4Springs' Brumby

Post up a thread for your EV. Progress pics, description and assorted alliteration
User avatar
4Springs
Site Admin
Posts: 638
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

IMG_20200301_093516174.jpg
IMG_20200301_093516174.jpg (122.94 KiB) Viewed 923 times
This photo shows the six rear battery modules in place. The blue string is to carry them with, and then tuck away once in place. All of the connections are at the tops of the modules. I've left the threaded rods on the shorter modules quite long still, with the idea that they will go all the way to the top of the box. This is to help prevent them rattling around if I go over a bump.
IMG_0406.JPG
IMG_0406.JPG (112.67 KiB) Viewed 923 times
This is the module sitting snugly in the front left battery box. This one has 20 cells. There is blue string, then also blue and orange wires for BMS communications. Then there is the 30 A fuse, and a laptop power supply. The PS is for battery heating, powering the Nichrome wire in between the cells. The PS is fed from 240 VAC, which is turned on or off by a temperature controller. The temperature sensor for the controller is sitting right next to the fuse, which was a bad place to put it. I have since opened this box again to move the sensor to a less 'warmed' position.

User avatar
4Springs
Site Admin
Posts: 638
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

IMG_20200321_155657414.jpg
IMG_20200321_155657414.jpg (132.55 KiB) Viewed 923 times
Photo of the new-look underbonnet. The battery boxes are longer, thinner and shorter than before. The controller has been moved from the end to the top of the right battery box (left in the photo). The grey box with the red isolation switch has also been moved from the end to the top, as has the new BMS Master (black box with green display).
I have a few covers I need to make up. These boxes now have the connections on the outside, meaning that they need to be covered before I let anybody near them.
IMG_20200321_155720511.jpg
IMG_20200321_155720511.jpg (128.23 KiB) Viewed 923 times
Here are the wired-up rear battery modules. I'm having some trouble trying to fit in one laptop power supply per module, which are for the battery warmers. Six of them make (ironically) a lot of heat! So I've sandwiched them between aluminium 'fins' to try to keep them cool. Here's a photo of them from the top:
IMG_20200321_155636135.jpg
IMG_20200321_155636135.jpg (153.38 KiB) Viewed 923 times
This arrangement got to about 50°C on a 20° day. They won't be used on a 20° day, but we'll see how we go.

User avatar
4Springs
Site Admin
Posts: 638
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

So - how'd it go?
As far as I know, nobody has tested these cells yet in a full-blown vehicle. And certainly nobody has tested my method for connecting them together into modules. My biggest fear is that the clamping arrangement that I developed won't take the current.
IMG_0342.JPG
IMG_0342.JPG (135.18 KiB) Viewed 920 times
The top and bottom should be ok, they use Francisco's spring system, which he has tested to 400 A continuous. The Brumby draws up to 500 A, but only for brief moments. More like 100 - 200 A most of the time.
The bit that I'm worried about is the bit in the middle. This clamps two brackets together over four battery tabs. The brackets are flat, and they bend a bit under the tension of the bolts, so they might not be making good contact along the whole surface of the tabs.

So, once finally all wired up, I was able to gradually switch things on. The brake booster works. The DC-DC converters work. The contactor works. The controller works! - the motor turns while in neutral.
The Brumby has been up on ramps for the past couple of months while I've been working on it. So let's take it down at last. I push in the clutch and spin the motor. Put it into reverse and carefully release the clutch - it moves backwards! So I got the polarity correct, well done. I get it down off the ramps and park it just outside the garage. Then I'm able to remove the ramps, vacuum the carpet and move all the accumulated junk. Then I put it into 3rd to put it back into the garage. But it doesn't move! Hang on, what have I done? Did I have it turned on correctly? Yes. Let's try 2nd gear - rocks a bit but goes nowhere. Ok, I'll try revving up the motor. This gives me some sluggish speed, but nothing like the roar I'm used to spinning up at car shows.
Some friends come over to do some AEVA stuff, and I tell them my woes and get them to push the Brumby back into the garage again for me. I have to then go with them to do what we're doing, and can't get onto it straight away. I have time to think though; are my cell connections really that bad? I guess I'll check for voltage sag at the controller - if the voltage drops when I rev it up then I can go back and check battery box by battery box to see where the problem is. Hopefully it is a connection between boxes. Or could there be a problem with the controller? Surely not, I haven't changed that.

To be continued...

antiscab
Senior Member
Posts: 2622
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Re: 4Springs' Brumby

Post by antiscab »

are you trying to drive on the pre-charge resistor only?
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
coulomb
Site Admin
Posts: 4098
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: 4Springs' Brumby

Post by coulomb »

antiscab wrote:
Tue, 24 Mar 2020, 20:40
are you trying to drive on the pre-charge resistor only?
That has a familiar ring to it. I seem to recall exploding a pre-charge resistor on a certain MX-5.
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.

User avatar
4Springs
Site Admin
Posts: 638
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

antiscab wrote:
Tue, 24 Mar 2020, 20:40
are you trying to drive on the pre-charge resistor only?
Ah, I know what it is like to drive on the pre-charge resistor. This happens every time I forget to click around to 'Start' on the key. It goes 'eh', and just moves enough to tell me that it tried. This was giving me more than that, but not nearly enough.

I measured the voltage to the controller while running the motor with my hand using the throttle box. It was nice and steady at 145 V (give or take). So I measured the voltage to the motor, and it was not much. I forget, but perhaps 10 V. This then seemed like the motor was doing what the controller was telling it to do, but the controller wasn't telling it to do much.
At this stage I thought about the throttle box. The Kelly controller takes a resistive input, but I found resistive throttle boxes very unreliable. So I have a Hall effect box, and use a Zeva Digital Pot Adapter to convert the signal. All of these connections were re-wired during the battery upgrade, since the controller was moved. I seem to remember that the resistive output from the Zeva unit was polarised. I looked it up, and sure enough it is. No particular way to tell which way around it should go, the instructions say to try it one way, and if it doesn't work, try it the other way. So I swapped the connections around and it worked!

So, now to test the car a bit more.
First I drove around the house. Then took out my non-contact thermometer and looked over all the connections that I can 'see' with it. Checked the temperatures and voltages reported via the BMS. All good.
Then I drove down the road a bit, turned around and came back. Check temperatures again.
Next was down to the end of the road and back - about 5 km. One connection seems a little bit warmer than the rest. It is one on top of the modules though, a type of connection that I've used for years. The ones I'm not confident about, the ones on the ends of the modules, these are nice and cool.
Do the same trip again, but with more oomph. No, I don't think I'm getting the power that I'm used to. It draws about 300 A maximum. I'm not happy with that, I should be able to get 500 A plus. Are my modules wimps after all?
That connection is still warm. About 35°C on a 20° day. Other connections of a similar sort are about 25°C. Is that some heatshrink caught in the connection? I'll get out the spanners to remove it. Whoops! Not even done up - just finger tight. It's a wonder it gave me 300 A. So I take it apart, remove that heatshrink and do it up nice and tight.
Now we're talking, I'm getting just over 500 A, and much more go than before.

User avatar
4Springs
Site Admin
Posts: 638
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

I'm not completely convinced that the Brumby has the full power that it used to. Neither is my wife, Zara, who took it into Launceston the other day. Mind you, we haven't driven it for a couple of months, and we've been driving our new Kona in the meantime. So perhaps we are just spoiled!

One potential culprit is the voltage. I've gone from 48 cells of LFP to 39 cells of NMC. The NMC has a wider voltage range. Here's the difference:
LFP: 173 - 149 V
NMC: 162 - 125 V
I didn't ever notice a difference in performance between fully charged and almost empty with the LFP. It used more current though when SOC was low. Since my 'SOC' meter is purely counting current, the lower half of the fuel gauge always went faster than the top half!

Another might be a dodgy connection somewhere. I'm pretty happy with the connections that I can poke a non-contact thermometer at. There aren't all that many of those though, so let's look at the BMS data for clues.

I can transmit data from the BMS to a computer. It is currently spitting out updates about twice a second. Here is my drive to work this morning:
Volts.png
Volts.png (203.54 KiB) Viewed 897 times
This graph plots the voltages of all 39 cells. I've driven about 15 km, over about 15 minutes. The x axis is just the number of readings, which are reported at a constant rate, so think of it as 'time'. The y axis is cell voltage. I've gone about 90 km on this charge so far. The resting cell voltages range from 3.60 to 3.65, which should equate to about 40% state of charge. They need a bit more balancing.
What I can see from this data (ok, maybe not from this graph, but when I zoom in on a section), is that the cell voltages follow each other up and down quite nicely. I was looking for one that was sagging more than the others under load, but I don't see any. This means that all my cells are about as good as each other, and the connections between the cells are good. If there was a connection that was a bit higher resistance then I'd expect to see the voltage sag on that one under load.

What the voltages won't show me though is a high resistance connection between modules. So let's look at temperatures for other clues:
Temperatures.png
Temperatures.png (232.85 KiB) Viewed 897 times
This is the same drive. The BMS reports temperature when I press a button on the dash. So there are only a few readings. The celltop board temperatures are not calibrated, and there is a bit of variation. So what I'm looking for here is not necessarily the highest temperature, but the one that increases the most under load.
All cells warm up during the drive, for example cell #38 goes from 10 to 15°C. But looking for the largest change, cell #35 is going from 15 up to 24°C. Cell #39 is going from 12 to 20°C. Cell #10 goes from 14 to 22°C.
So it may pay me to open up the battery boxes and look at the connections on or near those celltop boards.

francisco.shi
Groupie
Posts: 350
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: 4Springs' Brumby

Post by francisco.shi »

How did you attached the temperature sensors?
When I did some testing for the first time I found that if the contact overheats the tab conducts the heat to the cell quite efficiently.

User avatar
4Springs
Site Admin
Posts: 638
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

francisco.shi wrote:
Wed, 25 Mar 2020, 19:11
How did you attached the temperature sensors?
When I did some testing for the first time I found that if the contact overheats the tab conducts the heat to the cell quite efficiently.
The temperature sensors are on the celltop circuit boards - surface mount component very close to the negative terminal. So the temperature is of the tabs (and connectors) rather than the cells. One of the links that you sent me a while ago suggested that the best way to cool the cells is via the tabs, which fits well with your experiment (although in the other direction).

francisco.shi
Groupie
Posts: 350
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: 4Springs' Brumby

Post by francisco.shi »

This is what i did on my board but i found that when the bypass is running it warms up the board and that causes the sensor to detect a higher temperature. I placed the sensor very close to the contact and far away from. The hot parts of the bypass circuitry but it still detects a few degrees higher than the temperature of the cells. The heat seems to be conducted thru the board. About half of the board is used as heatsink for the power transistor and even it only gets slightly warm it does affect the temperature reading by at least 5°C or more. So I am considering to put the sensor on the side of the cell and run some wires to it. It is not really what I wanted but I seems like the only choice.
Also given the low thermal resistance of the whole pack I am wondering if I can just have only a few sensors instead of a sensor for every cell.

User avatar
4Springs
Site Admin
Posts: 638
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

francisco.shi wrote:
Fri, 27 Mar 2020, 04:51
This is what i did on my board but i found that when the bypass is running it warms up the board and that causes the sensor to detect a higher temperature. I placed the sensor very close to the contact and far away from. The hot parts of the bypass circuitry but it still detects a few degrees higher than the temperature of the cells. The heat seems to be conducted thru the board. About half of the board is used as heatsink for the power transistor and even it only gets slightly warm it does affect the temperature reading by at least 5°C or more. So I am considering to put the sensor on the side of the cell and run some wires to it. It is not really what I wanted but I seems like the only choice.
Also given the low thermal resistance of the whole pack I am wondering if I can just have only a few sensors instead of a sensor for every cell.
Here is one of the Low Cost BMS boards:
IMG_0409.JPG
IMG_0409.JPG (48.33 KiB) Viewed 849 times
The temperature sensor is TH1, on the right hand side. It's about 50 mm from the balancing resistors. My testing found no temperature increase when balancing. My balancing resistors were up to about 60°C according to my non-contact thermometer.

There may be a software fix for your temperature problem.
If you are slow charging from AC, you probably don't care about temperature while charging. And you don't care while sitting still. If you are like me, you only really care about temperature when driving the car. You want to be alerted if a cell gets too hot while discharging at high currents.
You could change your algorithm so that balancing does not occur during driving. This means your boards do not self-heat when you want the most accurate temperature readings.

francisco.shi
Groupie
Posts: 350
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: 4Springs' Brumby

Post by francisco.shi »

I did exactly the same as you but I use the board as a heatsink for the power transistor. So I will have to relocate the sensor.
Top side of cell top board.
Top side of cell top board.
15853659623345810300733166448637.jpg (1.15 MiB) Viewed 835 times
I even called the sensor TH1 :lol:
Bottom of cell top board.
Bottom of cell top board.
15853660539928946359089787843007.jpg (1.08 MiB) Viewed 835 times
My plan is to run the balancing all the time. If the cells track each other it will not run for very long.

User avatar
4Springs
Site Admin
Posts: 638
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

I discharged the battery down to a minimum cell voltage of 3.1 V the other day. Took me 160 km to do it.
Then I charged up to a maximum cell voltage of 4.1 V.
The next day I found the BMS reporting "Fault 1". This means that it wasn't hearing anything from the celltop boards. The first cell up the line is inside a box that's annoying to open, so I first concentrated on the connectors and wiring. No go though, so the box had to come open.
Inside I found that the celltop module was not talking, even when plugged directly in - so not a wiring fault. Voltage of the cell is ok (phew!), but the circuit board is hot, indicating that it is shunting.
I figure this is a hardware fault on the circuit board. So today I took the battery module out and inspected it on a bench. There's some weird goo on the circuit board, oh, and it's coming out of the top cell. Not good!
IMG_0415.JPG
IMG_0415.JPG (76.74 KiB) Viewed 816 times
IMG_0421.JPG
IMG_0421.JPG (66.22 KiB) Viewed 816 times
The goo has leaked over the entire connector, coating the tabs of the top 10 cells to varying extents. It doesn't look like it has reacted with the aluminium or copper, but that steel spring is showing a bit of corrosion, and the circuit board (bolted on top of that spring) obviously suffered a bit...
The goo must have been fairly liquid to start with, to travel so far, but it has evaporated to moist crystals in certain places. Anyone know what would be best to clean it up with? It soaks into a cloth ok, so I can get rid of most of it mechanically, but I don't want to wipe it down with metho if it's going to react with it.

Not sure what has caused this. Could be that I tightened that connector a bit hard, and it pulled the tab too much? Or perhaps just a dodgy cell, that took a good charge, discharge and charge again to show up as dodgy.

I have three cells that Francisco replaced for me. I'm thinking I might risk trying one of them out. Only one of them looks really bad, and to be honest there is one of them that I can't see a problem with now I look at it in the fading afternoon light.
I think I might open the other battery boxes and check over the other cells before I go much further...

User avatar
jonescg
Senior Member
Posts: 3444
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Re: 4Springs' Brumby

Post by jonescg »

Christopher - that electrolyte contains LiPF6 which is not corrosive itself, but reacts with moisture in the air to form hydrogen fluoride, LiOH and PF5.
These most definitely are corrosive towards both iron and aluminium.

I'd say you've pulled on the tab a bit too tight :(

I had no idea this clamping system used steel - at the very least use nickel plated steel, or just plain old aluminium if you have enough Alminox to throw around.

I've always treated pouch cells as unpeeled bananas - very fragile and prone to bruising. Also electrically conductive and energetic bananas...
AEVA National Secretary, WA branch chair.

User avatar
brendon_m
Senior Member
Posts: 578
Joined: Sat, 28 Oct 2017, 11:00
Real Name: Brendon McCarrol

Re: 4Springs' Brumby

Post by brendon_m »

I was curious so I looked it up, turns out a banana is good for about 0.8V so not quite as "electrically conductive and energetic" as lithium. :D
images.jpeg
images.jpeg (27.67 KiB) Viewed 808 times

User avatar
4Springs
Site Admin
Posts: 638
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

jonescg wrote:
Wed, 01 Apr 2020, 19:07
I had no idea this clamping system used steel - at the very least use nickel plated steel, or just plain old aluminium if you have enough Alminox to throw around.
Hmm...
Would you like to elaborate?
I've looked up Alminox, and it appears to be a compound to help aluminium surfaces make good electrical contact with each other. I haven't used that (didn't know it existed), I have aluminium and copper tabs clamped between extruded aluminium channel.
The springs on top are presumably mild steel. These are not part of the electrical path, they are there just to clamp the tabs together. They are bolted together with stainless steel bolts, and I may have used some bright steel washers.
Are you thinking that there might be problems where the different metals come together?

User avatar
jonescg
Senior Member
Posts: 3444
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Re: 4Springs' Brumby

Post by jonescg »

Yeah, aluminium and anything other than aluminium will set up a galvanic reaction, and iron is a good candidate as it's further down the reactivity table.
Alminox is good for any aluminium + other metal corrosion retardation. Not an inhibitor, but a slow-it-downer. Nickel and copper are well down the table, so they don't tend to react with much.
AEVA National Secretary, WA branch chair.

User avatar
4Springs
Site Admin
Posts: 638
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

I did some banana cleaning today.
The aluminium tabs are somewhat corroded - they will need to be polished up. The copper tabs have lost some of their (nickel?) coating as well.
I found another cell that suffered some damage. This is the one underneath the leaky one. Several spots around the anode. I didn't notice this yesterday.
IMG_0430.JPG
IMG_0430.JPG (84.34 KiB) Viewed 785 times
IMG_0429.JPG
IMG_0429.JPG (80.66 KiB) Viewed 785 times
So it looks like I have two cells to replace from this module. And I need to go through the other modules to see how they look.
Any ideas how to dispose of NMC cells that are fully charged? Discharge them first perhaps, or would that be dangerous in itself?

User avatar
4Springs
Site Admin
Posts: 638
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

jonescg wrote:
Thu, 02 Apr 2020, 18:04
Yeah, aluminium and anything other than aluminium will set up a galvanic reaction, and iron is a good candidate as it's further down the reactivity table.
Doesn't a galvanic reaction require a liquid? Electrolyte leaks aside, these connections should always be warm and dry.

User avatar
jonescg
Senior Member
Posts: 3444
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Re: 4Springs' Brumby

Post by jonescg »

It does need a solvent to solvate the ions, yes. But even humidity in the air will do that. If there's free liquid anywhere it's all over red rover, but even mild humidity will find a way.
AEVA National Secretary, WA branch chair.

rhills
Site Admin
Posts: 455
Joined: Fri, 25 Jul 2008, 01:57
Real Name: Rob Hills
Location: Waikiki, WA

Re: 4Springs' Brumby

Post by rhills »

I thought galvanic corrosion also required movement of electrons.

Which makes things a bit tricky when you want some electrons to move, but not others :-(
Rob Hills
AEVA Webmaster
2014 Mitsubishi Outlander Aspire PHEV
Petrol Usage to last refill: Jul 2014 - Apr 2020
Total Petrol: 720.1L
ODO: 61386
Av Consumption: 1.17 L/100km

User avatar
4Springs
Site Admin
Posts: 638
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

What a week!
On Monday I didn't drive the Brumby to work because I had a failed BMS celltop board. Today I'm wondering whether I even have a battery at all...

I opened up the rear battery box today. Both to look at some of those warm cells identified by the graphs above, but also to look for more electrolyte leaks. Moving some cell modules around, I can't see (or smell) any leaks.
The temperature testing I did previously suggested that celltop boards 20 & 25 (among others) were getting warm more quickly than others. These boards are at opposite ends of a group of 5 cells. So I pulled this module out and dismantled it. Takes about 45 minutes to dismantle, took about 3 hours to put together in the first place...
In this pile of 30 cells I found two with signs of leakage.
IMG_0442.JPG
IMG_0442.JPG (53.84 KiB) Viewed 754 times
These were the second and third from the bottom. The boards 20 & 25 were at the top, but no sign of leakage there. These two were on top of each other , with those spots extremely close to each other. Next to this spot was this mark on the aluminium sheet:
IMG_0436.JPG
IMG_0436.JPG (45.52 KiB) Viewed 754 times
This looks like it got hot. Unless this mark could be made by a gas venting? I'm guessing it got hot though, and one cell getting hot has made the other get hot and leak. These leaks have not exited the (Mylar?) coating on the cells.

Next I opened up the one remaining battery box, the front left one. This contains a single, 20 cell battery module. This stunk with the now-familiar smell of NMC electrolyte. Liquid was pooling at the bottom of the wooden box. I found four of these cells damaged, two of them quite extensively. These definitely look like they had been hot. Black burnt-looking holes in the cells, although still under the Mylar, nothing had penetrated that. I didn't take a photo straight away, and by the time I did, the holes had clouded over, presumably with the reaction to air as Chris said above.
IMG_0448.JPG
IMG_0448.JPG (63.32 KiB) Viewed 754 times
Here is the aluminium sheet next to these cells (they went at both ends):
IMG_0432.JPG
IMG_0432.JPG (95.62 KiB) Viewed 754 times
Again, three of the damaged cells were together, suggesting that one had set off the other two.

User avatar
4Springs
Site Admin
Posts: 638
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

So what happened? Why have these cells gone bad?
  • With this fastening system the top and bottom cells in each pile of 10 have quite an angle on the tabs. The first failure I saw was one of these. But since then I've found others that were not the furthest ones, but ones in the middle where the tabs are pretty flat.
  • It could be that I've damaged cells while connecting them up - clamped one bolt up tight too early, then creating tension when tightening the next bolt.
  • On the worst of the modules, the front left one, one of the compressing nuts had given way. So one of the four threaded rods was completely loose, with the other three still tight. I don't know if this could have had such dire consequences in a reasonably gentle discharge and then charge cycle?
  • Perhaps these cells are just duds? I don't know of anyone else who has tested them to this extent. If this is the case, then perhaps this has weeded out the bad ones, and I only have good ones left?
It seems that I'm down 8 cells, with probably more duds hidden away in the other 5 battery modules.
My original plan was 200 cells, 40 series 5 parallel, adding up to about 47 kWh. I killed 5 of them, so I built 195 cells, 39 in series, 45 kWh.
If I have enough cells then I should be able to re-configure to 4 parallel, 42 series. This would be 168 cells, or 39 kWh. 4 in series would mean less angle on the cell tabs. It would also mean smaller, lighter modules. The module system should cope nicely with this change, I just need to run the nuts further down the threaded rod. It means completely dismantling all modules, re-polishing all the connections and then re-assembling again.
And then I don't know if the remaining cells will do any better!

francisco.shi
Groupie
Posts: 350
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: 4Springs' Brumby

Post by francisco.shi »

I have had a pack of these cells assembled for over a year and didn't have any leak and they still seem to be working fine. I can not explain why you are having such a high failure rate. Is there any chance you could have pulled the tabs while tightening the bolts and damaged the inside?

User avatar
4Springs
Site Admin
Posts: 638
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

francisco.shi wrote:
Sun, 05 Apr 2020, 19:50
I have had a pack of these cells assembled for over a year and didn't have any leak and they still seem to be working fine. I can not explain why you are having such a high failure rate. Is there any chance you could have pulled the tabs while tightening the bolts and damaged the inside?
Assembled, yes, but tested as a pack? Mine were assembled for some weeks before I put them into the car, and there were no signs of failures then. It wasn't until I drove the vehicle, drawing reasonable currents and taking the pack down until the lowest cell was 3.1 V. Then I charged up again fully to 4.1 V.

It is possible that I damaged some cells while tightening the connections. I was very careful, but perhaps not careful enough? There are two arguments against this.
1) I'd expect that the tabs on the outside of the connections would be under the most stress, since they are at the greatest angle. This was not the case, so far I've had failures on the outside ones and inside ones.
2) The first module I found that had failures was the last one I assembled. I would expect that I'd get better at it over time!
The first cells I found to be failed were among the final ones that were sent to me as replacements, so at first I suspected that something was different about them. But I've found others now...

Post Reply