Bcm Recal Mod

[retepsnikrep]

I've been using my PHEV Insight for a couple of months now and it is running great. The only issue is the way it interacts with the Oem Bcm and the Soc three bar bug.

Basically the Soc falls normally and then it sticks at 3 bars for about 20-30 miles before it drops right down and you have to regen (wasting fuel) to get it back to 4 bars before you can use further assist, even though my Lithium Cells still have masses of charge remaining.

I tried to fool the BCM with a high voltage tapped battery supply made up of pp3 cells 9v and this worked for positive recalibrations but threw a code as soon as you used assist/regen as the actual Lithium Battery Voltage did not match the dummy battery I was fooling the bcm with.

Now I have gone a stage further and have a potential divider suppling the bcm with the various battery voltages. This potential divider is now supplied by the Lithium pack in normal mode, but can be swtiched to my dummy 180v battery at will using a 12v change over double pole relay activated by a dashboard switch.

This is very simple and I will be trying it tomorrow when I drain my pack down to an indicated 3 bars Soc. Once it reaches 3 bars for a few miles I will switch to Mima mode 1 preventing any assist/regen, and then activate the relay to apply the full dummy battery voltage to see if it will recalibrate whilst on the move without errors. If it does? then that basically fixes/negates the 3 bar bug, as you can simply refill the Soc gauge at will on your journey by doing a positive recal when reqd without having the penalty of forced regen.

I timed a positive recal today as well, and it takes 2 minutes, give or take a few seconds, so you need to predicte the terain ahead a bit, and allow for two minutes without regen/assist etc to acheive this on the move Soc refill !!

Couple of pics of my Bcm Fooler, it's basically a 180v battery with a potentail divider and relay, you could use a dc-dc converter to supply the dummy high voltage but I had these cells which should last as long as the shelf life (5 years or so) as the drain is only a couple of ma for two minutes every time you do a forced positive recal!

www.solarvan.co.uk/insight/BcmFooler1.jpg

www.solarvan.co.uk/insight/BcmFooler2.jpg


Edit

A few more thoughts.

I'll draw a schematic for this if it works later today.

I'll note the part discharged voltage of my cells. When they get to the first 3 bar bug fall off point.

You may simply be able to introduce an additional voltage in series with the pack rather than substitute a complete HV supply, that also means you only need a single pole change over relay. A 12v 1w dc-dc converter could do the trick here.

If you want to positive recal and cells are down to say 165v you activate relay which adds another 12v from the isolated dc-dc onto you own battery voltage to bring it up to the positive recal point. Say 175-180V.

If this all works it means the Phev pack can comprise any number of cells, and that the bcm tap points don't have to be included in the replacement pack saving some additional wiring. I had to use 50 cells as that allowed taps at the right points in line with the oem bcm points. Learning all the time.

If using a potential divider all the cells /fake subpacks will appear to be perfectly in balance under all loads to the Bcm, so you have to have some other method (bms) to monitor your pack to your own satisfaction. I have that so I'm quite happy.

[retepsnikrep]

Well testing was interesting.

Some good news and some puzzling.

1) BCM worked quite happily with pd and lithium as voltage source. No issues

So that means a phev pack can be any number of cells/chemistry and we don't have to worry about BCM taps matching cell numbers etc.

2) BCM would Positive recall when switched to fake battery 180v supply at will when stopped with ign on, in auto stop, or sat ticking over. Even when only one bar down at 18 bars after a recent recal to 19 bars, the time it took seemed in proportion to how far down the soc gauge was. At 18 bars it only took 10-15 seconds to jump to 19 again, but at 3 bars it took over a minute before it started rising. So that's excellent news.

3) Now when driving and with Mima Mode 1 engaged all assist/regen cut off it only worked about 50% of the time, the rest if gave the voltage error code forcing me to stop and clear the code. I wondered wether it was the momentary cut in voltage as the relay changed over but that must only be a matter of 50ms at most? I thought about adding an Cap to the pd supply so that during switch over the cap maintains the supply for 100ms or so. I'll try that and report back. I need a 200v cap say about 47uf. Does that sound about right. The pd load is 2ma. It could be due to the fact IMA is still operating in background doing the engine pulse smoothing thing and therefore detects the voltage imbalance between it's supply and the bcm.

[retepsnikrep]

Here is the schematic for my BCM Fooler for use with non Nimh Phev packs.

It allows positive recals at will when stationery with engine running, in autostop, or if stopped with ignition on but engine off.

It also works about 50% of the time on the move in Mima mode 1 with all assist/regen disabled.

www.solarvan.co.uk/insight/BcmFooler.jpg

Using this disables the BCM's ability to see unbalanced cells/packs so you must have a replacement BMS capable of providing this function.

[retepsnikrep]

I've re-done my BCM fooler as per my schematic above and incorporated a dashboard switch with indicator led to activate it. I have binned the 180v extra battery and it now uses a tiny 12v - 12v dc-dc converter and micro relay.

The dashboard switch activates the dc-dc and the relay and basically adds 12v to the Lithium battery voltage before feeding it to the BCM potential divider. The current consumption of the pd is only about 1-2ma.

When not activated the potential divider is simply fed from the Lithium battery voltage.

I find that when partially discharged the resting voltage of the pack is about 160-165V so adding 12v to that brings it up to the recal zone. I could change the dc-dc to a 12v to 15v model as they age and voltage drops down.

Works well so far just need to make it work while on the move now.

[retepsnikrep]

Quite a decent trip today with new dc-dc converter Bcm Fooler. Topped up Soc twice while stationery and three times on the move without error (Mima Mode 1) so unless that was a fluke appears the dc-dc version works better somehow than the dummy battery version I tried first.

I'll keep you posted, if that continues to work when on the move, that's the three bar bug almost defeated for phev Insights !!

[JimIsbell]

This was interesting. 2 weeks ago I had an IMA light come on that was caused by a loose connection on the DC to DC converter. I fixed it and then to clear the IMA light I removed 12V for a second or two by removing the ground (earth) connection on the 12V battery. This cleared the IMA light, but the check engine light remained. I ignored it for two weeks. Then today I decided to try again and leave the ground off for a bit longer. That worked and the check engine light went out. I then headed on the road to get a haircut. The ASST and the SOC were totally blank...as expected...but then after a mile or so the SOC came to life with one bar and started counting up. I expected it to go the 10 bar level on the SOC where it was when I disconnected the 12 battery to effect the reset. BUT...it just kept counting up all the way to a FULL 20 bars in just a few minutes. I has stayed there ever since. It was a 6 mile drive home and the SOC is still at 20 bars. I have NEVER seen 20 bars but once, and it was at night with the headlights on. This time it was with the headlights off and just a leisurely cruise down the road. For some reason the "reset" I did caused a positive RECAL and found the battery to have twice the charge that it started with.

I am now wondering just exactly which 12V consumer is the one that needs to be interrupted to do a "reset" Instead of doing the "course" adjust by disconnecting the battery I want to do a "fine" adjust by just disconnecting momentarily the exact wire that needs to be interrupted. I want to put a push button on the dash to effect a "reset" without taking a wrench to the battery. This could also be used to do a forced RECAL such as what you are trying to do here. Much less involved and no circuitry just a wire and a switch....yeah, thats circuitry,...but not much.

Anyone know EXACTLY which 12V consumer I need to interrupt to effect this reset?

[joecvt]

Many people perform the MCM reset process in order to reset the SOC. It is performed by removing fuse 18 for a few seconds. It is usually performed with the key out of the ignition so I'm not sure what will happen if you wire up a switch to do it on the fly. My guess will be an IMA and check engine light.

Fuse 18 also resets the climate control but it is still better than a total reset by removing the ground from the battery.

JoeCVT = Just your average CVT owner

[retepsnikrep]

I tend to agree with Joe if you remove fuse 18 on the fly whilst IMA operating I would expect an error but I have never tried it.

Forcing a hard SOC reset with a standard battery may, depending on it's actual soc (Not indicated) and voltage, produce a positive recal if the voltage is abover 170 or so.

What I am doing is different in that I am not hard reseting the bcm, I just flick my dash switch for about a minute and supply it with a higher fake voltage (+12v) added to the phev battery voltage (160v). This brings it into the positive recal range (+170v) and after a minute it recals to 19 bars. Avoding the three bar bug, no autostop, forced regen etc.

I can do my Recal

1) Whilst stationery with ignition on and engine off

2) Whilst stationery with ignition on and engine running

3) In autostop

4) And since I changed the BCM fooler mod to be powered by a dc-dc converter I have found I can also do it on the move in Mima Mode 1 with assist/regen disabled whilst it is recaling. I've tried this about 5 times now and no errors, previously I was getting them about 50% of the time.

So now when SOC falls to three bars for a while, after say 20-30 miles, when terrain allows I select Mima Mode 1, maintain speed with engine, flick recal switch, wait one minute then resume normal operating with full SOC and loads of lovely assist!!

I haven't tried doing a recal with ima actively assisting/regening at present.

[retepsnikrep]

Still no IMA errors with this mod, and I tried it with some trepidation driving along yesterday without disabling assist/regen (Mima1) and still no errors.

I still need to test that aspect further but it seems quite stable!

More testing today and drove several miles using assist/regen with it (BCM Fooler) turned on and no errors.

Seemes that as this sytem tracks the voltage (albeit with +12v added when activated) rather then remaining fixed at 180v or whatever my dummy battery supplied it does not give an error.

I found it worked best when no assist/regen was being used, as the battery voltage fluctuates quite wildly due to assist/regen when in use, and this keeps taking BCM out of the recal zone for a second or two and then it obviously has to wait another minute.

So Mr Three Bar Bug we have you at last!!!! Die!

[Mike Dabrowski 2000]

Well done Peter,
Nice solution.
FYI-I have a reset stitch on the underdash IMA reset fuse, and it will set a code if opened while driving.
I will put the car in neutral, ignition off, and while coasting, flip the switch to open the circuit for a few seconds. Then I turn the ignition back on, and keep driving.

Auto positiverecal?
Put an opto detector on the SOC bargraph, centered on the 4th bar from the bottom. When that goes out, apply the dc/dc voltage boost, until the bar comes back on.