grid charger/balancer

[Cobb]

Its truly universal? I may consider one for my tool box incase I come across a hybrid that needs a balance.

The i2 has 4 10 mm bolts near the breaker when the cover is removed that are hot when the key is on. 2 positive 2 negative. I balance mines with a beach key in the ignition and turned on. This way the blower operates and keeps the pack cool if needed.

When you do a force regen the car seems 100% at 117 volts. The balancing process seems full when its reads 120.2 volts.

My best indication other than frequent negative recals is the fact you can not obtain a 100% soc on the soc screen. Once done

Ive seen it take a good 24 hours to balance, 6 or so to fully charge at 350 ma. When balancing Ive seen the voltage wonder upwards to 127 and down to 109. I wait til it holds for an hour at a set voltage 120.2

FYI, since I balanced my pack with the enginer phev, its almost eliminated the p1586 ima codes when using ev mod for extended amounts of time. I balance every 90 days. Just make sure to hook a 12 volt charger to the battery up front as it will go flat in afew.

[Mr. Mik]

Originally Posted by fechter

...
...

Questions:
what is a typical end of charge voltage?
Is there any voltage limit that should be observed when doing a CC charge?

BTW, my discharge method is to simply drive to work. There is about an 800' climb on one hill that could fully drain the pack even when it was new.

Around about 172V at 20degC ambient temperature with active cooling.

The end of charge voltage is temperature dependent. A cool NiMH battery reaches a higher voltage; once it starts heating up, it drops it's voltage (= negative delta-V). Of course it also depends on the charge current.
If the battery is warm or hot to begin with, then the end of charge voltage stays lower. I think the battery heats up to the temperature at which the self-discharge rate equals the charge rate.
Running the cooling fans during grid charging is important because those cells that reach full SOC earlier can heat up quite severely; their negative delta-V will hide the ongoing voltage rise of the not-yet full cells. There may only be 10 cells (or less) that need much longer charging than the rest, their continuing voltage rise will be undetectable if uncontrolled heating of the rest of the battery hides it.

The heating of a fully charged NiMH battery continues after the end of charging, particularly if the battery is hot. The self-discharge rate increases with temperature, in extreme cases it can lead to thermal runaway. If you end a grid EQ charge with a hot battery, then 30-50% of SOC could be lost to self-discharge in the next few hours. Maybe that was behind the unexpectedly low energy content of your battery after the first grid charge?

I calculate the minimum charging time for the first EQ grid charge (rather than watching the voltage). I assume that some cells are at 0% SOC at the start, calculate how long I need to charge to certainly fill them up, then charge for that amount of time with constant active cooling of the battery. At C/10 I use around 140% of capacity for the first charge (8.4Ah for these cells) and 120% of capacity (7.2Ah) for subsequent cycles. At lower charge currents, like the C/17 at 350mA for these cells, I would over-charge to a higher multiple of the capacity, maybe 180%, because apparently the charge efficiency is lower (higher losses during charging). Lower charge currents are also much safer. Infrequent long and slow overcharging is often said to "rejuvenate" NiMH cells, but I have not tested this properly.

[Mike Dabrowski 2000]

My harness design will have a small isolation relay for both the PTC or other temperature sensor (thermistor) and the fan.
With the charger off, the PTC and fan/blower are connected to the stock circuits.
When the charger is turned on, the small pc boards with relays will switch the two circuits so the PTC/thermistor and the fan/blower are connected to the charger 12V and isolated from the cars internal circuits.
This eliminates any potential issue with back feeding 12V into the cars circuits, and allows full control of the fan and measuring of the stock temperature sensors.
I may also provide a bench harness/connector that would have alligator clips and a long insulated probe like my loaner to make quick safe connections to a pack, without requiring permanent connections.

The early civic packs are a bit tricky to charge on the bench as they do not have a fan, so a 12V fan mounted to a cut out foam block as an an interface would be required to get the cooling air to pass through the pack.

In the car, the early civics require the IMA cover to be installed as the stock blower draws air out of the IMA chamber, and the battery cooling air will only be drawn through the pack with the cover installed.
It is not possible to make a quick connection to the pack while in the car and still put the IMA cover on, so I made a small box with 12V fan that sits in the rear window over the air inlet, and blows air through.
The later civics feed the blower output into the pack first, so those can be charged with a quick connection and the IMA cover is not required.
All civic blowers require a PWM speed control signal and have a tach output, so the civic harnesses will have a small PC board with a microcontroller based PWM control circuit built in.
Today I will be placing the big order for 50 of each charger sheet metal enclosures.
I will have the aluminum chassis and covers powder coated with a bright golden yellow like the warning label. Should look very professional.
Grid Charger - MIMA Honda Insight Modified Integrated Motor Assist

Cobb,
Be careful of topping the pack with high currents, one can cook the cells even with blower running, as the stock system is designed to stop charging when near the top.
What current does the engineer system charge the pack with?

[Cobb]

115 volts and 15 amps. Its less that the 117 the car uses. I can see when its nearing full as I coast to a stop that the phev system cuts out.

[fechter]

Originally Posted by Mr. Mik

The heating of a fully charged NiMH battery continues after the end of charging, particularly if the battery is hot. The self-discharge rate increases with temperature, in extreme cases it can lead to thermal runaway. If you end a grid EQ charge with a hot battery, then 30-50% of SOC could be lost to self-discharge in the next few hours. Maybe that was behind the unexpectedly low energy content of your battery after the first grid charge?

I calculate the minimum charging time for the first EQ grid charge (rather than watching the voltage). I assume that some cells are at 0% SOC at the start, calculate how long I need to charge to certainly fill them up, then charge for that amount of time with constant active cooling of the battery. At C/10 I use around 140% of capacity for the first charge (8.4Ah for these cells) and 120% of capacity (7.2Ah) for subsequent cycles. At lower charge currents, like the C/17 at 350mA for these cells, I would over-charge to a higher multiple of the capacity, maybe 180%, because apparently the charge efficiency is lower (higher losses during charging). Lower charge currents are also much safer. Infrequent long and slow overcharging is often said to "rejuvenate" NiMH cells, but I have not tested this properly.

OK, that makes sense. My cooling was minimal on the first run and the cells got pretty hot. I now have a fan running and the temps stay down. I don't think I've run a long enough cycle to guarantee full balancing yet, but I can definitely see a huge improvement in performance after the 3 cycles I've run so far. This is very promising.

One issue is the SOC indicator will start out at whatever it was the last time the car was running. If I start charging with 50% SOC, the next day, the car starts out showing 50%. Sometimes it will recal, sometimes not, depending on whether I can get it to max out on regen. This seems to confuse the IMA system if it doesn't recal, as it can think the pack is nearly drained when it really has about half left.

[Mike Dabrowski 2000]

That will be something to look into. It would be great to get the car to do the positive recal after each charge.
Uhtrinity, Peter and others have posted some information on that.
I think the pack voltage needs to drop below a certain value before charging, and then when the system wakes up after the charge, the voltage must be higher than a certain value to get the positive recal.

I don't think the batteries actually continue to heat after charging, they are probably just equilibrating the higher internal temperatures with the cooled outside shell, so running the fan after the topping is finished is probably a good idea o dump all of the heat stored in the thermal mass of the cell.

I was not planing to providing a fan on off control, the fan was going to run whenever the charger is turned on, but that may not be the best thing if the charger will be used for long term maintenance charging. I will look into switching off the 12V to the fan. I still have a couple of more days work on the PC boards, so I can still build this into the design.

[retepsnikrep]

Originally Posted by Mike Dabrowski 2000

That will be something to look into. It would be great to get the car to do the positive recal after each charge.

I think the easiest way to do this as you are powering the fan and have a 12v supply available is to power the BCM during charging as well so it can count the current in. That way you will have 20 bars on start up with no waiting for a recal and a very lively warm battery . It's a minor power penalty with a lot of benefit. A 1A diode in the bcm power feeds and you tap in with a 12v supply and it's basically sorted. For those with the BCM gauge they will also be able to see the OEM soc increase and other data during charging. I see it as the only reasonable option to do what you want/need.

The only other way to do it is force the battery voltage to ~175 or so for at least two minutes undisturbed on switch on. Not really practicable with the OEM system without a BCM fooler and complicated relay switcher to fake a higher voltage on the BCM taps etc. Also the BCM voltage boost has to be kept to <12v or you will get an IMA error due to voltage difference between MCM and BCM. MCM will be seeing actual voltage, BCM will be seeing faked voltage If you drive the vehicle at the same time that gives another load of issues as voltage drops under load interrupting the recal cycle.

[Eli]

I really like the idea of powering the BCM while charging. That's something I want to explore.

I've planned on making a wireless battery pack voltage monitor, but if I were able to power the BCM while grid charging, that would make it very easy - With the wireless BCM Gauge, I could just bring the receiver into the house with me. Then you could monitor SoC and temperature, in addition to voltage. That would be awesome!

[retepsnikrep]

Originally Posted by Eli

I've planned on making a wireless battery pack voltage monitor, but if I were able to power the BCM while grid charging, that would make it very easy - With the wireless BCM Gauge, I could just bring the receiver into the house with me. Then you could monitor SoC and temperature, in addition to voltage. That would be awesome!

That's something we should explore on the BCM gauge thread, see you over there.

[Mr. Mik]

Originally Posted by Mike Dabrowski 2000

...
...

I don't think the batteries actually continue to heat after charging, they are probably just equilibrating the higher internal temperatures with the cooled outside shell, so running the fan after the topping is finished is probably a good idea o dump all of the heat stored in the thermal mass of the cell.

The NiMH self-discharge is an exothermic chemical reaction. The high internal gas pressure of a full cell reduces as the gases re-combine and release the energy as heat. Although the cell continues to produce heat, the temperature will usually fall once the higher core temperature has equalised with the surface temperature (unless in thermal runaway). You could test this by heating a stick with SOC <50% to the same temperature as another full stick under C/10 charge that has reached a stable temperature (40-50degC on the bench in 20degC ambient air without fan cooling). The over-charged stick will cool down much slower than the other one. You could also charge two sticks to 120%, then discharge one and compare their rate of cooling down.

I was not planing to providing a fan on off control, the fan was going to run whenever the charger is turned on, but that may not be the best thing if the charger will be used for long term maintenance charging. I will look into switching off the 12V to the fan. I still have a couple of more days work on the PC boards, so I can still build this into the design.

Switching it off could be beneficial in Winter. I thought you were including programmable temperature on- and off-points, including something like intake air and exhaust air temp difference? Like:"Turn on cooling if battery temperature is above 16degC AND intake air is below 20degC".