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 » Mon, 02 Feb 2009, 04:46

coulomb quotes coulomb in coulomb's post Image is that a new version of coulomb's law ? Image

Close.
I place fuses and each diode at the battery module so wiring is safer. I also use 800V PIV diodes to cater for (most) fault conditions.
(both wires to each module are fused at the module)

Never use earth return for traction pack currents. It is not safe and you risk super-imposing voltage on the 12V aux system.
Run a separate wire.

At present I don't use the changeover function of the contactor since the contactors I use are not rated for full pack voltage. I use the multiway charger plug to collect the module wiring together for parallel charging. This does expect the contactors to have openned ! (but both lines to modules are fused 3A)

I don't see an issue with your circuit if any contactor fails to operate. The diodes protect against that. The associated module will simply not get charged.

Can you re-draw with fuses at the diodes ? Your circuit is worth thinking about.

edit:spelling
Last edited by acmotor on Sun, 01 Feb 2009, 17:50, edited 1 time in total.
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, 16:01

acmotor wrote:Never use earth return for traction pack currents. It is not safe and you risk super-imposing voltage on the 12V aux system.
Run a separate wire.
I don't suggest this; only charging current. Traction current will use the ends of the batteries (+350 and -350 wrt chassis).
Can you re-draw with fuses at the diodes ? Your circuit is worth thinking about.
Sure, but right now I'm rushing off to the Ausindustry Green Cars Initiative event. Brisbane is first cab off the rank.

- Coulomb

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 » Wed, 04 Feb 2009, 06:58

acmotor wrote:Can you re-draw with fuses at the diodes?
Image
Edit: 2 right hand side diodes were wrong way around. Thanks, acmotor!

I'm not sure whether to include the contactors inside each battery group, or leave them, as shown, between groups.

I'd also prefer one charge bus rather than two, yet still allow the pack to be earthed in the centre.
Last edited by coulomb on Wed, 04 Feb 2009, 19:40, 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 » Wed, 04 Feb 2009, 21:38

Cut, paste and mirror has got the better of you !
Note diode A/K on positive charge bus.

I would still not use any earth return, even though your were thinking charge current only.
For two reasons....
Fault currents will also flow through earth (chassis). These may be more sizeable currents.
I would only use a single point connection (if used at all) between traction pack and chassis and this may be via a resistor anyway.

This second reason is also necessary to implement any earth leakage system.

Other than that, it is all starting to look very much like my arrangement (other than the changeover contactors as I have noted).
I do have some inline resistors to each module (1 ohm at present) to limit the unbalance of charge current when first starting the charge.

Try thinking one charger and multiple modules without the earth / centre and I think you will see there is a 2 wire bus that does the job.

I have included the contactors in the modules as this makes them a module !

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 » Thu, 05 Feb 2009, 06:52

acmotor wrote: Cut, paste and mirror has got the better of you !
Note diode A/K on positive charge bus.
Oops! Image   Corrected now, thanks.
I would still not use any earth return, even though your were thinking charge current only.
For two reasons....
Fault currents will also flow through earth (chassis). These may be more sizeable currents.
I would only use a single point connection (if used at all) between traction pack and chassis and this may be via a resistor anyway.

This second reason is also necessary to implement any earth leakage system.
Yes, I agree. Will change.
Try thinking one charger and multiple modules without the earth / centre and I think you will see there is a 2 wire bus that does the job.
I thought of a solution in the productivity booth (shower) this morning. Replace the contactor earth connections with connections to the most negative part of the pack. When all is switched, this should no longer be -350v anyway. (It might be floating, or earthed, or earthed via a resistor).

So you have a one bus solution for the Rodeo? Yet the pack is earthed (via a resistor) at the centre tap? (When the contactors are live).

I think I'd want all the contactors to be rated at full regen pack voltage.
I have included the contactors in the modules as this makes them a module !
In the sense that you can just daisy chain the hot ends, and parallel the charge bus connections, and there are no other components between the modules?

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, 05 Feb 2009, 07:20

coulomb wrote:So you have a one bus solution for the Rodeo?
Ah, no, I'm guessing this is already done in Red Suzi, with this box:
Image
BTW, are the fuses in there? Or in the modules? Image

When you say you can charge with up to 12 chargers... do you mean in series? I assume in parallel, for low voltage and safety. But maybe chargers don't parallel nicely.

You have me curious now.

Edit: minor rewording.
Last edited by coulomb on Wed, 04 Feb 2009, 20:23, 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 » Fri, 06 Feb 2009, 20:05

Here is version 3 of my battery paralleling circuit; many thanks to acmotor for his many suggestions:

Image

In this version, there is only one (set of) charger(s), so there is no possibility of a shock of two times the charger voltage. I've also made the contactors (a sort of distributed main contactor) part of the battery modules. Per acmotor's suggestion, there is only one connection to chassis.

The pack is still earthed in the centre, for maximum safety when active. However, the charger(s) have full negative pack voltage on them when active. This may stress their insulation, and a short to chassis or mains wiring will result in loss of on-board charging, along with a (small) fuse blow.

This circuit, more so than earlier earlier versions, assumes that the chargers and their mains inputs are not accessible when the contactors are active. Perhaps it would be better to have another contactor set that isolates the chargers when the contactors are active. Even so, the charge bus and its return are hot when the contactors are active, so these have to be suitably insulated (rated to half the pack voltage).

The 1K resistor near the bottom right corner is to bleed the VFD's bus capacitors without violent currents. It would need full pack voltage rating, and moderate power rating.

I'm still unsure where to put the pack fuses for maximum safety and effectiveness. At present, I'm thinking in series with the negative end of the leftmost battery module; that way any contactor flashovers would blow that fuse. Another one in series with the positive end of the right hand battery module would protect against pack+ shorting to chassis.

User avatar
evric
Site Admin
Posts: 500
Joined: Sun, 20 Jul 2008, 01:57
Real Name: Eric
Location: Adelaide SA
Contact:

Home grown BMS ideas !

Post by evric » Fri, 06 Feb 2009, 20:23

Do chargers like being paralleled?
Prius Plug-in Conversion: http://www.evplus.com.au ...Holden Barina EV: http://www.evric.kestar.com.au

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 » Fri, 06 Feb 2009, 21:28

evric wrote: Do chargers like being paralleled?
An excellent question, and I'm not that sure. For bulk charging, they are essentially switch mode power supplies, and each of these has a slight droop with load. I guess that means that some of them might not provide as much current as others. In extreme cases where the voltages are not well matched, only one charger might provide any current at all. This would not be bad for the chargers, it would just increase the time to charge the pack, and waste some investment.

There is also the possibility that one might switch to float mode (assuming lead acid here) before the others, and would be idle until the others switched to float mode.

Image Acmotor, is that why you use different multipin plugs for different numbers of chargers? To put the chargers in series and avoid current sharing problems? What if one goes to float when the others have not yet; could the output go negative, and would that cause problems? Perhaps you use only the CCCV method even on lead acid, which avoids the issue with one charger going to a lower voltage (float stage)?

I guess with a multipin arrangement as acmotor suggests, you can always split the pack up into suitable pieces (e.g. 3 48v sets if you have 3 48v chargers), and each set would be independent. You would use just single throw contactors, instead of the double throw contactors I have drawn. If you had say 11 modules and 3 chargers, you would split into sets of 4, 4, and 3 modules paralleled. The charger for the set of 3 modules would be idle for something like a quarter of the total charge time, which is not too bad. You can of course avoid this altogether by choosing the pack and module voltages carefully, as well as the number of chargers.

The other thing is that you need the multipin connector (socket only) to be rated at pack voltage, or half pack voltage if you use two (one on each half of the pack). Also, all but one modules are floating in this arrangement without the plug inserted, so some wires could float to plus or minus half the pack or so (with a few hundred picofarads of wiring capacitance behind it). Inserting the plug would presumably ground the modules.

Finally, I initially thought that the multipin arrangement doesn't lend itself towards having some chargers onboard, and some offboard (e.g. in the garage or at a reserved carpark at work). I suppose the onboard chargers could be wired to the multipin connector, and a special plug used to connect the onboard charger to the pack in parallel. The plug could stay with the EV near the socket, perhaps chained to the box with the connectors so it would not get lost. If the garage charger assumed the presence of the onboard charger to charge part of the pack, it could have a small circuit incorporated that sounded an alarm if the onboard charger wasn't outputting voltage (e.g. its mains connector not plugged in).

Then again, with the multipin connector and single pole contactors that just separate modules, there would be no need for the diodes. So what are you using the diodes for, acmotor? Image

So yet again, acmotor's solution is looking good. It allows great flexibility, without changing the wiring of the vehicle. If I get time, I'll do a fourth version with the multipin connector.

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 » Fri, 06 Feb 2009, 23:19

I am running out of secrets !!!!
coulomb is a cleaver cookie.

Connectors are 300VDC 6A IP67 18pin
Carefull choice of pin / voltage layout was made. (the fuses and diodes back this up as well)
There is one connector for +ve bank and one for -ve wrt centre(ground)

All good thinking there coulomb.Image

VFD has internal bleed down resistors and 'fast discharge' option in software already so don't worry about 1k resistor.

If you think of the principle of any control / charge wire leaving the module needing to be fused then you need to add a fuse to the -ve line (charge bus return) at the changeover contactor, as you have already commented on.
That is then all very close to what I am running other than I use a multiway connector to collect the -ve lines together.
A contactor could be added to your circuit (also as you note) to isolate the charging system -ve side from the charger(s).

Re the question re my charging box on red suzi. There are no fuses in there. Fuses are all are at the battery modules.

There are 'dummy' plugs that I put in to the charge box when not charging to retain the IP rating and also these contain the links that are part of the charge interlock / 'ign' Estop wiring of the EV.

Now all this is not perfect and a little fiddly to enter the charge mode but for a first off 600V EV it is probably quite a safe arrangement.

It does need to be a little more automated for Jo average to use.

BTW
The diodes perform an essential task (IMHO) as they stop any backfeed from the battery module to charger caps and also stop any fault condition of one cell / battery from draining any other modules that may be in parallel.

edit:spelling !
Last edited by acmotor on Sun, 08 Feb 2009, 20:03, edited 1 time in total.
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, 09 Feb 2009, 06:12

acmotor wrote:The diodes perform an essential task (IMHO) as they stop any backfeed from the battery module to charger caps ans also stop any fault condition of one cell / battery from draining any other modules that may be in parallel.

Ok, good points. But I still think it may be better not to parallel all the diodes' anodes before the multi-pin connector, so you can split the pack into separate pieces for charging.

Ironically, this eliminates the "charge bus" which is where this circuit started.

Example: 12 modules (6 each side of chassis), 3 chargers. You'd wire the multi-pin connector to split the pack into 3 sets of 4 paralleled modules. The negatives of each module could still be connected to the vehicle chassis (at one point only) for maximum safety when the contactors are inactive.

Then if you get a fourth charger, use a new set of plugs that splits the pack into 4 sets of 3 modules. If one charger dies, use your old cable that fed 3 chargers.

You can still parallel the whole pack with a third cable, for single charger use (on-board, solar, etc).

This way, no chargers are paralleled, so you never get any "lazy chargers", yet modules are paralleled to allow charging from a reasonable number of chargers. There is still at most one module's voltage potential at any point from the chassis (with the contactor inactive, obviously).

I think that will be version 4 of the circuit. Perhaps you don't have the anodes paralleled, acmotor? With 12 chargers and 12 modules, that would make sense, unless you lost say one charger. Then you might want to parallel all the modules and the 11 chargers to charge in 12/11th of the time, instead of twice the time (one charger driving two modules, and 10 chargers driving 1 module each).

Tritium_James
Senior Member
Posts: 683
Joined: Wed, 04 Mar 2009, 17:15
Real Name: James Kennedy
Contact:

Home grown BMS ideas !

Post by Tritium_James » Mon, 09 Mar 2009, 19:57

I realise I'm jumping in late to this topic, but one thing that scares me a bit with some of the circuits in this thread is failure modes.

For a BMS (and especially a lithium BMS) you need to think about what happens if something in your circuit breaks or ages. I think this is probably the biggest problem with the basic comparator switching a FET/resistor type designs - what happens when it's not sensing voltage correctly? This exact thing happened in the AC Propulsion balance nodes in the Porsche, they (after 6-8 years) started failing with the sense resistors and/or voltage reference getting out of whack. This means they were 'balancing' 12V PbA batteries down to 5 or 6 volts, which kills them pretty quickly!   This would potentially cause a fire in a lithium pack.

I think any serious BMS should have a micro on it (with watchdog, checksums, etc) and use two voltage sense circuits, which can be compared against each other. Any failure means that the unit defaults to not balancing at all, and shutting down the charger etc. Much better than over-discharging a lithium cell!

Just my opinion, of course! Please tell me if I'm wrong or paranoid!

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, 09 Mar 2009, 20:55

You are most welcome to question and even be hypercritical of us cave men. If we can't handle that them we limit our ability to learn !
BTW, sorry if I have been first to respond to some of your posts this morning, I just happen to be at the computer and am quickly typing before others get to their keyboards ! Image

Quite correct. Failure modes need to be considered.
Please humour me while I make some observations and excuses. Image

In your battery pack, was there other module or pack checks on voltage that would pick up the fault you describe ? As you know, reading things two ways can often pick up faults. BTW what BMS was being used ?

A TS lithium cell (LiFePO4) will not catch fire if discharged to zero. (it may be killed and have subsequent problems)

In the BMS circuit in this thread, the difference between micro drive and prog zener drive to the shunt transistor base is your point ?
Either system would require some shunt components (or charge shuffling components) so the question is, would a micro be better than the prog zener since the power side remains the same ?
The rest of the circuit is designed in the failsafe mode (on = good)
If there is a failure to 'on', the shunt probably still works as it is a separate circuit.
Regen (to me) means that the BMS at cell level needs to have a realistic shunt capacity. In the order of amps not mA.

Part of this overall BMS is monitoring module voltages (20 cells) and pack voltage (Yes, with a micro) to see it all adds up OK. This checks for wiring / terminal /contactor and fuse faults. It also checks for those inevitable BMS board failures !

A bipolar is used (normal fail to s/c in this type of circuit) along with parallel power resistors (normally fail open).
A bipolar failure will result (eventually) in a U/V alarm. resistor failure will result (eventually) in O/V alarm.

A micro is not immune to faults, so using a micro is not a gaurantee of function. ( I work with C8051F060,120 and 350 along with atmel mega and various PICs, mostly in C but also Bascom )
Yes, a lot of software goodies (and WDT) and the lack of (some) human assembly makes micros far more reliable than 'dumb' electronics.

Another factor is the cost. Once the BMS cost as much as or even 1/4 the price of the cell, there is the economic problem.

I have built a PIC based BMS. Reproduction cost was an issue and in reality there were probably more features than are actually required for Lithium monitoring. All you need is O/V U/V detection and some eq. shunt. I did also find that the environment on to of the cell was not friendly. Massive EMF, potentially high temps (particularly if there was a terminal contact problem). Once again, as you say, failure modes need to be considered.

No you are not paranoid. But lithium cells are not as evil in the BMS department as was early thought.

After all that, I do agree that the built in micro smarts are the future. I am probably just fighting the inevitable Image

Please, you are welcome to flame !
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 » Fri, 13 Mar 2009, 20:50

The battery paralleling circuit, which is getting a bit off topic, is continued at this thread.

Just to save people time searching (as I just did).

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, 23 Mar 2009, 16:16

I've been playing around with some TS LFP40s and trying various cell balancing methods and chargers. My design based on the EE Times article I saw a while ago is not yet perfected, but looking promising. I'll post details when I get it working OK.
In the meantime, I'm using a 64V, 6A current limited power supply to charge a series string of 16 cells with the voltage clamp circuit as shown to prevent over charging. I set the turn on point to 4.1V and the clamp is in full conduction at 4.13V. It carries the full power supply current (6A) easily with a 50 x 30 x 30 Al, L-shaped heatsink (3 mm thick) although I find that the clamp doesn't turn on for any cell until it is nearly fully charged and the current is down to an amp or so. At that level the heatsink gets barely warm.
For the low voltage end, I am just using a led voltmeter that comes on at 44V which is about 2.7V per cell and that has been adequate so far. The battery pack has now done 16 cycles using this setup and I can't detect any change - and I should hope not to!
It does rely on the charger being current limited, of course, but so do the batteries.
I built it (16 of "it") on a small pcb designed to fit onto the LFP40 terminals with the heatsink on top, so it just screws on, one to each cell.
I'll build a bigger one to fit LFP90 cells next - they are the cells I will likely use in my car, although I'd really like '160s, but the cost is a bit of a deterrent.
Image

User avatar
evric
Site Admin
Posts: 500
Joined: Sun, 20 Jul 2008, 01:57
Real Name: Eric
Location: Adelaide SA
Contact:

Home grown BMS ideas !

Post by evric » Mon, 23 Mar 2009, 17:11

Neville, Don't you think the clamp turn on point of 4.1V is a bit high? Ohters are using 3.65V etc.
Prius Plug-in Conversion: http://www.evplus.com.au ...Holden Barina EV: http://www.evric.kestar.com.au

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, 23 Mar 2009, 18:14

Good point Nevilleh.

One way to limit overvoltage is to have shunt Amp capacity > charger Amps. (provided you do not have regen !)

The fuse rating may be an issue. T03(P) TIP142 is Ic max 10A (pulse to 15 but that's not fuse blowing stuff) so a fast acting fuse of 6A may be more suitable given the charger current ? Mind you, it has all gone wrong by then. Just limiting the damage when you use only a semiconductor as the heater !

Check the heatsink capacity °C/W and follow the power de-rating curve for TIP142. Probably no issue given your observations so far.

Shunt voltage is still personal choice as a definitive number is still out there somewhere ! TS say 4.2-4.35 but....

Are you going to include over and under voltage monitoring ?

Have fun. Keep us posted !
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, 23 Mar 2009, 18:37

evric wrote: Neville, Don't you think the clamp turn on point of 4.1V is a bit high? Ohters are using 3.65V etc.

TS quote 4.2v as the fully charged maximum. This circuit turns fully on at 4.13v with the pot setting I use - the voltage gain is enormous with two darlingtons - so the voltage can't go any higher than that. Others are using a lower voltage for maybe two reasons: their circuit doesn't turn on so quickly, or, they are just being more conservative on the upper voltage limit. If you look at the TS charge graphs, the voltage is limited to 4.2v and held there for 70-odd minutes as the current falls to zero. At the time point where 4.2v is reached and the current starts to drop, the % charge is only 70.
The main reason I dropped it from 4.2v to 4.13v was that I read Tesla Motors have dropped their max charge voltage for 18650 cells from 4.2v to 4.15v "in the interests of increased battery life".
I have noted that if I charge the cells at 3.7V until the current drops to zero, then increasing the voltage to 4.2v causes a current flow (at the 6A limit) for only another 10 minutes or so, so they must be close to fully charged at that voltage.
As far as my car is concerned, I also wonder if I use 4.2v per cell as the max charge voltage, the whole pack of say 46 cells now produces 193.2v instead of the nominal 3.2x46 = 147.2. Sure it's only for a few minutes, but can my 156v rated controller handle that?
I'd be interested to other opinions.

User avatar
Johny
Senior Member
Posts: 3729
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Home grown BMS ideas !

Post by Johny » Mon, 23 Mar 2009, 18:45

Hi Neville. Why not remove the PNP darlington? The LM431 could drive the base of Q2 from it's Anode quite comfortably. Reduce semiconductor count by 1?
Reference R2 to Anode and place R1 from Anode to -ve (via fuse).
Like this but use your TIP142.
Image

Just a thought... so many less pins to solder.

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, 23 Mar 2009, 18:59

acmotor wrote: Good point Nevilleh.


The fuse rating may be an issue. T03(P) TIP142 is Ic max 10A (pulse to 15 but that's not fuse blowing stuff) so a fast acting fuse of 6A may be more suitable given the charger current ? Mind you, it has all gone wrong by then. Just limiting the damage when you use only a semiconductor as the heater !

Check the heatsink capacity °C/W and follow the power de-rating curve for TIP142. Probably no issue given your observations so far.

Shunt voltage is still personal choice as a definitive number is still out there somewhere ! TS say 4.2-4.35 but....

Are you going to include over and under voltage monitoring ?

Have fun. Keep us posted !


I designed the heatsink to dissipate 40 watts (Ic at 10A or so) and keep the junction temp below 150 deg, so its got a bit in hand. The fuse is that size because I had one and its main function as you surmise is to stop molten silicon spewing everywhere if I get something wrong!
I am working on a micro to monitor volts - since I know a bit about Atmel's AVR ones, I'm using an ATTiny25 Its only a couple of dollars and has 4 x 10 bit ADC in an 8 pin package and a serial comms capability that will do I2C. I figure to use one for each 4 cells to send volts to a master which may yet be my laptop, but might just be another Atmel chip driving a biggish LCD as I am much better at assembler and C than VB.
Problems yet to solve are bus addressing - I need to have the master assign addresses - and to decide if the accuracy is high enough at about 10mv resolution. I think it is and should be OK, especially if I set my upper and lower voltage levels a bit conservative.
The lower end particularly. Looking at the TS graphs again, the capacity drop from 3v to 2.5v is b--- all, so you might as well stop at 3v and not risk damaging the cells.

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, 23 Mar 2009, 19:03

Johny wrote: Hi Neville. Why not remove the PNP darlington? The LM431 could drive the base of Q2 from it's Anode quite comfortably. Reduce semiconductor count by 1?
Reference R2 to Anode and place R1 from Anode to -ve (via fuse).
Like this but use your TIP142.
Image

Just a thought... so many less pins to solder.

Not a bad thought. I originally used the pnp darlington to get heaps of gain and so a really fast turn on. I added the npn led driver later so I could "see the light". Might try your suggestion on my Veroboard one and see what the difference is.

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, 23 Mar 2009, 19:18

Well, I tried it and it isn't very good. Starts to turn on at 3.6v and is not in full conduction until 5.2v (without altering the pot setting from the original). I think the extra transistor is well worth it for the "snap-on" effect.

User avatar
Johny
Senior Member
Posts: 3729
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Home grown BMS ideas !

Post by Johny » Mon, 23 Mar 2009, 19:34

I should have looked harder. The TIP142 is not quite the conventional darlington (well not as I perceive them).
Image
The culprit is the 0.12 Ohm from base to ground on the second transistor. It gives it a sloppy turn-on characteristic. Done for speed me thinks.
Thanks for trying 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, 23 Mar 2009, 19:56

Yes, it's designed as a fast switch. I only used it because I had some from an SMPS project. If I was buying any, I'd probably look at something else as it is getting a bit old - the 140 would do OK as in this circuit it doesn't need the voltage rating of the 142.
Anyway, the circuit as is doesn't cost much and it works a treat!

User avatar
Johny
Senior Member
Posts: 3729
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Home grown BMS ideas !

Post by Johny » Mon, 23 Mar 2009, 20:01

I totally understand. I have various projects using TL494 PWM controllers which I use just because I have them!

Post Reply