Old BMS General Thread

[retepsnikrep]

Thanks Tesla, I thought Nicolai would have it sorted in a jiffy!

Reprogrammed slaves for temp Analog mode

Still tinkering, just finished temp Analog Master, and now fiddling with meter and led/switches which have to fit into the centre console cubby hole at front alongside MIMA display. It's the asthetic things that take the time Getting it look reasonable is sometimes the greater challenge.

I'm leaving the screen and digital master gubbins/cabling installed as I can work on that then alongside the up and runing analog temp system hopefully in few days.

Now hoping to get it installed/running tomorrow minus voltmeter which is on order.

The analog Master gives reduced functionality, but still covers the vitals.

Over Cell V audible Alarm when ignition on. (For driving regen)

Over Cell V visual Alarm at all times.

Over Cell V isolated charger cutback. (For controlling mains charger)

Under Cell V audible Alarm at all times.

Under Cell V audible Alarm at all times.

Temp is covered by normal IMA temp sensors when running and I have a simple digital gauge I may rig up as a temp measure.

Digital 0-199.9V meter also fitted running all time.

Digital Slaves (Analog mode) Test button propagates a low V alarm through the 50 slaves and activates low V alarm. If any of the 50 Slaves is off line/stuck/locked up etc signal won't make it to the last slave and therefore no Alarm will sound. Just press the button when you get in car to check they are on line, takes approx 500ms to respond.

[retepsnikrep]

It's In and working in analog mode

Been for a 25 mile trip just to check it out. BCM forced charge for a bit then gave an initial soc and then continued charging normally no positive recal. No IMA errors. Charge now at about 3/4 75% and I have just topped it off with the mains charger working under BMS control, seems fine and once one cell goto to Max V it held voltage steady and current started falling away. Zivan charger now adjusted to about 176V for the CV phase

Be interesting to see what happens next time I start car and if it notices the extra charge I put in when it was off?

(Greg) Seems the simple on/off for the charger opto works fine with the Zivan.

Just drove it normally for first test, no obvious issues. No over or under V alarms even under quite hard regen. Cells are very cold it's been about 0C all day and Insight BCM is doing it's normal NImh limited 50% assist. Seems normal.

Off to get an Indian Takeaway now as a treat/reward, I'll see what mpg I can manage on this 10 mile round trip.

Need to get some experience now with how it all interacts. Some cosmetic bits to do in boot as the cargo area is now a bit shallower so I need to cut the liner down 4 inches in height so it still fits. Need to finish bolting in cell trays from underneath as well and install pack voltmeter which should arrive in AM.

Some Pics

www.solarvan.co.uk/bms/Rear010109.jpg
www.solarvan.co.uk/bms/OnCharge010109.jpg
www.solarvan.co.uk/bms/IpuEntry010109.jpg
www.solarvan.co.uk/bms/Ipu010109.jpg
www.solarvan.co.uk/bms/Dash010109.jpg

Had to cut a small section of IPU cover to get clearance for cables.

Voltmeter will be going to right of leds in the blank area.

Edit

Didn't go for a further test in the end toady.

Secured batteries instead as a safety priority. They are bolted through side of wheel well into captive nuts on the two battery blocks/trays 6x6mm per tray. I'm pretty confident they would remian in situ now in the event of a roll over, I don't want to test it though

Charger finished cycle and turned itself off, some balancing taking place as I could see the cells turning the loads on/off. Charger was not voltage limited by BMS this time As cells get more into balance I may tweek the finishing voltage up a bit, I think I have it to low now but better safe than sorry during this testing phase. 176v divided by 50 cells averages out at 3.52v each at present.

Edit

This was the final version of my Analog Master schematic. Good job I made the Slaves fairly versatile

www.solarvan.co.uk/bms/AnalogueMaster010109.jpg

Edit

Interestingly I can still see the horrible noise on the Master Bus (Doesn't effect analog operation) but it will help when I start the filtering process after I have relocated the Digital Master. I will be re-wiring the Slaves with some nice twisted pair screened cable for the Master Bus and all the other links inside the IPU compartment. I've learnt a lot so far including how not to wire the pack for use in a very nosiy environment Didn't Edison say he found 999 ways how not to make a light bulb before he found how to do it

[retepsnikrep]

First Findings

45 Mile trip today 112mpg (UK) compared to my normal for this time of year abour 93mpg. Lot's to learn.

First thing I learnt was MIMA is vital, if you can't turn off forced charge and command extra power then it's a no go IMO. Well done Mike

OK How did BCM react? Well after topping it off last night via mains charger, it did not do a positive recal this morning. To be fair it was showing about 75% on dash gauge anyway. I started the journey with cells pretty cold at about 2C I had no autostop for entire journey due to cold air temp. Engine took quite a bit longer to warm up as the power was being offset by extra assist I altered my driving style from my normal peg instant FCD at 100mpg to peg it at 125mpg.

I used as much assist as was reqd to maintain 125mpg at 40-45mph over my 13 mile normal commute, traffic in town at end of journey was horrible, so that brought it down to the 112mpg. Initially charge fell away on gauge at about the rate I would expect for normal IMA use at this level, and it was def counting ah until it got to three bars. There it sat for the rest of the journey It appeared to be waiting for a nimh cell to go low which would have then triggered the gauge to fall to zero and stop assist etc. But as the cells didn't go low it just allowed assist ad infinitum. It was trying to force charge but that was disabled. It did not trigger an error, but just sat at three bars. Assist seemed to be limited to about 50% which was not a bad thing and may have been due to the battery temp or the indicated soc. Anyway it prevented me from abusing my new cells with some 100A assist

Battery temp at end of test was about 10C but car had been heated during hour long journey as well. No hot spots in pack when checked with infra red thermometer. Battery resting voltage was 165V at end of trip. 165/50 = 3.3v which is about the nominal voltage for these li-fepo4 cells. I don't know how many ah I used, but car is now on charge and I will know the kwh consumed from the mains in a couple of hours.

For the trip I disabled regen unless commanded with joystick, so it never charged except when braking/slowing down. I used manual assist and regen for this first test, sometimes when on a decent straight I put on one led of assist and then clicked the joystick to maintain it.

If one led fully lit is about 25A of assist that seems reasonable for the cells, which should be able to supply that load for about 90 minutes.

Sadly I'm back at work tomorrow and voltmeter hasn't arrived yet so that will have to wait.

[centrex]

Hi Peter
Re the noise problem you have.
Have you tried ferrite suppression sleeves these are the lumps you quite often see on laptop power supply cables and on the signal cable of computer monitors.
They came as plain tubes and split types that will clip around the cable, passing the cable twice through the core is very effective.
I have a great deal of success keeping rf interference out of microphone cables in gliders that I work on.
You may have to experiment with location on the cable for best effect.

Very interesting project you have.

[retepsnikrep]

Thanks Centrex. I shall be working throught the entire list of suppresion/filtering devices over next few weeks! It's still working well in Analog mode at the moment.

[GregsGarage]

Peter,

Great news about your car, I like the charging socket behind the rear number plate, reminds me of 70's american cars.

The test function of your digital slaves is brilliant. I am going to convert mine to digital with the analogue software and add the test function to my management module so that it checks the slaves automatically when the ignition is switched on. That is the main weakness of the analogue slave board, no way to know if one has stopped working.

Greg

[qdos]

Fantastic job Peter well done! Very tidy professional workmanship

Greg, We used to have a Ford Consul or was it a Corsair? with a fuel cap behind the number plate I used to love it as a 5 year old it was just so cool. I remember the Zephyr and Zodiac had it too

[retepsnikrep]

Peter, The test function of your digital slaves is brilliant. I am going to convert mine to digital with the analogue software and add the test function to my management module so that it checks the slaves automatically when the ignition is switched on. That is the main weakness of the analogue slave board, no way to know if one has stopped working.

Greg

I was quite pleased with that lateral thinking

As you can see it basically sends the low V signal in at Slave no 1, around all the Slaves and out at Slave no 50. It takes about 500ms for the signal to go round my 50 slaves running at the standard 4mhz. So the longest delay would be if Cell 1 triggered a low V alarm and the shortest would be for Cell 50. The only additonal wiring was two wires from Slave 50 back to my analog Master. I suggest you program yours for the standrad 4mhz speed as well.

You can tell which slave it not working as well if you get someone to hold the button down the data led would be lit on all upto the one with the fault.

Obviously if you convert to the Digital Slave layout now you will be able to use your Digital Master in due course. I suspect your vehicle is not the RF noise nightmare mine is.

We learnt a lot from your experience with the Analog Slaves (you were the first to install it) and I'm grateful for your continuing support.

I'm still chugging round with my new batteries trying hard not to abuse them in this freezing weather

Mpg is up 15% so far now running at a stable 108mpg on my commute.

Car develops so little heat though that even with rad block and hot air mod it barely gets up to temp on my 13 mile commute in this weather

I need to burn a minimum amount of fuel to get engine up to temp and provide cabin heating, so I don't expect ludicrous 150mpg figures until weather is a lot better!

Regards Peter

[ChrisBarron]

As you can see it basically sends the low V signal in at Slave no 1, around all the Slaves and out at Slave no 50. It takes about 500ms for the signal to go round my 50 slaves running at the standard 4mhz. So the longest delay would be if Cell 1 triggered a low V alarm and the shortest would be for Cell 50. The only additonal wiring was two wires from Slave 50 back to my analog Master. I suggest you program yours for the standrad 4mhz speed as well.
Regards Peter

Peter, I've read most of this thread and caught up with where you are.
I put my BMS on hold until I had a way to test how much energy is actually wasted during the balancing phase, when each cell burns off a few mW at peak voltage.

What is your estimate regards how much energy is lost during the balancing of any of your whole packs ? I've been wondering if it is worth going down the route of a biderectional dc-dc convertor on each cell whereby instead of just wasting the balancing charge it gets used to charge up a 6 volt cell, then, any cell which is low can take this charge and top itself up with it.

But before I go down that route I wondered how much energy we're talking about in the real world ? I'ld appreciate any comments.

Chris

[retepsnikrep]

Chris

Bugger all is the simple answer.

In my experience Lithium cells stay in balance very well. The balancing phase is pretty short and the energy wasted is a few wh at most. Def not worth bothering IMO.

50 x bi-directional dc-dc and the logic etc to control it doesn't sound cheap.

Peter