Old BMS General Thread

Threads relating to the BMS system begun by Peter Perkins

Moderators: GregsGarage, retepsnikrep

ChrisBarron
Posts: 244
Joined: Tue Dec 11, 2007 11:26 pm

Postby ChrisBarron » Tue Jan 06, 2009 5:48 pm

retepsnikrep wrote: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


That's pretty much what I thought too. I hate the idea of wasting anything (Yorkshireman at heart, living in Scotland,got a Jewish grandparent !) but it's not a lot at the end of the day, and probably is less than some connection losses.

User avatar
retepsnikrep
Posts: 1387
Joined: Sat May 26, 2007 4:50 pm
Location: North Yorkshire England
Contact:

Postby retepsnikrep » Wed Jan 07, 2009 3:19 pm

Touchwood. Still working fine! Weather still terrible here :(

Voltmeter finally working properly as it was also suffering from the IMA interference :roll: cured by a 0.1uf capacitor from Li-fepo4 negative to voltmeter negative supply rail. Supply is isolated anyway. That gives me a possible solution for the master interference.

I shall be re-locating/re-installing the master in next couple of days and running it alongside analog system. I've got a few clamp on ferrites/caps/clamp diodes etc now

I will reintroduce/re-connect each function individually starting with battery temp and battery current etc until it stops working, then I'll add the filtering to the circuit responsible which I am certain is the Master data bus 8)

I'm off on a 135 mile overnight trip on Monday 19/1/09 to see Greg in Scotland and talk BMS/EV's over beer! Image

I'm going to go for 135mpg (4.54L or One UK gallon) if I can.

Still tweaking pima to get better/best assist/petrol ratio.
Regards Peter

Two MK1 Honda Insight's. One running 20ah A123 Lithium pack. One 8ah BetterBattery Nimh pack.
One HCH1 Civic Hybrid running 60ah A123 Lithium pack.

ChrisBarron
Posts: 244
Joined: Tue Dec 11, 2007 11:26 pm

Postby ChrisBarron » Wed Jan 07, 2009 10:04 pm

retepsnikrep wrote:
I'm off on a 135 mile overnight trip on Monday 19/1/09 to see Greg in Scotland and talk BMS/EV's over beer! Image


What better thing to do !

I'm shelving my BMS for now because, as mad as this idea is, I think that if RF interference can be kept to a minimum then there is scope these days for a wireless BMS system.

I think the costs can be got down to £9 or £10 per cell.

I'm currently working on a project of which a part involves creating a short range wireless mesh at 2.4GHz (IEEE 802.15.4, probably Zigbee) .
Zigbee is relatively robust where interference is concerned and the prices of zigbee transcievers continues to fall. Some Zigbee devices are designed to run for years on a couple of AA cells. This periodic use of BMS and it's low data rate is an ideal application for it.

It would remove the cost of interconnection hardware, but, it will have to wait until my budget allows me to buy a development kit

User avatar
retepsnikrep
Posts: 1387
Joined: Sat May 26, 2007 4:50 pm
Location: North Yorkshire England
Contact:

Postby retepsnikrep » Thu Jan 08, 2009 4:40 am

Chris

Wireless link good idea.

I have certainly been thinking about this and have already bought a couple of 433mhz serial data modules to tinker with.

My idea is to use the wirless link to allow the Master to communicate with a remote display in the house so you can monitor charging etc on a seperate picaxe powered lcd display. You get remote alarm and direct data logging to a PC capability as well :wink:
Regards Peter

Two MK1 Honda Insight's. One running 20ah A123 Lithium pack. One 8ah BetterBattery Nimh pack.
One HCH1 Civic Hybrid running 60ah A123 Lithium pack.

ChrisBarron
Posts: 244
Joined: Tue Dec 11, 2007 11:26 pm

Postby ChrisBarron » Thu Jan 08, 2009 8:49 am

retepsnikrep wrote:Chris

Wireless link good idea.

I have certainly been thinking about this and have already bought a couple of 433mhz serial data modules to tinker with.



I've played with a few 433MHz modules and they seem a bit more susceptible to noise than higher frequency units. I built a GPS decoder and time relay unit which I fitted in my loft (just because I had the parts to hand) and built a couple of nixie tube clocks which used the 433MHz time signal as a clock source (with onboard backup if the signal got lost)

The problem seems to be in the type of modulation, even quasi AM-FM modules still suffer a bit from it. The 2.4GHz band stuff, such as IEEE 802.15.4 modules use much more complex digital forms of modulation which are designed to coexist in environments where ordinary 802.11 WiFi is being used.

I'm differentiating between 802.15.4 and Zigbee here, enough to say that 802.15.4 is the standard which needs to be observed, and Zigbee is an implimentation of a comms protocol which complies with 802.15.4. There are other 802.15.4 protocols too, my favourite (but only just being born so not sure how good it is yet), is called 6LoPAN , which allows IP traffic straight over a 802.15.4 link. That's all to do with 'The Internet of Things' and would mean that each cell could have it's own IPV6 internet address and assuming you had the correct link hardware you could check on your batteries from anywhere in the world.

As for cost, you can buy ready to program Zigbee controllers for £12.50, wwhich have the zigbee stack built in so most of the networking stuff is handled within it. Or for £5.50 Microchip cells a 802.15.4 transceiver, which can be loaded with a Zigbee stack or their own licence free 'MiWi' stack and you just control it from another cheap pic over SPI.

I've only used a picAxe once for a University introduction to microcontrollers project, and they seem okay to use, but they might get expensive in large numbers, though I'm aware their price is falling as time goes by too.

I can see where you're coming from with the 433MHz stuff, and used primarily for the home link when the motor controller isn't creating interference you should be able to get away with it. But if yoiu got a problem you could always step up to another higher frequency band. The thing to watch out for with those modules is that some are designed so that you must use Manchester Encoding, which is one way to keep a dc neutral signal voltage, but many of the moduels ought to come with circuitry to handle this now, like the 'Smart Radio' modules from www.rfsolutions.co.uk

I'm bogged down with a lot of other projects at the moment, and one of them could pay some bills so that will have to come first for now !

Chris

User avatar
retepsnikrep
Posts: 1387
Joined: Sat May 26, 2007 4:50 pm
Location: North Yorkshire England
Contact:

Postby retepsnikrep » Thu Jan 08, 2009 5:17 pm

I've re-fitted my master display cubby hole and put the Master board in a little recess behind drivers seat so I can get at it now without stripping down half the car! :roll: I might try and wire up current & temp sensors tomorrow on road towards full interference free operation.
Regards Peter

Two MK1 Honda Insight's. One running 20ah A123 Lithium pack. One 8ah BetterBattery Nimh pack.
One HCH1 Civic Hybrid running 60ah A123 Lithium pack.

User avatar
retepsnikrep
Posts: 1387
Joined: Sat May 26, 2007 4:50 pm
Location: North Yorkshire England
Contact:

Balancing

Postby retepsnikrep » Fri Jan 09, 2009 9:47 am

Balancing in progress,
charger set to 178v,
250ma balance load switch on at >3.60v,
about 40% of cells have reached balancing load switch on.

Pretty lights :D

www.solarvan.co.uk/bms/BalancingLight090109.jpg

www.solarvan.co.uk/bms/BalancingDark090109.jpg
Regards Peter

Two MK1 Honda Insight's. One running 20ah A123 Lithium pack. One 8ah BetterBattery Nimh pack.
One HCH1 Civic Hybrid running 60ah A123 Lithium pack.

User avatar
retepsnikrep
Posts: 1387
Joined: Sat May 26, 2007 4:50 pm
Location: North Yorkshire England
Contact:

Postby retepsnikrep » Thu Jan 15, 2009 1:46 pm

Brought current sensor and SOC calcs back on line with Master today. It's a factor of two out so boobed in the calculation code :roll: and I'll sort that bug out today. No interference issue at this stage. 8)

Might do pack temp sensors tomorrow, and then finally the data comms with the extra filtering.

Edit

Edited Master soc code and added 10x over sampling. Seems to work well on short test run. Oem Soc gauge fell from 19 to 3 bars (That's about 4ah). My soc display said I had used just over 4ah. May be reading slightly high we shall see. Now on charge.
Regards Peter

Two MK1 Honda Insight's. One running 20ah A123 Lithium pack. One 8ah BetterBattery Nimh pack.
One HCH1 Civic Hybrid running 60ah A123 Lithium pack.

GregsGarage
Posts: 870
Joined: Tue Apr 01, 2008 5:27 pm
Location: Galashiels, Scottish Borders
Contact:

Postby GregsGarage » Wed Jan 28, 2009 11:03 pm

Peter,

I have been thinking about how to implement your digital slave boards with analogue software on my Fiat. The main weakness of the analogue slave board design is that if a board fails you may not notice. This has been partially rectified with your latest design, the test button tells you that all the slaves are working and that the low voltage warning circuit is o.k. The only thing not tested is the high voltage warning, so a fault there will still go undetected.

As you are aware, I am using the the original analogue slave board design with my own EV Management Module (details can be found on this forum). I can easily implement your test button sequence with this unit, so that every time the ignition is switched on it sends a test pulse and waits for the corresponding signal back from the slaves. That just leaves the problem of how to verify the high voltage warning.

So here is my proposed solution. Modify your slave software so that a low voltage warning activates both low and high optos. My module would send a test pulse and then count the number of high pulses until it gets the final low and high pulse. If it doesn't count 1 pulse per cell then it flags a fault. A high voltage warning would still trigger just the high opto on its own, so it is still possible to distinguish between a high and low voltage warning.

Another idea is that it might be possible to just use the low opto for both high and low warnings. I think this may be possible because my module runs different routines when driving and charging. When charging I only monitor the high voltage warning. For driving its a little more complicated because I have regen, but my module knows when I am using regen and can figure out if a warning is for a high or low cell voltage. Haven't completely thought through this last idea, but I think it could work.

Greg
Greg Fordyce

Daewoo Matiz
http://www.evalbum.com/4191

User avatar
retepsnikrep
Posts: 1387
Joined: Sat May 26, 2007 4:50 pm
Location: North Yorkshire England
Contact:

Postby retepsnikrep » Thu Jan 29, 2009 2:26 am

All good ideas, fire away and be my guest.

I'll tinker with the software in due course as well. But I'm busy at the moment with my Insight Oem Bcm *Battery Control Module" fooler module.
Regards Peter

Two MK1 Honda Insight's. One running 20ah A123 Lithium pack. One 8ah BetterBattery Nimh pack.
One HCH1 Civic Hybrid running 60ah A123 Lithium pack.


Return to “BMS thread”

Who is online

Users browsing this forum: No registered users and 17 guests