Home grown BMS ideas !

How do you store and manage your electricity?
User avatar
acmotor
Senior Member
Posts: 3595
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Home grown BMS ideas !

Post by acmotor » Sat, 24 Jan 2009, 21:56

Yes, there is a lot of battery development in the pipe line. The very development we should have had years ago. AND the development that will make the hybrids look like vintage cars.

I guess to some extent my decision to buy TS was that it was the best value offered so far and it was ACTUALLY available.
Otherwise we all just grow old waiting for the promised EVs and batteries !

I'm glad you follow the circuit.

Image
iMiEV MY12     105,131km in pure Electric and loving it !

EVLearner
Groupie
Posts: 51
Joined: Tue, 06 Jan 2009, 02:57
Real Name: Malcolm Moore
Location: Turramurra

Home grown BMS ideas !

Post by EVLearner » Sat, 24 Jan 2009, 23:48

Hi AC Strange as it may seem, but the circuit that you have is a virtual digital realisation of an expanded analogue voltmeter circuit.

Am I guessing correctly that the separate circuit on the left hand side is an over-voltage shunt to assist in self-balancing the series connections of battery voltges. If this is the case, then I would look to be integrating the two circuits so that the component count is cut by at least 3.

I like the idea of directly mounting the boards on the batteries - that is the general idea - isn't it? Image

User avatar
acmotor
Senior Member
Posts: 3595
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Home grown BMS ideas !

Post by acmotor » Sun, 25 Jan 2009, 02:47

Correct.
LHS is voltage shunt.

I did consider combining O/V and shunt but that would make them at the same voltage and that is not quite the plan. Shunt voltage can (needs to) be set independent of O/V.
i.e. shunt is normal operation O/V or U/V is not if you get my drift.

The shunt is set to come in at 4V so eq has a chance to work before any O/V 4.2V) signal (O/V must shut off or at least considerably reduce the charging current (to less than the shunt current))
Also when used with regen the shunt does its job (1-2kW on my pack) without giving an O/V.

The other thought was to keep the voltage monitoring as just that. Separate from the shunt (and more likely to fail) circuit.

I am pleased you are examining the circuit and making suggestions.
Go for it !
I will try to justify my thinking but realise that it will probably have holes in it !

Boards directly mount on TS battery (40Ah) but could be adapted to others in the range.
This idea is quite blatently stollen from others!!!
Some people have a one size fits all PCB. I am not the optimistic yet !

iMiEV MY12     105,131km in pure Electric and loving it !

EVLearner
Groupie
Posts: 51
Joined: Tue, 06 Jan 2009, 02:57
Real Name: Malcolm Moore
Location: Turramurra

Home grown BMS ideas !

Post by EVLearner » Sun, 25 Jan 2009, 04:18

Hi acmotor

Well that answered the question that I was obviously going to ask - what are the voltage thresholds? Does the programmable zener have a sharp knee, and how sensitive is it to temperture (I will check these on a data sheet if I can lay my hands on one!)

I had a thought about arranging a VCO (voltage controlled oscillator) into the Opto, so that the frequency could be proportional to the voltage (over a limited range) - that way the microprocessor could scan and read the cell voltages, as well as know the cell is OK - any thoughts on that?

User avatar
acmotor
Senior Member
Posts: 3595
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Home grown BMS ideas !

Post by acmotor » Sun, 25 Jan 2009, 04:41

Accurate 1.24V 1% for the zener, very sharp (mVs) knee. See a TLV431 data sheet (5 pin sm pack) just google for pdf.
O/V 4.2V , U/V 2.5V , shunt 4V although I may move this a bit.

VCO is interesting idea.
My earlier micro version sent serial data via opto but this meant the interconnections were not a simple one wire daisy chain. Also the 4800 baud data rate was too slow to scan 220 cells. 5 bytes x 220 cells = 1100 bytes at onlt 480 bytes/sec and this was before handshake. It was looking like taking 5 seconds for a scan.

As per previous posts, I was reluctant to leave simple analogue as I had experienced micro crash.

Please explore the VCO idea further though as it has merrit.
iMiEV MY12     105,131km in pure Electric and loving it !

EVLearner
Groupie
Posts: 51
Joined: Tue, 06 Jan 2009, 02:57
Real Name: Malcolm Moore
Location: Turramurra

Home grown BMS ideas !

Post by EVLearner » Mon, 26 Jan 2009, 04:48

TLV431 is almost a reference diode and a virtual 741 all in one - nice bit of work there! And at $0.12, resistors are almost more expensive!

With the VCO I was thinking about 131 kHz for 13.1 V, and direct opto scanning into a BCD counter, gated into an 8 or 16 bit bus, on a 10 msec synch per cell or about 2.2 sec for 220 cells - which is still too slow! Better KIS (Keep It Simple).

With the shunt currnt management, has this been tried yet? It concerns me that the TLV431 may leave the switched transistor in an active state - not saturated , and I can't easily see how to impliment a little hysterisis into the circuit so that the transistor 'switches on' and 'switches off' - bit like a schmitt trigger - maybe a 2.2 M from the collector to the TLV input might cause the circuit to switch of the switching transistor does not 'switch' - just a thought!

User avatar
acmotor
Senior Member
Posts: 3595
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Home grown BMS ideas !

Post by acmotor » Mon, 26 Jan 2009, 06:16

Played with that. Caused oscillations. But I wasn't really bothered since the dissipation from the transistor was in worst case only 2W and it is rated at 20W. It is mounted on plenty of PCB copper by design.
It is all just running linear analogue and seems to work fine.
The transistor saturated at 1A is more than 1V anyway, so it is not really 'saturation' at these low voltages. A FET could be used with low enough Vgs and Vds but the shunt is just a heat loss system so KIS.

After an hour shunting 1.5A at 4V at 25°C ambient the transistor tab is at 45°C and the resistors at 65°C. Measured with IR thermometer.
Real life ambient will be higher but then I see no eveidence from the test bank that eq. will need to run for that much time.

Shunt resistors could be from (edit: forgot to finish the sentence) 1 ohm to some high value so shunt current can be selected between about 3A and not much at all depending on charger rate and regen etc.

The VCO idea still interests me. Keep thinking on it. How does the sync work between BMSs ?
Last edited by acmotor on Mon, 26 Jan 2009, 10:59, edited 1 time in total.
iMiEV MY12     105,131km in pure Electric and loving it !

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Home grown BMS ideas !

Post by Nevilleh » Mon, 26 Jan 2009, 21:38

I have used a couple of SMPS in series - 2 x 27V @ 5.4A - quite successfully as a "48V" charger. They are adjustable from a nominal 27V +- 10% and they are current limited at a bit over 5 A. I am charging a 16 x LFP40 pack so I set the two supplies to their max which equates to 4V per cell. They provide CC then CV, admittedly not quite at the TS spec of 4.2v/cell, but I haven't noticed a significant capacity drop using this voltage.
To use them for PbA I would set the o/p to 56.4 or maybe a bit less so as to get CC up to that voltage. A normal PbA charger would drop to the float voltage once the current decreased (at 56.4V) and you won't get that, but if you're fussy you could drop the voltage manually. The advantage of these things is you can buy 2 of them for about $70!
There are a number of SMPS providers that import off-the -shelf stuff from China at pretty good prices. Just watch that the over-current protection is current limiting rather than shutdown/reset.
And you can get higher current ones for a bit more money.

User avatar
acmotor
Senior Member
Posts: 3595
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Home grown BMS ideas !

Post by acmotor » Mon, 26 Jan 2009, 22:39

Yes, SMPS are quite versatile and the outputs are well isolated from supply or ground.
Nevilleh, what equaliser are you using in the 16 x LFP40 string ?
iMiEV MY12     105,131km in pure Electric and loving it !

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Home grown BMS ideas !

Post by Nevilleh » Tue, 27 Jan 2009, 03:09

I have to admit that my equaliser is a digital multimeter.
I have been checking the cell voltages whilst charging and just recording them to get a handle on what they do. So far I have found less than .08 volts variation between the 16 cells I am testing. I stick my trusty Dick Smith variable power supply on each one after that to bring it up to 4.2 volts and record all the times etc. The variation is surprisingly small and I wonder if I can get away with just bringing the total pack voltage up to 60 volts. This leaves the cells at voltages from 37.1 to 37.9 and I wonder if this is enough charge - taking them up to 4.2 volts doesn't seem to give appreciable increase in driving range, although I have not measured this precisely as my route varies.
While this is happening I have been working on the design I saw in eetimes-india and just about have a working prototype.
I'm running the cells in an electric scooter and I use the voltmeter to tell me when I am down to about 44volts, then I check the individual cells to see how low each one is and again the variation is quite small. I guess I am accepting a small range/capacity compromise by stopping when the worst cell is at 2.5volts and not worrying about balancing at that point.
My BMS prototype is for only 4 cells but easily extendable to 16 if it works as promised and that will automatically take care of balancing low voltage and high voltage cells and still give me a readout of the lowest and highest voltages.
I have a BMS from TS and it gives the cell temperatures as well, but I'm not sure if this is really needed. Possibly if I ran 5C discharge currents for long times, temperature might be something that I need to monitor too. But charging at 20 amps only increases the cell temperatures by about 7 08 degrees over the couple of hours needed for a full charge. At 5 amps there is not much increase at all.
It's fitted to a big motorcycle and I don't want to pull it to bits just yet! It's interesting that it also shows a voltage range of about .08volts from best to worst and I wonder if it does indeed attempt any balancing at all.
I do need to come up with a good, cheap solution as I plan on fitting 200-odd LFP40 cells to my electric car project and I don't want to destroy them any earlier than necessary.
Just in passing, it strikes me that a couple of 12v 15 amp SMPS's connected (parallel the outputs) across each side of the +- 330volt battery pack would provide the 12volt power needed to run all the other car stuff!

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Home grown BMS ideas !

Post by Nevilleh » Tue, 27 Jan 2009, 03:18

Sorry, need to reply to my own post! Duh..
Shift the decimal point one place left in those thirty-something voltages! Also, 7 to 8 degrees.
Must be the couple of glasses of Aussie red.....

antiscab
Senior Member
Posts: 2327
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Home grown BMS ideas !

Post by antiscab » Tue, 27 Jan 2009, 08:08

a point on charge voltage for TS.

measuring recoverable AH on discharge, the difference between charging to 3.4v and 4.2v was around 0.75AH for a 40AH cell.
as long as current has fallen to around 0.1A at a voltage higher than 3.4v, the cell is charged.

im not sure what other ramifications a lower charge voltage has.

Matt
Matt
2017 Renault zoe - 25'000km
2007 vectrix - 156'000km
1998 prius - needs Batt
1999 Prius - needs batt
2000 prius - has 200 x headway 38120 cells

User avatar
acmotor
Senior Member
Posts: 3595
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Home grown BMS ideas !

Post by acmotor » Tue, 27 Jan 2009, 08:11

I also wonder if there is a downside to LV charging of TS.

I know that with my SLA if you go to 13.6 to 13.8V and wait for current to drop (12-16 hours) the battery is full. Same as 'fast' charge to 14.5V.
iMiEV MY12     105,131km in pure Electric and loving it !

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Home grown BMS ideas !

Post by Nevilleh » Tue, 27 Jan 2009, 20:35

I did read on the Tesla Motors web site that they had dropped the charge voltage by .1v in the interests of battery longevity. Seems like you can do more harm by over-volting the things than anything. Mind you, they are/were using 18650 cells. I'm really starting to wonder if just monitoring the cell voltages is going to be sufficient. I suppose they might get more and more out of step as the number of cycles goes up so you need to re-balance them from time to time to try and keep them together for as long as possible.
Can anyone tell me why Li cells should be so expensive when Li is the 3rd most abundant element? The processing doesn't seem that complicated.

User avatar
woody
Senior Member
Posts: 1711
Joined: Sat, 21 Jun 2008, 02:03
Real Name: Anthony Wood
Location: Mt Colah

Home grown BMS ideas !

Post by woody » Tue, 27 Jan 2009, 21:24

I did some sums in another thread to see if the drop in minerals market would likely affect battery prices, the raw materials came to 10-20% of our price.
The lithium patents should show you how to do it but legally prevent you :-)
I think the chinese suppliers have ignored the patent laws.
There are photos somewhere online of a TS cell being disected, that should help you if you are going to DIY :-)
Planned EV: '63 Cortina using AC and LiFePO4 Battery Pack

User avatar
juk
Groupie
Posts: 179
Joined: Sun, 02 Mar 2008, 02:53
Real Name: Craig
Location: Perth

Home grown BMS ideas !

Post by juk » Wed, 28 Jan 2009, 01:32

Lithium carbonate is selling for around US $8 per kg up from US $0.50/kg a few years ago, which works out to about US $180 per car for our purposes. Remembering that that is for battery grade material, non battery grade material still sells for only 1USD/kg.

Here's a recent diatribe i pulled from my archives:

There’s been chatter here and there about how much recoverable lithium there is in the world, and whether our move toward electric vehicles powered by lithium-ion batteries will create a “peak lithium” scenario.

It all started with William Tahil of U.K.-based Meridian International Research, who back in January 2007 published a paper questioning whether the automotive sector’s expected embrace of lithium-ion technology for next-generation plug-in vehicles was a wise move. Tahil is a fan of the zinc-air battery, largely because “zinc is the only metal which can sustain large battery production in the volumes required by the global automotive industry.” Needless to say, Tahil’s first report whipped up a firestorm of controversy, as you’ll see from some of the comments in a past post here.

Geologist R. Keith Evans published his own report in March 2008 in response to Tahil. Evans’ conclusion: “Concerns regarding lithium availability for hybrid or electric vehicle batteries or other foreseeable applications are unfounded.” Tahil returned fire four months later with a July 2008 report, arguing that Evans failed to make a distinction between practically recoverable lithium carbonate and resources where lithium concentrations are too low to economically exploit. Evans’ document, wrote Tahil, “is not useful for the industrial and strategic planning purposes of the battery and automotive industries. It confounds geological lithium deposits of all grades and types with economically viable reserves that can be realistically exploited and relied upon as a dependable source of sustainable supply by the mass production scale of the automotive industry.”

Evans, keeping the debate alive, issued a quick retort. He argued that it wouldn’t take much of a price spike to economically recover lithium from spodumene deposits, which Tahil had ruled out. He added that other sources of lithium can also be extracted economically as the price of lithium creeps up, which will be necessary to unlock these reserves. “A rise from the current levels is probably necessary but the cost of carbonate in batteries is a very small percentage of the battery cost,” wrote Evans. “Where hectorites, geothermal brines, oil field brines and jadarite stand on the cost ladder remains to be determined.” Evans also called Tahil’s report “alarmist” and “ludicrious.” Ouch.

Of course, Tahil raises other concerns, such as energy security. It doesn’t make much sense, he argues, to move away from oil and all its geopolitical risk and toward lithium, which offers up another batch of geopolitical risk. China, for example, has its own lithium reserves but it’s unlikely to share that with the west. North America gets its lithium mostly from Chile and Argentina, and while Bolivia has huge reserves, that country is beginning to behave like Venezuela. In fact, according to TIME, both Toyota and Mitsubishi have been knocking on Bolivia’s door, hoping to get in on the lithium action, but nobody is answering. Mitsubishi has said that demand of lithium will outstrip supply in less than 10 years unless new sources are found. (Hat tip to Earth2Tech)

Perhaps Tahil’s assessment isn’t so ludicrous, after all. Besides, it’s not an issue of whether the resource exists, it’s a matter of who holds it, how much of it is accessible, at what cost, and at a given time. We saw what that perfect storm of factors did to the price of oil. And unlike oil, lithium batteries will be part of the cars when you buy them; we’re not talking fuel that you pump in later after the vehicle has been purchased. The question must be asked: How would a rapid, steep climb in the price of lithium affect automotive sales? Even if it was a short-term climb, it could have devastating effects.

Toronto-based TRU Group Inc., a leader in lithium resource research, issued a report last week — commissioned by Mitsubishi — which flicked at some of those long-term supply issues but didn’t seem overly concerned. In the short term, TRU said the economic downturn is actually creating a lithium glut. The market, it wrote, “will be pushed into oversupply this year through 2013. Global use of lithium will decline sharply by at least 6 per cent in 2009 and demand is unlikely to bounce back any time soon as consumers put off buying laptops or cell phones containing lithium batteries.”

Notice that there’s no mention of electric or plug-in hybrid vehicles in this period. The impact of those markets won’t begin to be felt until 2013, before which any introduction of the vehicles will be quite limited. Come 2015 the market will regain momentum:

The long range, however, remains bright because new and large uses for lithium will start having a major impact on demand within the five year horizon: Lithium use in electric vehicle batteries and lithium alloys for aircraft. TRU forecasts that demand will be strong and sustained in these two segments over the long term 2020. The industry does need at least one of the announced pipeline production projects to come into production and also could do with another new project as the market tightens around 2015-2017. New lithium producers still will need to be cost competitive with existing salt lake brine based producers in South America and China. Emerging technology may make some of the undeveloped medium sized (brine) lithium resources quite attractive. Certainly the industry through expansion and development of new resources will have no problem meeting demand.

The company said it would post its full report sometime on Tuesday.

So, does all this make you feel more comfortable with the lithium supply-demand situation? China and Chile certainly can’t complain. That said, this isn’t just about forecasting out to 2020. Lithium needs to support decades of growth in both the consumer electronics and automotive sectors, and while recycling of lithium will help, will it help enough?

That said, unlike oil/gas/diesel, the battery is part of the car and can easily be swapped out with different chemistries. By 2020, who knows what chemistries will lead the energy-storage race? To quote GM vice-chair Bob Lutz: “People keep saying we’ve used up the whole periodic table on battery composition and that lithium-ion is about as good as it gets. I don’t believe that.”

Besides, it’s not only new chemistries that could come along, so could technologies that blend different chemistries and energy-storage systems. I’ve got a piece today in MIT Technology Review about a new energy-management system developed by Indy Power that can take two or more different batteries/storage systems and balance them off against each other in a way that optimizes both performance and system life. The system is flexible, allowing multiple combinations with only a software upgrade. It means a car could be designed in the future that blends a little bit of lead-acid, a little bit of lithium-ion, with a touch of ultracapacitor.

If it could be manufactured for less than a 100-per-cent lithium-ion vehicle, if it got better performance, and if the life of each battery system was extended as a result, this could be the way to go…
Last edited by juk on Tue, 27 Jan 2009, 14:34, edited 1 time in total.

User avatar
coulomb
Site Admin
Posts: 3611
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Home grown BMS ideas !

Post by coulomb » Thu, 29 Jan 2009, 08:51

woody wrote: I saw a4x4kiwi contactors which split his pack into 12 x 48V packs and a 24V pack for charging. It made sense for me to join them all in parallel to 1 48V pack for use with a big charger.

The way to do this was DPDT contactors:
    48V Bus + ---O  O--- + 48v block - ---O  O--- - 48V Bus
                   /                        \
prev 48V block ---O  O-----(bypass)-------O  O---  - next 48 V block
When the contactor was energised (show), the blocks join together to make the 600V chain, when unenergised, they join onto a common 48V bus for charging.

The reversing contactors have 4 terminals on each end. I thought they were double pole double throw, but they aren't are they, a reversing contactor is pretty much quad pole single throw, which you can wire as double pole double throw, I don't think I can do the bypass thing anyway unless I use even more hardware.
I think you might be wanting this:

Image

It has the unfortunate effect that each group of cells has three contact sets in series with it (two from one DPDT contactor). But this way, each cell can be active, bypassed, or charging, and you can arrange for the no-power situation to be all paralleled to the bus (charging position).

Edit: Presumably, the coils of all charge/run contactors would be paralleled, so all cells would be in the run (or bypass) position, or the charge position. Also, you could use 1, 5, or as many chargers in parallel as you can afford. The charger bus would have to be thick cables, though, or the last batteries on the bus could be undercharged (though hopefully, the BMS would overcome this).

Image These contacts obviously have to be of the break before make variety, and DC rated at full pack voltage Image Edit: and pack current.

Edit: There are ways to reduce the number of contacts per group to 2, but I can't see any safe way that avoids race conditions.

- Coulomb
Last edited by coulomb on Wed, 28 Jan 2009, 22:11, edited 1 time in total.

User avatar
woody
Senior Member
Posts: 1711
Joined: Sat, 21 Jun 2008, 02:03
Real Name: Anthony Wood
Location: Mt Colah

Home grown BMS ideas !

Post by woody » Sun, 01 Feb 2009, 07:25

coulomb wrote: I think you might be wanting this:

Image

It has the unfortunate effect that each group of cells has three contact sets in series with it (two from one DPDT contactor). But this way, each cell can be active, bypassed, or charging, and you can arrange for the no-power situation to be all paralleled to the bus (charging position).

Edit: Presumably, the coils of all charge/run contactors would be paralleled, so all cells would be in the run (or bypass) position, or the charge position. Also, you could use 1, 5, or as many chargers in parallel as you can afford. The charger bus would have to be thick cables, though, or the last batteries on the bus could be undercharged (though hopefully, the BMS would overcome this).

Image These contacts obviously have to be of the break before make variety, and DC rated at full pack voltage Image Edit: and pack current.

Edit: There are ways to reduce the number of contacts per group to 2, but I can't see any safe way that avoids race conditions.
Bypass is hopefully rare enough for it to be a manual operation, but depending on the pack location (e.g. On Ute Tray vs Under Body) it may be desirable to have it remotely switchable.

What did you use to draw the circuit?

cheers,
Woody
Planned EV: '63 Cortina using AC and LiFePO4 Battery Pack

User avatar
coulomb
Site Admin
Posts: 3611
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Home grown BMS ideas !

Post by coulomb » Sun, 01 Feb 2009, 17:06

woody wrote:Bypass is hopefully rare enough for it to be a manual operation, but depending on the pack location (e.g. On Ute Tray vs Under Body) it may be desirable to have it remotely switchable.
You might change your mind if you have an older pack with a handful of bad cells.
What did you use to draw the circuit?
MS paint Image

- Coulomb

User avatar
acmotor
Senior Member
Posts: 3595
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Home grown BMS ideas !

Post by acmotor » Sun, 01 Feb 2009, 19:33

As a pack ages the monitoring of battery condition becomes more important as Coulomb says. Although the infancy failure needs to be monitored as well, it usually occurs when you are watching things closely anyway.

It seems that many EVers just wait for something to go bang. I have seen some rather scary battery demolition even in lead acid ! Although some failures can be spontaneous, these usually result from pulling too much current !

Fortunately, most batteries weaken evenly so the user gets the message to re-power without individual battery failure.

All this makes BMS important.

The use of contactors to bypass is IMHO not a requirement as batteries are generally reliable and you may be better off stopping altogether and going for the tow truck if there is a problem. This is assuming there is a BMS to say stop now !

One factor to consider is the contact resistance of all the contactors.
This can vary from 10 to 30mohm and will contribute considerable to the effective ESR of the battery pack.

Coulomb, keep thinking about the parallel charging of modules though. Add in some diodes in the charging lines. It would be intersting to see if you come up with something similar to my (yet to be posted !) circuit. Your ideas will probably be better than mine so I will sit back ! Image
iMiEV MY12     105,131km in pure Electric and loving it !

User avatar
coulomb
Site Admin
Posts: 3611
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Home grown BMS ideas !

Post by coulomb » Mon, 02 Feb 2009, 00:14

acmotor wrote:The use of contactors to bypass is IMHO not a requirement as batteries are generally reliable
Well, it just seems criminal that one bad cell out of perhaps 220 can stop you from getting home. Bypassing a module would allow an almost normal ride home.
One factor to consider is the contact resistance of all the contactors.
This can vary from 10 to 30mohm and will contribute considerable to the effective ESR of the battery pack.
Eek, that is a lot. A contactor per cell is definitely out, then. And I think bypassing might have to be done with an insulated spanner after all.
Coulomb, keep thinking about the parallel charging of modules though. Add in some diodes in the charging lines. It would be interesting to see if you come up with something similar to my (yet to be posted!) circuit.
I'm up for a bit of design, but what's the requirement? Are you concerned that paralleling so many groups of batteries might cause a lot of current flow as they even out?

Perhaps concerned that the last battery on the bus will get less voltage, at least till the others have charged up?

Edit: Can you find a way to use less contactor contacts with diodes used only for charging? That seems worthwhile.

Something else I've overlooked?

- Coulomb
Last edited by coulomb on Sun, 01 Feb 2009, 13:17, edited 1 time in total.

User avatar
acmotor
Senior Member
Posts: 3595
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Home grown BMS ideas !

Post by acmotor » Mon, 02 Feb 2009, 00:52

Yes, cell failure can be an issue but some things we need to have faith in. Computer memory used to be parity checked. Vehicles used to have crank handles and they used to carry real spare tyres. My present dino burner has done 180,000 km without need of the spare (don't tell it though).
Monitor to know there is a problem but don't complicate or expensivate (yes folks, it makes sense as a word !!).
Mind you, module bypass is still an option.

Parallel charging does to some extent need to be controlled/ monitored so that individual modules don't receive overcurrent, although I have not seen this on the 12 modules in red suzi. Once the batteries are equalised once, they share charge quite well.

The idea is to use some blocking diodes rather than all contactors to distribute the charge to modules. This also limits any backfeed between modules in the case of a cell failure during charge. i.e. parallel charging has its downsides.

Your edit is onto the idea.
iMiEV MY12     105,131km in pure Electric and loving it !

User avatar
coulomb
Site Admin
Posts: 3611
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Home grown BMS ideas !

Post by coulomb » Mon, 02 Feb 2009, 02:23

acmotor wrote:Coulomb, keep thinking about the parallel charging of modules though. Add in some diodes in the charging lines. It would be intersting to see if you come up with something similar to my (yet to be posted !) circuit.
Ah, I think you might mean something like this:
Image

Now there is only one contact per group of cells. The diodes only carry charge current for one group, so they only need to be rated at fully charged pack voltage, and half the total charge current, plus some safety margin. Edit: diode current rating was misleading.

The movistor and fuse is in case the diodes break down and put (partial) pack voltage on the charge bus. Hopefully it would protect the other diodes and cells, and blow the fuse.

Edit: There should be fuses in series with each diode; the one in series with the charger is then redundant.

For extra effect, put a neon bulb (if they are still obtainable) in series with a pair of 100K resistors (for voltage rating) across the fuse, so you can see that there is trouble, and on which half. Unfortunately, if it's the first group or two, the neon may not light.

The vehicle chassis could be used as the common for one side of the charger bus, since the charging current will not be enormous.

A potential problem with this scheme is what happens if a contactor doesn't switch with the others. Kablooey Image   

Edit: Actually, that's not the case with the current version of the circuit. Please disregard the above sentence.

Also, each contactor has to break before the first contactor makes. So there is certainly room for improvement.

As before, you can have as many chargers as you want, though they have to be paired now. You could have say one pair in the vehicle for opportunity charging, and 6 more at home. (But that means a high current, relatively low voltage connector to connect to the external chargers, as well as a mains connector, which is perhaps not ideal). Maybe you could have all the chargers in the vehicle, but turn some of them off if you only have a 10A mains supply.

Also, this scheme does leave the possibility of a shock from two groups (so 140v DC shock if you manage to get your fingers across the two 70v charge buses). I suppose we could invert the negative batteries and put the whole lot in parallel. No need for paired chargers, either. Hmmm. Next post perhaps.
Your ideas will probably be better than mine so I will sit back ! Image
I doubt that. Maybe I've saved you the trouble of drawing part of your circuit, though.
Last edited by coulomb on Sun, 01 Feb 2009, 16:51, edited 1 time in total.

User avatar
coulomb
Site Admin
Posts: 3611
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Home grown BMS ideas !

Post by coulomb » Mon, 02 Feb 2009, 02:35

coulomb wrote:I suppose we could invert the negative batteries and put the whole lot in parallel. No need for paired chargers, either. Hmmm. Next post perhaps.
Oops, the negative side doesn't "invert" quite so easily.

So this arrangement might have to be confined to conversions where the positive and negative pack halves are well separated, either in separate boxes, or one at the front and the other at the back.

Improvement suggestions welcome.

User avatar
coulomb
Site Admin
Posts: 3611
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Home grown BMS ideas !

Post by coulomb » Mon, 02 Feb 2009, 03:58

If the pack fuses are put in series with the battery connections to ground, then at least any contactor problems like make before break or a flashover will be safe. (In the sense that blowing a multi-hundred amp fuse is ever "safe" Image)   Of course, that leaves all the batteries in one half of the pack at more than half the pack voltage; one end (across the fuse) is at full pack voltage. So that's probably a nono.

Post Reply