youtube video of replacement of nimh with lifepo4 hv batteries

@2k1Toaster is our resident battery expert.

 

no one has accomplished a lithium replacement to my knowledge. dr. prius has been working on one for quite a while, with no success.

 

Using the Toyota BMS that is geared for NiMH with Lithium of any kind is asking for a fire to burn your car down. It will work just fine 99% of the time because of the 40% to 80% limit if you can keep your lithium range also in the low end range. But there are corner cases that the NiMH system has no problem with that a lithium system will overvoltage and make fire, or under-voltage and kill.

 

This guy has been running one for about a decade.
Antiscab's NWH10 Prius plugin Lithium conversion - AEVA Forums

There are now dozens of Honda conversions to lithium from NiMh.

Lithium conversions are still a DIY science project.

 

Check out what the Honda guys are doing. There are some bright people over there working together to come up with a replacement for OEM honda packs.

LTO cells as NiMH repacement | Honda Insight Forum

 

I have wanted to try a lithium conversion for years. It's the pesky 1% that keeps getting in the way.

 

it's a fools errand

 

Yeah I've been contacted by some of them too. And they have ignored my warnings. I won't even sell my NiMH to fit into a Honda IMA system. I have owned these systems, I know these systems. The Honda IMA system is severely crippled and it treats the battery extremely poorly. Toyota beat them to the market with a good hybrid. They made the Insight, first gen, and Toyota released the Prius right away and blew em out of the water. Scrambling to compensate for what they thought was a great lead in technology, they robbed life from the batteries to play the higher instantaneous mpg. Our civic hybrid had 3 brand new replacement packs under warranty. Everyone I know with an IMA system, had a replacement pack under warranty at least once. And their "fix" years later was to basically reduce the useful range again just like Toyota did from the start to preserve pack life, but the mpg's dropped significantly.

 

I was looking at an LTO conversion to replace the NIMH battery. 98 20ah 2.3v cells in series. Putting the voltage sense wires on every 7th cell.

Between 2-2.5v 16.5 usable ah.
2.8v upper voltage charge cut of is 274.4v so the system can't overcharge the cells.
With 7 in series rather than 6 the cells won't get too low to cause a system fault.
160amps continuous charge and discharge rating.
1/3 of the original packs internal resistance.
More efficient and significantly lower self discharge rate.

The only issue I see with this is that Prius won't use all of this extra capacity? If it's coulomb counting and only going to use a very small portion of that pack.

 

The Prius does coulomb count, but it also calibrates based on voltage during light/idle loads. So that part of it should work just fine.

 

Well that makes things far more interesting.

The 98s 2.3v LTO has a nominal voltage of 225.4v

I believe the lowests the battery ecu allows is 202v? At 7s between voltage sense wires that is 2.06v per cell which is above the rapid fall of voltage. Besides with such a larger capacity the Prius has more time to try and charge the battery before it got to that point.

What I've been trying to find is the upper voltage limit for a cell block and the upper limit of the pack as a whole before the system reports faults. Is it 19v per cell block?

Would be great to be able to charge it and get a few of kilometers of extra assistance.

 

Use a bluetooth OBDII reader and Hybrid Assistant / Hybrid Reporter apps on any android device. Then you can run your own tests for the battery voltage limits. For the discharge try leaving the Prius in neutral with the AC on. I think this will discharge the battery lower than anything except running out of gas or having the gas engine quit running. For the charge side just use a forced charge with the brake on hard, car in drive, and the gas pedal floored.

A large capacity LTO battery could respond differently than the stock NiMh pack. I suspect the voltage would tend to stay around 2.3V per cell and the extra capacity would go unused.

 

I have a copy of techstream, and have rebuilt a few HV batteries. So pretty familiar with the system. Have driven around with techstream logging data to see some stats( was surprised to see pulling 160amps when there is 125amp fuse). Have seen the low voltage fault codes from bad cells/ modules but was unsure what the high voltage limits of the system were before the battery control module would report an issue.

The LTO cells can't be damaged by the voltage range of the system, so that means I could charge them to higher than the stock NIMH with a charger. However need to know at what voltage, or sustained voltage will trigger a fault code. No point charging to 240v/250v/ 260v to get those extra ah if the computer doesn't like it.

 

I have been considering the same type of conversion. That pesky 1% scenario that burns down your car keeps getting in the way.

So I ran a forced charge and discharge to record the voltage extremes of the stock pack. On discharge I kept my Prius in reverse to a 44% SOC and then shifted to neutral to continue the discharge. AC was set to max cold, high beams on, and rear defroster on to maximize the current draw. At 44% SOC the current draw was about 8 amps. At a 25% SOC the master warning light came on. I continued to a 18.5% SOC. At that point the current draw was 3 amps. From 44% to 18.5% SOC the block voltages stayed steady at 14.5 volts. So 14.5 volts per block seems to be a hard limit for the stock pack. As the SOC drops the amp draw drops to maintain the 14.5 volts. I have seen block voltages go below 14.5 volts by running out of gas. To end the discharge test at 18.5% SOC I shifted out of neutral to park and the gas engine fired right up. On a forced charge the block voltages hit a high of 18.5 volts at 50% SOC and 75 amps.

The 14.5 to 18.5 voltage spread per block works out to 203 to 259 volts for the whole pack. It's easier to look at the block voltage to determine what lithium cells can match the constraints. With 7 LTO cells to replace a block their voltage would range from 2.64 to 2.07 volts. That's within the safe operating range for LTO cells. At the mid range of 16.5 volts per block the 7 LTO cells would be at 2.36 volts. That's close to full for a LTO cell. That does not leave much headroom for charging to a high SOC to make a plug in conversion. As a non plug in conversion 7 LTO cells per block seems like it will work.

For a plug in conversion you could try 8 LTO cells per block. The LTO voltage range would then be 2.31 to 1.81 volts. That's still within the operating range for LTO. At the mid range of 16.5 volts the LTO cells would be at 2.06 volts. That's close to empty for LTO cells. A large
(50 Ah or about 10 kWh) LTO pack could be charged to 259 volts and the Prius would just use up the extra voltage and Ah's.

All of this looks good on paper but until somebody tries it, it's just speculation. And there's that pesky 1% that burns down your car.

Matthew100, are you looking at salvaged LTO packs from Honda Fit EV's? That seems to be what's available.

 

Yes, it can go below to the UV limit of just about 180VDC for the pack and on the upper end, the car sometimes gets to 270VDC or a bit higher. Iout can get as high as 120A driving through the Rockies. In reverse you're limited to the current draw because of the HSD architecture.

It's these 1% cases that are a killer.

 

Well here are some numbers that may shock and amuse.
Pulling a 160amps. This was done after reconditioning and balancing a battery. Still had to replace a module afterwards ... and pulled harder on the battery. Same levels of volt drop but the min-max variation was tighter.