HV Pack Voltages & Grid Charging

On the Honda side, people there build chargers with output in the low 100s of mA. Many NiMH manufacturers including Panasonic spec indefinite float charge rates of 0.033C to 0.05C -- 215 to 325 mA. Probably the best practice would be to read the SoC from the car just before shutdown, then program the timer to have the charger deliver a 250 mA trickle charge overnight that would result in 85% SoC. It is a lot easier to have a plug-in conversion kit do it for you. I looked into putting an HV charger into mine but finally decided that I would have spent just as much time fussing with the charger every day as I would have just by waiting for the Enginer pack to do it.

I don't know if there was a definitive post-mortem for aminorjourney's catastrophe, but I do know that you cannot charge NiMH packs in parallel without -dV/dt detection on each pack. The basic problem is that at full charge, the voltage starts to drop as the temperature rises, so a fully charge pack in parallel with undercharged packs accepts a greater share of the charging current, which then pushes temp higher and voltage lower. Now you have all charger current going to one pack, and the other packs in parallel are beginning to discharge into the overheated pack. Pretty quickly there is a thermal runaway. When I charge individual modules at 5 amps, the temperature rises quickly on approach to full charge, and continues to increase even after charging current is removed. Even if that Zivan had cut off in time, it may have been too little, too late.

 

Would a 200mA indefinite charge at 200V damage a NiMh battery?

The thing I can't understand is that charging the stock battery seems very difficult but when I'm' at 2 bars, the car will charge it to 7 within 5 minutes.

 

Charging the stock pack with the enginer kit is quite quick too, it'll go from 2 bars to eight in maybe 10-15 minutes.

personally i would never try and put -any- charge into the HV pack without the system in ready.

 

The Enginer kit doesn't actually charge the battery to that level. Because the battery ECU will adjust the SOC high if the voltage is high, you see a "fake" charge which would disappear if you shut off the kit. It takes a longer time for the battery to actually get to this level, so initially, the SOC (bars) read artificially high.

 

200ma is a safe indefinite float charge for the Prius pack, provided it is properly mounted (constrained) and the ambient temperature is cool.

I have done this many times without trouble. I would suggest if you were charging unattended, to also power up the battery fan if possible and/or have a temperature cut off so that the battery charger will shut off at 120F.

Interestingly, The Battery ECU actually has a mode that will allow it to automatically control an external charger and charge to 80% SOC IIRC.

 

That is interesting. I wish there was more information on that.

I keep coming back to the idea of a charger that would output anywhere from 200mA to 1A would work. In addition, have temp sensors, current sensors and perhaps me standing there holding a laser heat thermometer on the battery.

I realize this is stated in a simple manner but from reading about NiMh batteries, heat is the tell tale sign that the battery is fully charged. It seems like a secondary sign is that the voltage starts to drop. It seems like to me that if you monitor those two things closely, your chances of a failure drop greatly.

 

That goes against my direct observation. If i turn off with 8 bars, and turn back on (with the kit switched off) i still have 8 bars. How long would you expect it to take the battery ECU to reduce the reported SC?

 

Yes, this is one that I don't understand either. My impression was that the HV battery ECU was a talker, and not really a listener. The Enginer kit plugs in downstream of the current sensor, and the car must be in Ready for current to flow to the battery through the system main relays. If the battery ECU sees charge coming in at a higher voltage, wouldn't it just think that either the engine was running to charge the battery, or there was regenerative braking going on? I suppose I can hook up the AutoEnginuity and count some Coulombs...

 

It may not right away. Depending on several factors, it may coulomb count until a lower voltage which will trip the adjustment routine and you will see the SOC drop suddenly.

Cutting power to the Battery ECU will force it to attempt to "guess" the SOC based on voltage and temperature when power is restored.

-Phil

 

I'm not sure what you mean by Talker/Listener. If you mean the CAN bus, then you are mostly right; it sends 4 different can frames periodically with no input needed. (talker) However it also receives other frames as well. (listener)

If you put power in/take power out, PAST (car side) the current donut (hall effect current sensor), the coulomb counting function will be intact and you will not have errors. If you put it in directly to the battery, you will have invalid SOC reported as I described above.

-Phil

 

Thanks Phil. That's what I meant, the battery ECU is not continuously receiving CAN messages from the HV ECU telling it how much current to expect. Enginer kits are supposed to be connected on the car side of the SMRs, so any Enginer charging current to the HV battery will flow through the current sensor.

Now, one could argue that, other than the charge just before shutoff, it is a waste to charge the traction battery with the Enginer kit, due to Peukert's...

 

Well, if you could EASILY and SAFELY charge, it would give you a whopping 600 watt-hours or roughly 2.4 miles of extra capacity. (and that assumes you are down to zero bars when you start charging!) I just don't think all the extra hassle and danger to the pack is worth it.

 

pEEf, as Seilerts says the Enginer kit is connected on the car side of the pack. I believed this meant that the coulomb counting is correct and that the reported SOC is real, and you seem to be saying the same thing.

Also, in the above scenario (charging just the HV pack), if you have a 15 mile round trip commute and usually get 50 mpg you use 0.3 gallons of fuel. If you charged from 0 to 8 bars, and got your 2.4 miles of EV range, your "MPG" goes up to 59.5MPG using 0.25 gallons of fuel. Nothing stellar, but interesting. might save you $50 a year in fuel.

 

Yes, you are right, I screwed up on that and forgot that the Enginer kit connects (properly) downstream of the current sensor. I am confusing this with a kit putting too much charging current here and setting codes. The proper way around this would be to intercept the CAN frames from the Battery ECU and lower them by the amount of charge current you are adding. (some conversions have hooked directly to the battery to avoid setting codes, and this is not a good idea!)

But it might cost you a pack, or worse; your entire car! (Ask Nikki Bloomfield, or the other handful of people who have had BAD BAD experiences charging the OEM packs!)

 

I've seen and heard the stories. After looking through all of the information, the main reason why I won't charge the OEM battery is due to cost. A good charger is really expensive.

What keeps me wondering is the fact that my car will go from 2 bars to 7 bars in less than 5 minutes. I can't wrap my head around why it is so difficult to grid charge an OEM pack.

 

It's not difficult if implemented properly, but as you indicate, it's neither cheap nor trivial to implement a proper charging system. The trouble and expense for 600 watt-hours of energy is hard to justify. I'd recommend you spend your efforts on an additional pack with more capacity if you are serious about implementing a PHEV. (Such as my solution)

 

I certainly discourage the external grid charging to the OEM pack, due to the extreme risk involved leading to fire and total loss of the vehicle.
But if you do, please keep your phone/camera and or photo camera handy; I and everybody else would like to be entertained.......
The pictures shown here are not inventions or photoshoped, had been the results of peoples distractions and or mistakes.
Murphy's law is around your corner....

 

Those pictures show multiple NiMH packs being charged in parallel at several amps, which is basically impossible to do safely. It is possible to charge a single pack at a couple hundred milliamps without catastrophe -- Honda Insight modders do it all the time, no Zivan required. But anyone who tries this on a Prius really, really, really needs to know what they are doing.

 

If you left the car in Ready, and applied power to the 500V bus it would safely charge the HV battery (i believe) but again, this is not going to be cheap, easy, or safe(in terms of wiring etc).

 

This will set a code! (ask me how I know! =)

Because the Prius OEM pack is so small, if you wanted to charge it, simply do as Seilerts suggested and charge at low constant-current. My tests indicate as long as the pack is cool (keep the fan running) you can charge indefinitely (no controls) with 200ma. This will usually charge your pack fully overnight.

You can also make use of the battery ECU's external charger control mode, put +12v into pin A3 and then hook a relay up with it's coil to pin A11 and +12v. This will also enable the fan to run as needed. (I'd recommend you also provide +13.8v to the battery so it doesn't go dead running the fan.)