on a 5km ride before they went flat.

If I can find a good place that sells cheep second-hand IBM chargers I will buy one and see how it works with that before I give up on my old packs.
 
totally normal ride on my overnight pack. The details are beyond my current knowledge base, but the big picture is pretty clear - exceed the right charging voltage (somewhere around 8.3v) on those packs and you're in for an unpleasant event of some kind.

the fact that your results on 8.4v at a short time in encouraging as far as gen II (or maybe gen I) packs goes Kurt. I'm now thinking if you can get down to around 8.3v your going to have a much different feeling for the cells you own.
 
I can't help but think that these Third Generation Prius cells are seriously undervalued, compared to the discussions on LiFePO. They represent a mature technology, with reliable charateristics, and beautifully executed construction.

The availibility from used battery packs, makes them affordable. They are easy to assemble into packs, safe and abuse tolarant. The high C rate makes them really useful.

:idea: In larger packs for small EV's a combiled open source BMS and LVC would be of real benifit, allowing a lot higher portion of capacity to be safely used. Something similar in execution to the 24 cell LiFePO BMS but extended to 200V or 300V range, might be a significant kickstart a better choice for small EV conversions, than say Thundersky cells :?:
 
Agreed.

I've probably had a few dozen emails from folks wanting to know how to charge these cells in larger configurations than the stock Prius. I'm obviously very happy and comfortable with them compared to A123 cells. I've taken to charging them in series - they just don't seem to get out of balance.

One thing I'm starting to appreciate I didn't before - they are not self discharging like typical NiMH cells. I've a single cell that hasn't dropped at all in over 6 months.
 
One problem is you cant connect have them in parallell when charging, right?
Because when one cell gets fully charged it's voltage will drop a bit, which will cause a large current to flow to it from the other batteries in the group.
 
If I understood you right you are charging and discharging them all in series, so the voltage drop will not be a problem for you.

I had a plan to use the whole pack (which contains 28 batteries in series), but use them in a 4p7s or 3p9s configuration, until I realized they can't be charged in parallel.
 
I am doing that for my packs - but my point was that each module (a single plastic unit) in the Prius pack is actually 6 cells in series - the module voltage is 7.2 nominal - so a Prius pack is essentially 28S6P. I'm not convinced you can't safely charge them in parallel - as long as you're not trying to get a full charge into them.
 
You may be right about the charging. Are you charging constant current and cutting at a voltage?

Since every module is 6 cells in series the total pack is 6*28s = 168s.
 
bearing said:
I had a plan to use the whole pack (which contains 28 batteries in series), but use them in a 4p7s or 3p9s configuration, until I realized they can't be charged in parallel.

You can use them in parallel with the right charging setup. They should behave fine in parallel during discharge. It should be possible to make some kind of switching scheme to disconnect the strings for charging and reconnect them for discharge. This switch could be mechanical or electronic. An electronic switch using FETs for each string could be made automatic.

During charging, you could use a separate charger for each string or switch a single charger from one string to the next and charge sequentially. You could also use a charge controller for each string that operates off a single large supply. Separate chargers would be the easiest.

I think this kind of approach will be more practical than trying to deal with the high voltage of one giant series string.
 
Lapwing said:
I can't help but think that these Third Generation Prius cells are seriously undervalued, compared to the discussions on LiFePO. They represent a mature technology, with reliable charateristics, and beautifully executed construction.

The availibility from used battery packs, makes them affordable. They are easy to assemble into packs, safe and abuse tolarant. The high C rate makes them really useful.

:idea: In larger packs for small EV's a combiled open source BMS and LVC would be of real benifit, allowing a lot higher portion of capacity to be safely used. Something similar in execution to the 24 cell LiFePO BMS but extended to 200V or 300V range, might be a significant kickstart a better choice for small EV conversions, than say Thundersky cells :?:
I agree, I'm old NiMH addict as well. My last packs lasted close to 5 years and tons of abuse and well over 1,000 discharge/charge cycles to them. They still work even now, just at about a 5AH capacity. Not only that, the technology has matured a little more. When I bought my packs, they were 12AH, but the new ones are 13AH at only slightly more than I spend for the original ($250 vs. $289), plus I got more voltage. I started with (2) 24 volt @ 12AH packs back in 2004 and now I've moved up to (3) 25.2 volt @ 13AH packs. The new ones currently do near exactly 13AH in a free wheel spin test (just letting the wheel spin at a low wattage, takes a day to drain them) and under a big load (actual riding), still hit 12AH, with less heat to boot.

I remember my old ones, full power, hot enough to cook eggs on those things after a discharge, but the new ones just feel "warm to touch" now rather frying pan hot. After the abuse my last ones went through, I'm going to baby these new ones. No more leaving them outside (in the shed) to freeze/roast in the weather while charging. No more loaning them to friends (who run them dead and bring them back to me), I loan them my old ones next time, 5AH should be good enough for them. :D

Don't get me wrong, I'm glad to see the Lithium tech take off and I'm certain that's where the future lies, but don't count out the old faithful just yet, the NiMH still have a good place in the market, just as you said, tried and true technology now, no surprises about how it will perform or how long will it last.
 
Back
Top