barnard's Toyota MR2 - now Mr240

Post up a thread for your EV. Progress pics, description and assorted alliteration
User avatar
evric
Site Admin
Posts: 500
Joined: Sun, 20 Jul 2008, 01:57
Real Name: Eric
Location: Adelaide SA
Contact:

barnard's Toyota MR2 - now Mr240

Post by evric » Tue, 08 Jun 2010, 15:06

Re. Signage...Individual letters and numbers are available from Sprint Autoparts.
Prius Plug-in Conversion: http://www.evplus.com.au ...Holden Barina EV: http://www.evric.kestar.com.au

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

barnard's Toyota MR2 - now Mr240

Post by Johny » Tue, 08 Jun 2010, 15:27

barney wrote:Anyone able to suggest a source of some ways to add signage to the rear of the MR240v? I'd love something saying 'electric', or even All-electric, ....
These are the ones I'll be using http://www.stickyletters.com.au/designs.html

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

barnard's Toyota MR2 - now Mr240

Post by evric » Tue, 08 Jun 2010, 15:27

www.EVParts.com also have the "Electric" decal. Goto "Street Vehicle" on the top menu, then "Decals & Emblems" on the left menu. EVParts postage to Aus is $33.00US but Cloud Electric is about $76.00US. This is a lot of money for one decal! Can anyone do any better? I want to get some as well.
Prius Plug-in Conversion: http://www.evplus.com.au ...Holden Barina EV: http://www.evric.kestar.com.au

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

barnard's Toyota MR2 - now Mr240

Post by evric » Tue, 08 Jun 2010, 15:31

Johny wrote:
barney wrote:Anyone able to suggest a source of some ways to add signage to the rear of the MR240v? I'd love something saying 'electric', or even All-electric, ....
These are the ones I'll be using http://www.stickyletters.com.au/designs.html


Stickyletters certainly have a good range, but Sprint are about half the price. Unfortunately the Sprint website is useless.
Prius Plug-in Conversion: http://www.evplus.com.au ...Holden Barina EV: http://www.evric.kestar.com.au

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

barnard's Toyota MR2 - now Mr240

Post by Johny » Tue, 08 Jun 2010, 15:40

No Sprint in Victoria Image
I haven't actually looked at SuperCheapAuto for chrome letters so I've added that to my 'To Do' list.

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

barnard's Toyota MR2 - now Mr240

Post by evric » Tue, 08 Jun 2010, 16:01

Checkout the list of Emblems suppliers on the Parts or Links page on the AEVA Adelaide website.

http://www.aevasa.kestar.com.au/parts.htm

Elektroplate are offering Free Shipping and their prices are good.

Is EVWORKS thinking of adding decals to their list of products? There is a lot of room to spare on their "Signage and Labels" page.

If anyone finds any more suppliers, please let me know, and I'll add them to the list.
Prius Plug-in Conversion: http://www.evplus.com.au ...Holden Barina EV: http://www.evric.kestar.com.au

antiscab
Senior Member
Posts: 2489
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

barnard's Toyota MR2 - now Mr240

Post by antiscab » Wed, 09 Jun 2010, 00:14

there is a place around the corner from EVWorks that does decal.

I can't remember the name, but I'll make a note of it next time I come past.


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

barney
Groupie
Posts: 77
Joined: Thu, 09 Oct 2008, 07:29
Real Name: bdc
Location: Perth AUS
Contact:

barnard's Toyota MR2 - now Mr240

Post by barney » Sat, 14 Jul 2012, 08:08

I havent been a good correspondent recently - lurking and watching other developments, some of the great bespoke EVs and learning a bit more about batteries.... but not enough, sadly.

I am worried about my traction pack, and would appreciate some guidance. Please dont tell me its too small for the car... whether this is true or not, that is the fun of a one-off design. We were not a manufacturer who gets development time, so the MR2 became ZEV240 and now Id like to work with what Ive got. And I was hoping for a few thousand recharges, not a six hundred, before having to renew the pack.

In summary it is performing very poorly, after being pretty sprightly in its first two years. Also the range is dropping slightly too. I suspect only a few batteries, but I do not know how to read the data...

Here are some details about the traction pack, and Ill attach in a second post, some voltage tables and graphs as a result of logging the batteries (8 at a time) until all 45 were recorded. Data recorded late May. A word version of this report is available - just ask.

------- Battery Report on ZEV 240's traction pack, June 2012) ----------

Background
My EV is a converted MR2 – rego ZEV240 - a DC conversion now two years old, and still great fun to drive. There seems to be a problem with the traction battery (45 100Ah Lithium batteries c.3.3V ea, total c.150V). Here are the issues and some data from a data-logger (which does 8 batteries at a time) and graphs of the same data – six sets all told.

Issues
Not all batteries are _exactly_ the same, of course; and that is why ZEV240 has a BMS (battery management system), whose major purpose is to get them all up to full charge equally by the end of the charging cycle. The red BMS error light indicates there is an error condition, either constant or flashing under different circumstances. Since about January this year I have seen the light come on during normal driving. Didn’t happen before, that I noticed. Recently the red BMS light appears even if I accelerate only moderately. Watching the gauge, this is at approx. 135V (and 45 x 3 = 135). This suggests that the av. Voltage below 3V/cell is one of its error conditions.

Also since about Jan there is a gradual loss of power and I notice I cannot get away quickly if I need to, eg turning right across traffic, which one has to do occasionally. It may be exacerbated by winter chills, maybe below 5°-10°, but Im sure it started much earlier (ie in Australian summer, Dec-March).

Note that the pack still settles back to about 149V or 151V after overnight charging, so no battery seems to be performing seriously below average at the start. My feeling is that, over the last 3-6 months, the ability to accelerate has dropped significantly, but the 'mileage' has only dropped by a few percent. Table 1 below shows the actual data by month since January. I can imagine that if say 1 battery out of the 45 were very weak, (about 2% of the capacity) then that might account for it, since they all share in the load at times of peak power (and one is very weak, drawing the rest down), but each gives its best as they are all slowly discharged (hence the drop of only a few percent). That’s not exactly how it is, I know, but it leads to the big questions –

1 which battery or batteries, are weak? And the next question is:
2 How did it happen so quickly (within 2 years for batteries with a predicted life of 8+ years) and
3 what can be done?

Table 1. (old link stopped working-fixed)**
Image


The Test
Drove the same course 6 times, moving the data logger each time to a different set of the 6 groups of batteries. Took car on a circuit total time approx 4.30 min from top of driveway and back. Each test pushed the batteries four times, the first 3 times were pedal-to-the-floor acceleration to 60k/h. On the timeline these are about 1:15min, 2min and 2:45min. The fourth – less pressured – was back up the driveway at about 3:45m or 4:00m. Batteries were at ~70% at start, ~40% after the six tests, so not ideal testing. As the whole traction pack voltage drops slightly during use, the later V measures would be slightly lower than the early ones.


The Data
The data logger records the voltage accurate to 3dp roughly every second for each battery. Each test cycle is reproduced here in two forms: firstly a table of data with 8 columns and the time count down the left, but only the middle period c.1:15m to 2:45m or less; secondly a graph of the voltages with voltage up the left scale and full timeline along the bottom from zero to c.4:30m. The eight cells are shown as 8 different coloured lines. The lowest point is usually identified and its value listed, so it can be located in the corresponding data table.

Next post should have all six sets of data tables and graphs etc.


Last edited by barney on Tue, 14 Aug 2012, 22:29, edited 1 time in total.
Plan: Promoting EVS as pico-powerstations. My EV is handmade MR2>ZEV240 (150V 9" DC Warp motor, SkyEnergy batts)
Mr240-electric handmade car

barney
Groupie
Posts: 77
Joined: Thu, 09 Oct 2008, 07:29
Real Name: bdc
Location: Perth AUS
Contact:

barnard's Toyota MR2 - now Mr240

Post by barney » Sat, 14 Jul 2012, 08:38

Part 2 of Traction Pack report for ZEV240, now 2 years old.....
o o O O== In which the six sets of tables and graphs are presented, and feedback is sought on how to recover from apparent battery problems ==O O o o

(and the graphs and charts are becoming visible - apologies...)

Tables graphs and observations (sorry, seems the pics all disappeared; and now magic is bringing them back)**
First set A, UNDER the car...Set A: 1-8 (Pack at c.70% at start of test cycle).
               A1            A2           A3            A4           A5           A6           A7           A8
Image

Getting smarter... the Conditional colouring now has a minimum of 1800 (ie 1.8V) rather than using the min of the 8pack.... And here is the graph of that set.

Image

|| The question is: What can you tell from this graph? Which batteries should I replace, and why!
}} All graphs are similar. Eg, Is it the batteries that go lowest, or those that recover slowest (not the same).


Batts#9-16:
Image

Image


Set A 17-21
        17              18       19          20         21
Image

Image

I think Batt #s 22-29. Second set done last as I learned what to do…
(Pack capacity down to c.55% by now)

Image

Image


Next: 30-37:
               30            31            32           33           34            35             36            37
Image

Image


Then finally, 38-45…(Pack dropped to c.40% by this time)
               38            39            40           41            42          43             44            45

Image

Image


Summary:
If dropping below 2V and recovering slowly is a measure, then batteries #5, #17, #36, #45 are issues. (Table 2)

If dropping below 2.5V at any time is a measure, 10 batteries are identified (Table 3):

If final ‘rest’ voltage (ie after a minute or two’s recovery) is a measure, then investigate #4, 6, 14, 21, 25, whose final ‘resting’ voltage low. They dropped much more than all others, lower than many identified as weak in Table 2.

Which of these identifies the truly weak batteries that need to be replaced?

Image


Image



What about pack voltage o/night settling to 149V? or 151v? Is this good? Sure it is, the pack only started at about 152V in the start…And 3.3V x 45 cells ~ 150V. Ok, relax.

The Criteria for judgement
Dont know, of course; that’s why Im sharing this data. But looking at the graphs it seems that some batteries recover within minutes to within 0.01% of the average voltage, while others take a long time to do so. One obvious distinction, as I created and looked at these graphs, led to building Table 2 above… In this case ‘the weakest’ were those that…
a) fall below 2.0V under load, and
b) do not recover to quite the same level as or better than others within minutes (altho this is only millivolts difference, apparently, but it might be significant).

The third table shows all whose voltage fell below 2.5, regardless of their recovery (or the state of the battery pack, which dropped from 70% to 40% during the test process).

The Cause
There are many explanations possible, and I list a few that occur to me. Of course which explanation is better could arguably depend a little on how many batteries are playing up.
**Is it age? The batteries are meant to be good for some thousands of recharges. Ive seen figures of 2000 and 3000 in writing. This assumes the average DoD is kept above 80%, which I almost always do. Once it got to 19% as we got home. The table 1 estimates are for 100% DoD simply to provide comparisons with ICE vehicles, both petrol and diesel, which shouldn’t be drained either.
**Is it over-stressed batteries? The pack is not large, and the ammeter can (or used to) show a figure of over 300A momentary draw, if you accelerated hard. This has taken its toll and they are giving up already? This seems unreasonable since I don’t always drive hard, rather I am always keen to get the best mileage I can, as shown by the fact that I have recorded monthly data recordings since I started, and they are all about minimising W/km and cents /km.
**Is it corrosion? I was shown how to clean all battery contact surfaces before installation, and tried to clean them very thoroughly, using cleaning agent and finally a soft rubber, before we attached the BMS module onto each battery. Perhaps some of these connections have oxidized enough to reduce apparent battery performance of a few?

Solutions/Ideas that occur to me
Limit the controller to a max of Amp draw, to reduce performance but also reduce strain on the battery pack? (Reduce future strain on all batts)
Move any battery that is problematic to the (easily accessible) top pack and retest? (simplifies future testing/replacement; and addresses Corrosion)
Swap some of the critically weak batteries with new ones and retest? (Go straight to the problem - may fix it)


Your comments are welcomed! Please help me fix my traction pack! Barney
Last edited by barney on Wed, 15 Aug 2012, 10:20, edited 1 time in total.
Plan: Promoting EVS as pico-powerstations. My EV is handmade MR2>ZEV240 (150V 9" DC Warp motor, SkyEnergy batts)
Mr240-electric handmade car

User avatar
coulomb
Site Admin
Posts: 3780
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

barnard's Toyota MR2 - now Mr240

Post by coulomb » Sat, 14 Jul 2012, 15:23

barney wrote: **Is it corrosion? I was shown how to clean all battery contact surfaces before installation, and tried to clean them very thoroughly, using cleaning agent and finally a soft rubber, before we attached the BMS module onto each battery. Perhaps some of these connections have oxidized enough to reduce apparent battery performance of a few?

I think that's the first thing to check. It could be as simple as some of the cell terminal bolts have come slightly loose with vibration. So you could either just torque up all the connections to the specified torque, or you could run a non-contract thermometer over the connections after a drive, and see if there are any terminals that seem hotter than others. (But they could be hotter because the batteries are high resistance, and therefore getting hotter internally). [ Edit: this could be used to choose one or two terminal connections to take apart and inspect thoroughly for corrosion. Or of course, you could use your data to choose these one or two, then check if the data improves. If it does, clean up at least the ones that the data suggests are a problem. ]

While your pack is a little on the light side, if your maximum, not-often-seen peak current is 300 A from the pack, then (assuming you have 90 Ah cells), that's 3.3C, which is not outrageous at all. Sky Energy batteries were supposed to be 4C peak at about the time you bought yours. So I'll be disappointed if that turns out to be the cause. (Not half as disappointed as you, I imagine Image .)

[ Edit: bolts -> cell terminal bolts ]
Last edited by coulomb on Sat, 14 Jul 2012, 05:27, 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.

antiscab
Senior Member
Posts: 2489
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

barnard's Toyota MR2 - now Mr240

Post by antiscab » Sun, 15 Jul 2012, 06:51

what is the continuous discharge rates like Barney?

do you spend much time on the freeway at 100kmh?
how much current do you draw at that speed?

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

User avatar
weber
Site Admin
Posts: 2626
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

barnard's Toyota MR2 - now Mr240

Post by weber » Sun, 15 Jul 2012, 15:45

Hi Barney,

I agree with Coulomb.

How are your cells clamped together? Can you post photos?

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

User avatar
jonescg
Senior Member
Posts: 3036
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

barnard's Toyota MR2 - now Mr240

Post by jonescg » Sun, 15 Jul 2012, 17:51

Hey Barn,

If you think you have a few dud cells, just replace them with new ones. While mixing old with new is not recommended, spending another $10k on a new pack is hardly an option either. In fact, if the newly replaced cells start to cause trouble in another 600 cycles, you have at least got your 1200 cycles albeit costing a couple hundred dollars more than planned.

But definitely go through and check all the terminations and make sure there's no corrosion and that the bolts are torqued down properly. While the pack is small, it's not exactly undersized. Hope you enjoyed the 'Revenge' movie!
AEVA National Secretary, WA branch chair.

zeva
Senior Member
Posts: 422
Joined: Sat, 15 Dec 2007, 05:09
Real Name: Ian Hooper
Location: Australia
MSN: sigmunky@hotmail.com
Contact:

barnard's Toyota MR2 - now Mr240

Post by zeva » Sun, 15 Jul 2012, 23:38

My initial thoughts:

- I agree about the possibility of corrosion on battery terminals reducing performance. Terminals should always be cleaned before installation, but corrosion (Al2O3 in particular) reforms over time if you don't have Noalox (or similar) keeping moisture & oxygen out. I know people who have noticed a *big* difference after re-cleaning terminals (and adding Noalox) after even ~6 months. You can Noalox from EV Power, BTW.

- As you noted, cold weather can make a big difference. The ~10% change you've noticed since February could be largely due to that.

- It sounds like your cells have seen discharges down to 0% SoC (say <3V/cell). This can definitely reduce cycle life dramatically.. Discharging to 100% DoD will halve or worse your cycle life compared with say 70-80% DoD.
Ian Hooper
--
"Never doubt that the work of a small group of thoughtful, committed citizens can change the world." - Margaret Mead
http://www.zeva.com.au

barney
Groupie
Posts: 77
Joined: Thu, 09 Oct 2008, 07:29
Real Name: bdc
Location: Perth AUS
Contact:

barnard's Toyota MR2 - now Mr240

Post by barney » Wed, 15 Aug 2012, 08:48

Thanks to everyone who contributed and all the feedback. You asked lots of questions so Ive been collecting the answers, and now trying to answer them, restore broken links (they did work once, I think?) and get this sorted so I can enjoy driving again. And learn a bit on the way. Like changing the max current draw on the Zilla controller so say 300A max, or 3C, or some voltage drop ceiling... (min 135V?)
Dont actually have any idea of what is reasonable yet; this AEVA website is perhaps my saviour here....)
Plan: Promoting EVS as pico-powerstations. My EV is handmade MR2>ZEV240 (150V 9" DC Warp motor, SkyEnergy batts)
Mr240-electric handmade car

barney
Groupie
Posts: 77
Joined: Thu, 09 Oct 2008, 07:29
Real Name: bdc
Location: Perth AUS
Contact:

barnard's Toyota MR2 - now Mr240

Post by barney » Thu, 16 Aug 2012, 08:22

ZEV 240's traction pack... feedback round 1:

Thanks to everyone who made suggestions and comments. I'll try and answer them briefly here.
Corrosion was the first thing Coulomb and Zeva commented on, and of course I hope they are right and it is not an underlying battery weakness.... I assume that the batteries with the lowest drops in the log charts are good initial candidates to check, although Coulomb suggested a non-contact thermometer (are these easy to borrow, btw?) to measure all terminal temps after a drive looking for hot connections. At this stage the motor bay pack is easier to get to than the lower petrol bay tank!

One approach suggested to me is to move all the apparently dicky batteries from the lower pack, which is hard work to remove, up to the motor bay pack and then it will be easy to measure test & replace. Sadly this means only older ones are left below yet you'd want the newest ones there to minimise access issues later!

Rather than moving batteries around to do more testing it would be great to be able to identify all the dicky batteries based on the current information alone; and then do the battery swapping AND install the new batteries below, all in one pass.

Next question is the condition of the batteries themselves, addressed by Coulomb, Antiscab, weber, jonescg and zeva. The fact that the BMS battery light comes on whenever I start putting my foot down, is a concern. As Rod Dilkes (BMS designer) said with mastery of the understatement, 'a battery is very weak'. I hope that he too is right, and its only one! I have always faithfully kept the battery DoD to less than 80%, and have records to prove this, as Ive been keeping a monthly spreadsheet since the beginning. Once it got to 19% remaining, otherwise it has always been above this. But the Zilla controller has a maximum amp draw set of around 500A, iirc. It may be lower; Id love to build a serial cable so I could examine the Zilla monster's settings myself, but couldn't find the wiring specs to match to a Psion 5 palmtop I still have lying around. The first thing I should do would be set it to max out at 300 or 400A, I can see.

Antiscab: Continuous draw when on the freeway, where I do max c.90kph - perhaps half of all journeys - seemed to be around 75A to 115A on level parts.. perhaps it averages at about 90A? By my calculations I average about 170W-190W/km between recharges, does this seem reasonable.?

Weber: pic of the battery clamping setup, in this case in the upper pack, now with recently added Junsi Cell-Log wires on each battery is below (to se a larger version, click the link below the pic):
Image

Larger pic

So, I like the suggestion that I replace the bad ones, putting up with the fact that they wont match the existing 100Ah Sky Energy ones. But which exactly ARE the bad ones! Who is ready to suggest which ones should be replaced, and anything else I should do, to maximise the remaining life of the rest of the traction pack?

Barney
[edit: Corrected spelling Junsi; added _sml and ref to full sized pic]
Last edited by barney on Wed, 15 Aug 2012, 22:35, edited 1 time in total.
Plan: Promoting EVS as pico-powerstations. My EV is handmade MR2>ZEV240 (150V 9" DC Warp motor, SkyEnergy batts)
Mr240-electric handmade car

User avatar
weber
Site Admin
Posts: 2626
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

barnard's Toyota MR2 - now Mr240

Post by weber » Thu, 16 Aug 2012, 18:11

I think we can forget loose links and corrosion, now that I see from the photo that you have braided links and everything looks very clean and dry, and when I read your description of careful preparation, and when I look again at the curves you provided.

Although I would still like to know how your cells are clamped together, i.e. in the north south direction relative to the photo. i.e. how are they prevented from swelling.

But I now strongly suspect your charging regime is the problem. I think the cells are just badly out of balance. i.e. those 10 that fall below 2.5 V during your tests are simply at a low state of charge. This would be because they have slightly higher self-discharge than the others and your charging shuts off too soon and never lets them get fully charged, and so over time they have gradually fallen further and further behind the others.

See this post today by Johnny, and its responses.
viewtopic.php?title=johnys-electric-vog ... 707#p38503

What charger do you have? What voltages, what currents, and what algorithm does it use and how is it controlled using BMS information?
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
weber
Site Admin
Posts: 2626
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

barnard's Toyota MR2 - now Mr240

Post by weber » Thu, 16 Aug 2012, 18:18

You might use your data logger on a set during a charge cycle.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
jonescg
Senior Member
Posts: 3036
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

barnard's Toyota MR2 - now Mr240

Post by jonescg » Thu, 16 Aug 2012, 20:39

The EV-Power BMS is great for monitoring, but not so good for balancing. This is usually because most chargers cut out when the charge current drops below half an amp or so, when really, for the balancing to occur properly the current needs to remain on at half an amp for another hour or so while the BMS modules do their thing.

I'm with Webber - try a balance charge (if possible) or if not, try to charge up the individual low cells and see how things perform.
AEVA National Secretary, WA branch chair.

zeva
Senior Member
Posts: 422
Joined: Sat, 15 Dec 2007, 05:09
Real Name: Ian Hooper
Location: Australia
MSN: sigmunky@hotmail.com
Contact:

barnard's Toyota MR2 - now Mr240

Post by zeva » Thu, 16 Aug 2012, 21:27

An assortment of comments…

- If you're drawing 500A from 100Ah Sky cells, it is to be expected that they will sag down to below the 2.5V threshold where EV Power BMS modules signal a warning - not that it would be damaging the cells, just the BMS modules are getting confused into thinking the cells are flat. Limiting the Zilla to 300 battery amps (3C) is likely to keep the voltage above 2.5V, but vehicle performance would take quite a hit.

- The braided connectors are tin plated so don't tend to have corrosion problems, but the aluminium terminals on prismatic cells *definitely* do, and even if you cleaned the terminals when first assembled, aluminium oxide tends to re-form over time unless you also added Noalox (or similar) to keep oxygen out. Anyway, looking for hot connections is probably a good way to see if Al2O3 issues are coming into play.

- I've done a bit of research lately into the requirements of balancing from a BMS, and have found that it's actually far lower than most people seem to think. As far as I can discern, cells in series only get out of balance due to differences in self-discharge. Other than internal self-discharge, cells in series always necessarily have the same number of electrons flowing in or out of them.

Self discharge with lithiums is in the order of 1% per month. Differences in self discharge is a fraction of this, amounting to say a few % capacity per year. For a typical EV-sized pack, that amounts to approximately 10mAh per day of imbalance. With a 500mA shunt, that is about 1 minute of balancing time required.

For comparison, many BMS modules themselves will consume over 100mAh per day just for monitoring (~5mA consumption). In fact, for many BMSs, their power consumption (and differences between) exceeds cell self discharge.

So it's certainly still possible that your cells have strayed out of balance, and doing a manual balance before jumping to any other conclusions is a good idea. If you have access to a laboratory power supply, you can use that to bring all cells up to 3.65V individually after doing a bulk charge with your regular charger.
Ian Hooper
--
"Never doubt that the work of a small group of thoughtful, committed citizens can change the world." - Margaret Mead
http://www.zeva.com.au

User avatar
coulomb
Site Admin
Posts: 3780
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

barnard's Toyota MR2 - now Mr240

Post by coulomb » Fri, 17 Aug 2012, 03:04

barney wrote: ... although Coulomb suggested a non-contact thermometer (are these easy to borrow, btw?) to measure all terminal temps after a drive looking for hot connections.

Well, they're easy enough to buy, and many people would have one, so I assume it would be easy enough to borrow.

Here's one for $50:

http://www.jaycar.com.au/productView.asp?ID=QM7215

Image

Dick Smith used to have one, but their site is useless for searching these days.

They're really useful, you'll wonder how you lived without one. They don't work in every situation, but for things like hotspots in electronics, they work well.
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.

drgrieve
Noobie
Posts: 12
Joined: Thu, 28 Apr 2011, 18:14
Real Name: David grieve
Location: Sydney

barnard's Toyota MR2 - now Mr240

Post by drgrieve » Fri, 17 Aug 2012, 03:36

zeva wrote: An assortment of comments…

Self discharge with lithiums is in the order of 1% per month. Differences in self discharge is a fraction of this, amounting to say a few % capacity per year.


I gather that self discharge is closer to nothing per year. I've seen a few reports on cells sitting on the self for over two years still reading the same static voltage within a few millivolts.

But I like where your train of thought is going.

I would encourage the OP to try the pack without a BMS for a while and see how you go.

Before you do that, I'd recommend you discharge each cell manually until its static voltage (resting for a hour will do) is 3.0V. They are now bottom balanced.

Then charge the pack at around 20 amps (if you can) until the voltage is 3.5V, then charge holding the voltage and dropping current until current is 5 amps. When current is 5 amps terminate the charge. This is known as Constant Current - Constant Voltage charging.

By counting the AH that went into the pack in step b, you know your packs usable capacity (20% to 95% SOC).

After an hour your pack voltage should average 3.33 to 3.35V. This is nice safe resting voltage approximately 95% SOC after a charge.

Now when you use the pack, count the AH coming out, and make sure you charge the pack before you use up the AH calculated previously.

Now what would be best if you do some testing on your "weaker" cells - charge as above but to 3.65v instead and discharge at 10 amps until 2.7V to see what your full AH is. Then charge back to 3.65 (terminating still at 5 amps). Then take 5% of ah out. Wait for cells to rest. This should be your target charging static voltage. Take another 75% of ah out. This should be your bottom balancing static voltage target.

By doing this testing your will gain confidence in understanding your cells behaviour.

Every month or so take a voltmeter when your pack is near empty (20% SOC). Your cells should all read exactly the same voltage give or a take a few millivolts. After a few times you will get bored and perhaps you might do it yearly if bothered.


roddilkes
Groupie
Posts: 176
Joined: Tue, 18 Sep 2007, 18:07
Real Name: Rod
Location: WA, Australia
Contact:

barnard's Toyota MR2 - now Mr240

Post by roddilkes » Fri, 17 Aug 2012, 11:31

Hi Barney,

What is your peak charge voltage?
That is, what is the highest voltage your charger gets to before switching off?

antiscab
Senior Member
Posts: 2489
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

barnard's Toyota MR2 - now Mr240

Post by antiscab » Fri, 17 Aug 2012, 14:27

jonescg wrote: The EV-Power BMS is great for monitoring, but not so good for balancing. This is usually because most chargers cut out when the charge current drops below half an amp or so, when really, for the balancing to occur properly the current needs to remain on at half an amp for another hour or so while the BMS modules do their thing.


I strongly don't recommend trying to do a balancing charge - remember 3.65V @ 0.1C is undercharged but 3.5V @ 0.001C is overcharged

The per-cell shunt balancers are good for slowly getting a pack top balanced, or for dealing with a single cell that has developed a small amount of self discharge

Putting a small current through the pack continuously is bad (increases pack internal resistance)
drgrieve wrote:
I gather that self discharge is closer to nothing per year. I've seen a few reports on cells sitting on the self for over two years still reading the same static voltage within a few millivolts.


This is my findings aswell, *but* cells that have been sagged below 2V under load (even fully charged cells) develop internal shorts resulting in some cells with self discharge
drgrieve wrote: I would encourage the OP to try the pack without a BMS for a while and see how you go.

Before you do that, I'd recommend you discharge each cell manually until its static voltage (resting for a hour will do) is 3.0V. They are now bottom balanced.

Then charge the pack at around 20 amps (if you can) until the voltage is 3.5V, then charge holding the voltage and dropping current until current is 5 amps.


I strongly recommend leaving the BMS in place - it's not just keeping the self discharge of the bad cells in check, its also a means to stop the worst cells sagging below 2v again

The bottom balance and limit average voltage to 3.5v approach only works when you have a new pack (cell capacity variation ~1%) and a variable voltage charger

You need the variable voltage charger as when the capacities of the cells within the pack start to drift, the voltage of the weakest cells start increasing as the stronger cells stay in the linear region longer (comparatively)

With a BMS, the charger will stop once the voltage of any cell reaches 4v

The charger also won't charge if you reverse any cells, alerting you to the problem then and there

without a BMS, eventually one of the weak ones will go above that voltage, and then you will be playing the replacing cells game

without a BMS, a charger will have no problem charging a pack with a reversed cell. If you are lucky, the additional 3.5V the rest of the pack sees will just cause the next weakest cells to overcharge.

If you are unlucky, the voltage never goes high enough and the charger never shuts down

As far as swapping the worst of the worst for new cells, it will work fine, but you will still be limited by the weakest of the remaining cells.

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

antiscab
Senior Member
Posts: 2489
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

barnard's Toyota MR2 - now Mr240

Post by antiscab » Fri, 17 Aug 2012, 14:34

drgrieve wrote:
Then charge the pack at around 20 amps (if you can) until the voltage is 3.5V, then charge holding the voltage and dropping current until current is 5 amps. When current is 5 amps terminate the charge. This is known as Constant Current - Constant Voltage charging.

*snip*

After an hour your pack voltage should average 3.33 to 3.35V. This is nice safe resting voltage approximately 95% SOC after a charge.


An easy way to tell if a pack has been overcharged is if the cell voltage doesn't fall back below 3.4v after it has been fully charged

I had a pack (15 cells) where every cell had been charged up to 3.65v
I removed everything from those cells and put them on a shelf

8 months later I came back and every cell still showed 3.65v

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

Post Reply