Low cost BMS

How do you store and manage your electricity?
T1 Terry
Senior Member
Posts: 991
Joined: Thu, 30 Sep 2010, 20:11
Real Name: Terry Covill
Location: Mannum SA

Low cost BMS

Post by T1 Terry » Fri, 18 Nov 2011, 18:54

It would interesting to fit a cell logger across 6 cells and see if the voltage changed between cells while under load. Theoretically each block of 3 cells should remain the same voltage if they evenly shre the load, that is if I understand the whole thing correctly. I've tried it on my 4 x 90Ah cells in parallel pack for my house batteries but the most load I can apply is 200amps so that's not much more than 0.5C, the voltages were exactly the same for the 10 minute test. I'm guessing 3C would be a different story.

T1 Terry
Green but want to learn

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

Low cost BMS

Post by Nevilleh » Fri, 18 Nov 2011, 20:30

Johny wrote: If you swap to a 16 x 4 line display you could show a block for each cell still shunting with the last line as text. 15 cells per line would make it easy to see which ones were which. I think I use character 0xFF for a block in the Vogue SOC display.
Image


Excellent suggestion! I'll look into doing just that. I had thought of displaying the cell numbers, but that needs an even bigger display. The blocks idea is much better.

User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Sat, 19 Nov 2011, 04:10

Tritium_Johny wrote: Weber, thank you so much for showing me the error of my ways. It must be great to be so smart. And good looking too.

Yeah, OK. I admit it. It's a forgery. The only person who's ever gonna write that, is me. Image

Seriously: Are you still not convinced about the unequal distribution of voltage drops, and therefore current through each cell?
me wrote:What everyone misses ... is that the upper left and lower right straps (between cells) are carrying approximately twice the current of the other two. Therefore they have approximately twice the voltage drop.
And even when we see that, we tend to assume that since top and bottom are mirror images, it will still come out the same for all cells. But it doesn't.

The more cells you parallel, the bigger the difference between the middle cells and the outside cells. Take an extreme case: 7 cells in parallel, connected to the outside world by diagonally opposite corners. Normalise everything to 1 ohm per strap and approximately 1 amp per cell to keep the maths simple.

Image
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

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

Low cost BMS

Post by Nevilleh » Sat, 19 Nov 2011, 12:57

Weber, your diagrams and stuff are things of beauty! How do you draw them and then post them?
I am stuck with crappy drawings like this:
Image

Consider 2 cells in parallel with the current going in and out of one cell as shown above.
The equivalent circuit is as shown also, where Vc = cell voltage, Ri= cell internal resistance, Rc= resistance of cell connections, Rl = load resistance and Vl = volts across load. The current flowing in the left hand cell is i1 and the current in the right hand one is i2.
Then           Vl = Vc - i2.Ri = Vc - i1(Ri + Rc)
Therefore      i2Ri = i1(Ri + Rc)
And          i2/i1 = (Ri+Rc)/Ri = 1 + Rc/Ri
So if the connecting link resistance is 10% of the cell internal resistance, 10% more current will flow in the cell that the load is connected directly to.
The Ri of my SE40AHA cell is about 4 milliohms. My Cu links measure 22 x 3 x 46 and that gives a resistance of 15.3 micro-ohms, according to the resistivity of Cu. So 1 + Rc/Ri is .0038 and therefore i2 = 1.0038.i1 or a difference of .38%.
I suppose this could easily be extrapolated to 3 cells, but I just can’t be bothered as I don’t think the current imbalance is significant.

Go on guys, shoot me down in flames
Image
Last edited by Nevilleh on Sat, 19 Nov 2011, 05:51, edited 1 time in total.

User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Sat, 19 Nov 2011, 17:25

Nevilleh wrote: Weber, your diagrams and stuff are things of beauty! How do you draw them and then post them?

Thanks Neville. These are done using the drawing tools in Microsoft Word, but some others by Coulomb and me are done using Open Office Draw. Both are object or vector oriented rather than pixel oriented. To post them, I zoom to the size I want, hit the Print-Screen key, paste it into Microsoft Paint, crop it, and save it as GIF for uploading to the forum.

But hey, there's nothing wrong with scanning a hand-drawn diagram. I have posted photos of my whiteboard too. The ease of doing that may make the difference as to whether you can be bothered posting it at all.

I know this parallel cell strapping stuff is a bit off-topic. Thanks for tolerating it Neville. But I just realised that in the 7 cell case the diagonal connection is not the best we can do. Assuming our figure of merit is the difference between the maximum and minimum voltage drops per cell, the best arrangement is to come in one cell from the corners. Doing that brings the max-min difference down from 9 to 4, as well as reducing the average voltage drop significantly.

Image

There isn't meant to be anything significant about the overlapping of links in the above two drawings. I didn't take much care with that, they're not even symmetrical.

And this "one cell back from opposite corners" rule gives the best result all the way down to 4 parallel cells.

With 4 cells in parallel "one-in from the diagonals" has the same max-min difference as using the diagonals, but the average voltage drop is more than halved. I must check whether Pascal realised this.

I found someone else who has figured this stuff out.
http://www.smartgauge.co.uk/batt_con.html
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

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

Low cost BMS

Post by Johny » Sat, 19 Nov 2011, 17:30

weber wrote:
Tritium_Johny wrote: Weber, thank you so much for showing me the error of my ways. It must be great to be so smart. And good looking too.

Yeah, OK. I admit it. It's a forgery. The only person who's ever gonna write that, is me. Image...
...
OK, I admit that I was looking at it with one-cell blinkers on. You have shown me (in your infinite wisdom) that the cumulative current along the cell connects causes centre cells to not have the same load as those nearer the corner connects to the next cell set. (Good pick-up Image )

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

Low cost BMS

Post by Nevilleh » Sat, 19 Nov 2011, 17:46

Weber, thanks for that, I've never tried the drawing tools in Word, might give it a shot.
Not sure if the battery connection stuff is all that much off-topic, its all to do with monitoring batteries and interesting enough in its own right anyway.
Had a look at that SmartGauge link and he puts it very well. My links are much lower in resistance than his at only 46mm long and 66 sq mm in cross-section, so the differences in currents are very much smaller, but I take it on board and certainly the best config for 3 cells has to be diagonally opposite corners.
Did you agree with my circuit analysis (added as an edit to the last post)?
I did work it out for 3 cells, centre fed, and using those values of cell internal resistance and connecting link resistance, then with 500A flowing I calculate there is 166.5 amps flowing in the outer cells and 167 A in the centre one.

BTW, the bms finally got all the cells to 2.8V, it took 54 hours! I now have it back on charge and it will shut off when one cell reaches 3.6V. Be interesting to see what sort of range I get and how long it takes to balance next time.
Last edited by Nevilleh on Sat, 19 Nov 2011, 07:40, edited 1 time in total.

User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Sat, 19 Nov 2011, 19:45

Nevilleh wrote:Did you agree with my circuit analysis (added as an edit to the last post)?
I did work it out for 3 cells, centre fed, and using those values of cell internal resistance and connecting link resistance, then with 500A flowing I calculate there is 166.5 amps flowing in the outer cells and 167 A in the centre one.
I can't fault your equations. It's only the numbers you're feeding into them that I question. But first let me say that the important thing you made me realise is that the apportioning (i.e. percentage) of current, and therefore charge, does not depend on the magnitude of that current.

In other words, it isn't only heavy discharges that matter. They will share unequally for the entire charge or discharge cycle, except near the two ends of the cycle where the internal resistance of the cells increases enormously and forces them to share more equally.

Unfortunately the bulk resistance of the links is only a small part. The interconnect resistance is dominated by the contact resistances at each end of each link. This can be particularly high at the aluminium terminals if great care is not taken. It would be good if someone could provide some measured figures for this, but I would not be surprised to find 250 uR per contact, which means your Rc is more like 1 mR, i.e. 25% of your internal resistance.
BTW, the bms finally got all the cells to 2.8V, it took 54 hours! I now have it back on charge and it will shut off when one cell reaches 3.6V. Be interesting to see what sort of range I get and how long it takes to balance next time.

Hoorah! Good on you for being willing to test this idea throughly. I look forward to future results.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Sat, 19 Nov 2011, 21:26

I wrote: "They will share unequally for the entire charge or discharge cycle, except near the two ends of the cycle where the internal resistance of the cells increases enormously and forces them to share more equally."

That last bit is wrong. They won't share more equally near the ends, they will swap roles. e.g. When the cells that have been supplying most of the current hit bottom, suddenly the cells that have been having an easy time of it will be forced to supply all the current.

Thanks for doing that analysis Neville. I assume the centre-fed group of 3 has exactly the same current-ratio as the side fed group of 2. Namely I1 = I3 = I2 * Ri/(Ri+Rc). Does anyone want to have a go at a formula for the diagonally fed group of 3?
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

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

Low cost BMS

Post by Tritium_James » Sat, 19 Nov 2011, 21:47

weber wrote:It would be good if someone could provide some measured figures for this, but I would not be surprised to find 250 uR per contact, which means your Rc is more like 1 mR, i.e. 25% of your internal resistance.


I've only got results for copper/copper resistances. Test setup: 300A lab supply, fluke77 multimeter on mV range. Test specimens: laminated 90Ah thundersky cell busbars, bolted together with M8 hardware.

Weathered copper strap: 71.7uR
Same, but undo and redo connection: 35.0uR
Cleaned (wire brush): 34.5uR
Cleaned, with contact grease: 29.3uR
Cleaned between laminations: 27.7uR

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

Low cost BMS

Post by Nevilleh » Sat, 19 Nov 2011, 22:20

Tritium_James wrote:
weber wrote:It would be good if someone could provide some measured figures for this, but I would not be surprised to find 250 uR per contact, which means your Rc is more like 1 mR, i.e. 25% of your internal resistance.


I've only got results for copper/copper resistances. Test setup: 300A lab supply, fluke77 multimeter on mV range. Test specimens: laminated 90Ah thundersky cell busbars, bolted together with M8 hardware.

Weathered copper strap: 71.7uR
Same, but undo and redo connection: 35.0uR
Cleaned (wire brush): 34.5uR
Cleaned, with contact grease: 29.3uR
Cleaned between laminations: 27.7uR


That all makes sense and I have to ashamedly admit that my resistance figure for the Cu strap was for one only and there are, of course, two! So that means I should have been using 31.something uR. And I didn't take into account the resistance of the terminal connection, mostly because I have no way of guessing that. So if I use TJs figures - presumably for 1 strap - and double them, my Rc is more like 60 uR and could be as high as 140 uR.
That turns my 1.0038 into 1.035 so the % difference could be as high as 3.5%. And if I have 170A flowing in my middle cell, the others have only 164A in each. Is that a significant difference? Starting to be, methinks. But that is a "worst case" condition. Oh well, one of these days I will move the straps to the diagonal opposite corners - but I won't rush into it.

T1 Terry
Senior Member
Posts: 991
Joined: Thu, 30 Sep 2010, 20:11
Real Name: Terry Covill
Location: Mannum SA

Low cost BMS

Post by T1 Terry » Sat, 19 Nov 2011, 22:40

Just on the drawing thing, a print screen and paste to power point is much easier to work with. Every thing can be moved around and size adjust, duplicated or completely deleted. The text boxes have spell checker and the saved JPEG file is much clearer from poer point than from paint.

T1 Terry
Green but want to learn

User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Sun, 20 Nov 2011, 00:13

Thanks TJ. Sounds like my guess was too high. I based it on the fact that Anderson rate the silver/silver contacts on their SB50s at 200 uR. But I guess the pressure is much higher with a bolted connection.

I assume that was a single contact you were measuring, between two TS90 straps, measured close to the contact. Please confirm.

Neville, we could assume that the contact for a 40 Ah cell with M6 bolt would be about twice the resistance for a 90 Ah cell with M8 bolt. Lets call it 60 uR to 140 uR. So I'm thinking you'd have 4 of those and two of your 15 uR bulk resistances. That's 270 uR to 590 uR for Rc. That's 6% to 15% of 4 mR.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

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

Low cost BMS

Post by Tritium_James » Sun, 20 Nov 2011, 01:18

Yeah it's hard to beat a bolted joint for contact pressure!

Yes, that was a single contact, measured at the back of the straps behind the contact surface.

Fastening hardware was bolt, washer, strap1, strap2, washer, spring washer, nut, all stainless steel.

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

Low cost BMS

Post by Nevilleh » Sun, 20 Nov 2011, 12:28

Finished charging yesterday and the charger reports that it delivered 121.4 AHrs in 8hrs 6mins. At that point the highest cell was 3.60V and the lowest was 3.37V. The total battery voltage reported by the bms was 152.6 for an average of 3.39V.
I switched on the dc-dc converter which was drawing maybe 1/2 amp and after about 25 minutes the high cell was down to 3.37V and the low one was 3.32.
Not quite sure what all this means yet, but I'll do the balancing again after running the car to get one cell down to 2.8V and see what happens.
I note also that when I ran the battery down last time, the low cell was at 2.77V and I noted that 30 mV difference stayed pretty much right through the charge cycle ie at the start of charging Vh=2.80 and Vl=2.77V.
Pretty hard to stop the car right at 2.80V as you still have to get home!
I was also a bit surprised that the charge was just over 120 AHrs as I expected a bit more.

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

Low cost BMS

Post by Nevilleh » Sun, 20 Nov 2011, 12:40

weber wrote:

Neville, we could assume that the contact for a 40 Ah cell with M6 bolt would be about twice the resistance for a 90 Ah cell with M8 bolt. Lets call it 60 uR to 140 uR. So I'm thinking you'd have 4 of those and two of your 15 uR bulk resistances. That's 270 uR to 590 uR for Rc. That's 6% to 15% of 4 mR.


I'm not sure about that as I note the Sky Energy/CALB straps are quite a bit larger than the TS ones for the same cells. Also, the pressure from a 6mm bolt would not be too different from an 8mm bolt, depends on the tightening torque and that's limited by the Al and Cu cell posts. Also, I don't have four, there are only two per cell.

My CALB straps are 22 wide x 3.0 thick and the length between holes is 46 mm. What are the dimensions of your test straps TJ?
The extra straps that I had made locally are even larger, a bit wider and a bit thicker, but they are used for the connections between each block of 3 cells.

Next time I'm out in the car, I'll try and get an observer to note the bms reported voltage and the voltmeter reading at some constant current and that will give me an idea of the total wiring resistance. Perhaps not relevant to the cell links, but useful to know. I have 95 sq mm cable joining all the things together.

Edit: I have a further thought in that I have some TS40 cells sitting on the floor here and I might connect 3 of them as a triple with the same straps I use in the car, stick on a 130A load (which is what I used for testing individual cells) and see if I can measure the voltage drops across the links. I'm sure I have a multimeter here somewhere with a mV range on it. Sort of like what TJ did.

Also, I wonder how one wold know when one cell in a triple is failing? Would that show up just as reduced capacity for that triple? I guess one reason for having a bms is to detect these sort of events!

More edit:
Having a coffee and started doodling the 3-cell diagonal feed analysis and before you know it, here it is!
In case you can't read my scanned, jpg'd, pencil sketch, the results are that the two outer cells have the same current and the middle one has more by the factor 1+Rc/Ri where Rc is the resistance of a single link. Whoops, I made a mistake in the last line - the factor is Ri/(Ri+Rc) and the current is LESS, not more! - excuse is the 'phone rang just as I was finishing.
Maybe I made some other errors, someone could check it for me? This is a surprising result, if true. Stick in Rc=250uR and Ri=4mR and you get .94 ie the centre cell carries 6% less current than the two outer ones.
Image
Last edited by Nevilleh on Sun, 20 Nov 2011, 03:35, edited 1 time in total.

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

Low cost BMS

Post by Tritium_James » Sun, 20 Nov 2011, 15:19

Neville, it's too early in the morning for me to check your maths :P   but a quick spice sim with your resistance values shows a 5.88% difference in cell current, so I think you've got it right.

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

Low cost BMS

Post by Nevilleh » Sun, 20 Nov 2011, 16:21

Thanks for that, I've rechecked it and I think its right too.
A very interesting result. If Ri goes to zero, no current flows in the centre cell! Of course, if Rc goes to zero, the currents are all equal. With the centre connection, same result if Rc+0 but if Ri=0 then all the current flows in the centre cell. Isn't maths a wonderful thing?
Putting some numbers into those equations:
Let Ri = 4 mR and Rc = 125 uR per link. Total current = 500A.
Then for the center connection, the side cells have 163.26A flowing in each and the centre cell has 173.47A flowing. The centre cell has an extra 10.21A, so its working a bit harder, nearly 6%.
For the diagonal opposite connection, the side cells have 168.52A flowing and the centre one has 162.96A, its working a bit less, about 3.4%.
And so we have proven, without a doubt, that the diagonal opposite configuration means the cells are stressed more evenly, about twice as good!
Anyone want to have a go at weber's 7 cell arrangement?
Last edited by Nevilleh on Sun, 20 Nov 2011, 06:32, edited 1 time in total.

User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Sun, 20 Nov 2011, 19:05

Thanks Neville. You've only half corrected the error at the end. It is kind of in the last two lines. What you want is
i1 = i3 = i2(1+Rc/Ri)
so the current in the middle is less, as you would expect.

Some readers may be confused as to why we say the imbalance in the diagonal fed case is halved relative to the centre-fed case (for 3 cells). I note that Neville changed his definition of Rc between the two cases. In the first case he lumped two connecting links together and called them Rc.

In the terms we are using now. The centre fed case has
i1 = i3 = i2/(1+2Rc/Ri)

I doubt anyone is parallelling 7 cells, so I think it would be far more interesting to stick with 3 cells and examine a more realistic model that includes the effect of different choices of link stacking, to convince yourselves that the way I've shown above is the best.

To do that we can break Rc (connector) up into Rb (bulk) and two lots of Rk (kontact). So Rc = Rk + Rb + Rk.

You could also work out what ratio of Rk to Rb would give theoretical perfect sharing with my proposed stacking. I say "theoretical" because we are ignoring variations in cell internal resistance which make it always a bad idea to parallel cells, except for the convenience of packing them better into limited spaces.
Nevilleh wrote:
weber wrote: Neville, we could assume that the contact for a 40 Ah cell with M6 bolt would be about twice the resistance for a 90 Ah cell with M8 bolt. Lets call it 60 uR to 140 uR. So I'm thinking you'd have 4 of those and two of your 15 uR bulk resistances. That's 270 uR to 590 uR for Rc. That's 6% to 15% of 4 mR.

I'm not sure about that as I note the Sky Energy/CALB straps are quite a bit larger than the TS ones for the same cells.
Larger straps will have little to no effect on the contact resistance, which is limited by the area of the top of the cell terminal, and in the strap-to-strap case it is limited by the area compressed under the bolt. Contact conductivity falls off with some high power of the pressure, as you get further from the bolt.
Also, the pressure from a 6mm bolt would not be too different from an 8mm bolt, depends on the tightening torque and that's limited by the Al and Cu cell posts.

The pressure will be much the same, but the area will be different by approximately the square of the bolt size = 6x6/8x8 = 36/64 ~= 0.5.
Also, I don't have four, there are only two per cell.
Two links per cell, each link has two contacts. That makes 4 contacts.

Now that you have redefined Rc to be for a single link I expect it to be between
60+15+60 = 135 uR
and
140+15+140 = 295 uR

[Edit: Replaced "Contact resistivity falls off" with "Contact conductivity falls off"]
Last edited by weber on Sun, 20 Nov 2011, 08:28, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

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

Low cost BMS

Post by Nevilleh » Sun, 20 Nov 2011, 20:38

All the above noted. I re-wrote my equation to give i2 = i1(Ri/(Ri+Rc)) which shows the centre cell is less than the other two.
Your comment about the area: The hole is bigger, so the 8mm bolt has less area! The contact area is really the cell electrode dia less the hole size and I use a big,flat, brass washer under my bolt heads, so it really is a bit difficult to have much idea of what the contact resistance is. Never mind, back to bms'ing.
Drove my car 16 kms today and the Vh is now 3.31V and Vl is 3.28V. I did note that one cell (#9) dropped to 2.88V under full acceleration and the highest stayed at just on 3.1. These were different numbers to those with no load.
My control unit does glitch up now and again and it doesn't seem related to motor load as it has done it just coasting. I haven't put the UART reset code in yet, but it gives me a few things to think about.
Weber, have you got your bms working OK yet and if so, what did you do? (Clues, clues)

User avatar
woody
Senior Member
Posts: 1715
Joined: Sat, 21 Jun 2008, 02:03
Real Name: Anthony Wood
Location: Mt Colah

Low cost BMS

Post by woody » Sun, 20 Nov 2011, 21:41

parallelling cells is always bad
maybe true, but aren't the TS / CALB cells just parallelled internally anyway? Same as LiPo?

I've seen a cylindrical cell setup (not headway, the premium mob) which links all cells in series and the parallel links are 150amp fuses (solid links which taper to a few sq. mm in the middle)

This could be the best alternative? Having 3 series links between the cells instead of 1...
Planned EV: '63 Cortina using AC and LiFePO4 Battery Pack

User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Sun, 20 Nov 2011, 21:56

Nevilleh wrote:Your comment about the area: The hole is bigger, so the 8mm bolt has less area!
I'm guessing that's a joke, but I'm not seeing an emoticon, so just in case ...

bolt head area minus hole area
or
washer area
still vary approximately as the square of the thread diameter.
The contact area is really the cell electrode dia less the hole size
I assume you mean pi/4 times (the square of the cell terminal dia less the square of the hole size). Yes. That's what I meant by "the area of the top of the cell terminal".
Edit: I have a further thought in that I have some TS40 cells sitting on the floor here and I might connect 3 of them as a triple with the same straps I use in the car, stick on a 130A load (which is what I used for testing individual cells) and see if I can measure the voltage drops across the links. I'm sure I have a multimeter here somewhere with a mV range on it. Sort of like what TJ did.
That would be great! But it is difficult to make this measurement on an actual cell. I suggest filing some of the anodising off the outside of the big aluminium terminal nuts to allow for multimeter contact.
I was also a bit surprised that the charge was just over 120 AHrs as I expected a bit more.
I expect that's just because it's not yet as well balanced as it could be. I expect the overall capacity to increase after the next bottom-balance.
My control unit does glitch up now and again and it doesn't seem related to motor load as it has done it just coasting. I haven't put the UART reset code in yet, but it gives me a few things to think about.
Weber, have you got your bms working OK yet and if so, what did you do? (Clues, clues)

Sorry Neville. We still haven't put the contactors back in. We pulled them out to put noise suppression caps in their economiser circuits and decided we'd better actually attach them to the car with something other than their cables Image when we put them back this time. So we are still working on contactor box mountings, in some very tight spaces.

My biggest suspicion with ours is that we need to increase the drive current to the optocoupler LEDs and reduce the pullup resistor on the optocoupler phototransistors by about a factor of 10. And add a filter capacitor across the "UART" input giving -3 dB at about the third harmonic of the fastest comms square wave.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Sun, 20 Nov 2011, 23:05

woody wrote:
parallelling cells is always bad
maybe true, but aren't the TS / CALB cells just parallelled internally anyway? Same as LiPo?
Yes, and I wouldn't be surprised if that led to some premature failures. Cylindricals have an advantage there. But the green and purple HED200's from NilCO2 http://www.nilco2.com.au/products.html are interesting. They are prismatics with terminals on opposite ends, like a cylindrical.
I've seen a cylindrical cell setup (not headway, the premium mob) which links all cells in series and the parallel links are 150amp fuses (solid links which taper to a few sq. mm in the middle)

They don't sound like fuses. They sound like arc lamps. Image
This could be the best alternative? Having 3 series links between the cells instead of 1...
Yes. And once you do that, why do you need the parallelling links, except at the ends of the series strings?

Sure it lets you save on BMUs, but then you are left in doubt about the state of the cells the BMU isn't directly connected to.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

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

Low cost BMS

Post by Nevilleh » Mon, 21 Nov 2011, 12:16

Just a further thought on measuring the link + contact resistance:
A single cell with a link bolted to each terminal, the othe ends of the links bolted to the same cable I use for the car wiring and so connected to a load of 130A (3C for my 40 AHr cell). I can measure the current and the voltage drop across the link reasonably easily and accurately, hence the link resistance.
And according to my theory (the one missing the emoticon, but I tend to use ! instead) the smaller I make the bolts, the better it should be!!

I am now wondering if my bms glitches are coming from the 12V supply, so I am going to add some BIG filters to the input and see if it makes any difference. It only glitches occasionally, so its not a huge problem, ie the thing is still very usable and giving good data, but I want to know why it is doing it at all.

Weber, what Baud rate are you using with your bms? Also, do you have the electrode terminal dimensions for different size cells? I see they don't give them in the specs and I have only 40 AHr ones.
Last edited by Nevilleh on Mon, 21 Nov 2011, 01:28, edited 1 time in total.

User avatar
coulomb
Site Admin
Posts: 3779
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Low cost BMS

Post by coulomb » Mon, 21 Nov 2011, 22:30

Nevilleh wrote: Weber, what Baud rate are you using with your bms?

I can answer that: we're using 9600 bps. Any slower, and it would take a very long time to propagate messages throughout our looong string of 228 cells. [ Edit: as Weber points out next post, we have two "parallel" (in comms terms) strings of 114 cells. So our string length is "only" 114 cells Image ] Our packet sizes are a lot larger than yours, too, by the sounds of it. For example, a voltage response might be
\012: 3324 V
which is about 14 characters with a newline and CRC byte. The slosh is to make the line a comment as far as other cells are concerned. The 012 is a 3-digit cell ID; 3324 is the voltage (in millivolts), V is the unit, and there are spaces for readability (I guess we could lose those spaces). I suspect your equivalent is about 2 bytes.
Last edited by coulomb on Tue, 22 Nov 2011, 09:06, edited 1 time in total.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

Post Reply