Nevilleh's BMW

Post up a thread for your EV. Progress pics, description and assorted alliteration
Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Nevilleh's BMW

Post by Nevilleh »

Yah, a piece of coat hangar wire 90cms long (a standard length once you cut the twisty bits off) measures 37 milliohms. Another good source of resistance wire is the galvanised iron wire they sell in gardening and hardware shops for tying up plants or something. Its a lot lighter, typically 18 g (1.25mm) but has a resistance of around 130 mR per metre. You can of course, parallel up several strands to increase the current carrying capacity.
I've used it before wound on a piece of ceramic tile to act as a discharge load for Li cell testing. It has a fairly high +ve temp coefficient so you end up with a lot more resistance than you started with when it gets hot! My cell tester had the thing immersed in an old ice cream container full of water to keep it reasonably stable.

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

Nevilleh's BMW

Post by Nevilleh »

I haven't tried field weakening yet and I'm not sure I'm going to bother as the performance is quite adequate as is.
I have collected quite a bit of data on the cells now, via a serial port from the bms master and I'm trying to format in a way that conveys useful information at which time I will start another thread and post it all. I'm wrestling with tying to get meaningful graphs from a mass of numbers in Excel and I wondered if anyone had any suggestions for software that might do a better job more easily?
One problem I have is importing the data - Excel can do it from a text file OK, but I have strings of 45 cell values separated by 0xFF and I can't figure out how to make Excel import it that way. Be grateful for any suggestions!

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

Nevilleh's BMW

Post by woody »

Nevilleh wrote: One problem I have is importing the data - Excel can do it from a text file OK, but I have strings of 45 cell values separated by 0xFF and I can't figure out how to make Excel import it that way. Be grateful for any suggestions!
Send it to me, I'll send it back as a CSV, and tell you how I did it. (probably a Perl one-liner. You can get Perl for windows from ActiveState)
Planned EV: '63 Cortina using AC and LiFePO4 Battery Pack

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

Nevilleh's BMW

Post by Nevilleh »

Will do, look for a PM.

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

Nevilleh's BMW

Post by Nevilleh »

Many thanks to Woody for writing a Perl of a program that converts my raw cell voltage data into a csv file that can be opened in Excel and hence graphed to show how the cells perform under load or charging or whatever.
For example, here's a single sample of 45 cell voltages in my compressed, raw form
4E 4F 4E 4F 4F 4E 4E 50 51 52 4D 55 4E 4D 4D 4C 51 57 56 54 4E 4E 4F 51 4E 4E 4E 4F 4D 4E 4E 50 51 50 50 51 4E 54 53 52 4F 4E 4E 4F 4E FF
(the FF is the terminator)
Woody's little program:
#!perl -wl

use strict;

my $count = 0;
my @sample = ();

while (<>) {
     @{$sample[$count++]} = map {
          1263405 / ( hex($_) + 316 )
     } (split)[0..44];
}

foreach my $cell (0..44) {
     print join (
          ",",
          map {
               sprintf("%0.0f",$sample[$_][$cell])
          } 0..$count-1
     );
}

turns it into this:
3207
3198
3207
3198
3198
3207
3207
3190
3182
3174
3215
3151
3207
3215
3215
3223
3182
3135
3143
3159
3207
3207
3198
3182
3207
3207
3207
3198
3215
3207
3207
3190
3182
3190
3190
3182
3207
3159
3166
3174
3198
3207
3207
3198
3207

which is a single sample of each cell's voltage in millivolts.
It works for any number of samples (I think!) and lets you make pretty graphs.

I use a terminal program to save the original data file, then Woody-ise it, stick it into Excel and print out lovely graphs! Just proves you don't need to know anything yourself as long as you have experts to help.

I have also revised the cell module pcb to allow it to fit three different terminal spacings, the CALB and Winston 40 AHr and the 60 AHr (81 mm). I took the opportunity to increase the board space for the shunt resistors and provide more copper all for better heat sinking and the result is the temperature rise when running at 3.6v such as you would use for top balancing is now just on 80 deg C - which is quite acceptable. I haven't bothered to measure it at the bottom balancing voltage.
Here's a picture of the new board:


Image

Hand soldered, by the way! It was the first one made to check the new layout and the rest (in my scooter) are oven baked. The un-populated 5 hole row is for the ISP header. I don't bother fitting that any more, just hold a header in place with my finger for the 2 seconds it takes to program and verify.
Last edited by Nevilleh on Wed, 10 Oct 2012, 03:59, edited 1 time in total.

User avatar
coulomb
Site Admin
Posts: 3833
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Nevilleh's BMW

Post by coulomb »

Nevilleh wrote: Woody's little program: ...
Wow. That sure does nothing for my belief that perl is a write-only language. Woody... what's the constant 1263405 got to do with anything, and how can it work without a comment justifying its value? Surely even a Perl God would forget why that value is what it is a few weeks after writing it   Image It seems to work well, though.
Just proves you don't need to know anything yourself as long as you have experts to help.
Well, I like to have a clue, and of course you still need to know what to ask.
Here's a picture of the new board:

Image
Nice. Though R11 and two other components seem to be a bit close to one of the mounting holes, where a cell bolt and lock or other washer need to go. Maybe they are low enough in height that the socket needed to tighten the bolt will clear them, but it seems a little tight.
The un-populated 5 hole row is for the ISP header. I don't bother fitting that any more, just hold a header in place with my finger for the 2 seconds it takes to program and verify.
That's very similar to the JTAG connectors on our boards. The problem I find is when debugging, so you have to hold the connector on for some time. And if you lose contact for a millisecond, our JTAG driver software gets all upset.
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.

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

Nevilleh's BMW

Post by Nevilleh »

I'm to blame for that constant, not Woody! He just put it in because I told him to and a comment saying "Neville said" would not be helpful. Actually, it comes from 1023 *1235 which is Vref in mV x the ADC max count. My upside down use of Vref and Vs, plus the compression of the 10 bit ADC count down to one byte is the reason for it:

V (mV) = ((1023*1235)/(ADC count)

and ADC count = stored byte + 316

so you can see where all the magic numbers come from. This is also why the voltage range is restricted to 2.21v to 3.99v, but a string of bytes is much better than a string of 10-bit numbers.

Those parts on the board are just less than 1mm high and a washer under the bolt head lifts it high enough for a socket to clear them.

If I were using the debugger, I'd solder in a header, but the design is stable now and the connector is only used for initial programming.

I wonder if "Mother of Perl" might be more appropriate than Perl God? His little program impressed me enough to download Perl and have a look for myself. See what you mean about a write-only language though.

I probably should have put all this under the BMS topic, but the two are closely connected, if you'll forgive the pun......
Last edited by Nevilleh on Wed, 10 Oct 2012, 07:08, edited 1 time in total.

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

Nevilleh's BMW

Post by woody »

Dear Mr Reverse Polish,

The program was a one-liner, but Neville wanted columns and rows swapped, and quoting didn't work right for him on windows. The random numbers came straight from Neville...

perl -nle'print join ",",map {1263405/(hex($_)+316)} (split)[0..44]'
BMS1.log > BMS1rows.csv

The Perl operators are pretty simple:
Join - makes a string by joining a list with a given character, e.g. A Comma or space
Map - does the same thing to each element of a list on the way through
Split - splits a string into a list, dividing on white space.
Print - duh

And the options:
-n - reads each line of input into the default variable, $_ and run the code on each line.
-l - take the carriage return off $_ on the way in, and add it back when printing
-e - run code from the command line

Cheers,
Woody
Perl Guy
Planned EV: '63 Cortina using AC and LiFePO4 Battery Pack

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

Nevilleh's BMW

Post by Nevilleh »

My motor coupling failed recently. The one between the two motors, that is. Thought I'd describe what happened as a warning to other people who might couple two motors together.
I used an old BMW rubber "doughnut" drive shaft coupling modified to do the job and I thought that would be quite satisfacory, bit like a Rotoflex but much cheaper! What actually happened is that the auxiliary drive shaft on the rear motor completely snapped off. This is the shaft out the front and the one that is couple to the other motor's main drive shaft.
The main drive shaft is 1 1/8 inch and the aux one 7/8. but I thought that ought to be able to handle the motor torque OK. There's no warning in the motor specs about limiting the max torque on that shaft.
Anyway, my very clever engineer mate looked at it and decided the two motors were not in good alignment resulting in a flexing of that shaft which caused it to eventually snap off. My rubber coupling which is supposed to take up any flexing had gone as hard as a brick and we think that, with the misalignment, caused the problem.
I bought a new rubber coupling which is infinitely more flexible and we spent a lot of time aligning the motors resulting in shimming the mounting plate of the front motor to get it all in line. By the way, I simply swapped the front motor with an unused aux shaft for the rear one which saved having to do any work on the motors - or indeed replacing the one with the broken shaft.
If you are looking for thin aluminium shim on a Sunday when nothing is open, go to the local shop and buy a can of Coke. Its about .2 mm thick and perfect for the job! You can even drink the Coke, although not recommended.
Anyway, its all back together now and driving as well as before. I might look up a mechanical engineering textbook and see if I can work out what the max torque on a shaft that size can be.
I'd like to add that a new BMW coupling was over $300 and I probably could've bought a Rotoflex for not much more than that!

Renard
Groupie
Posts: 219
Joined: Sun, 29 Aug 2010, 18:55
Real Name: Robert Fox
Location: Cobargo NSW

Nevilleh's BMW

Post by Renard »

Nevilleh wrote:
If you are looking for thin aluminium shim on a Sunday when nothing is open, go to the local shop and buy a can of Coke. Its about .2 mm thick and perfect for the job! You can even drink the Coke, although not recommended.


This remark brings to mind a passage in 'Zen and the art of motorcycle maintenance' which has always stuck in my mind, ever since I read it some forty years ago.
The narrator's friend rode a BMW bike whose handlebars worked loose, and the narrator fixed it by finding an aluminium beer can by the roadside and cutting a piece out, shimming out the shaft so that the pinch grip of the handlebars had something to bite on.
Expecting appreciation from his friend, the narrator was puzzled to find that his friend was upset and uncomfortable.
Then he realised that his friend would have been oh so happy if a special 'BMW shim' had been flown in from some gleaming Fabrikwerks in Germany. That would have sustained the image of a shining, romantic, mythic machine which was transporting him in magic style. A beer can by the road completely destroyed the image.
So the narrator came to realise that there were two types of people, and most people on this forum are, like myself and the narrator and Neville, pragmatic types, but I suspect some contributors here have leanings the other way.
Renard

User avatar
weber
Site Admin
Posts: 2643
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Nevilleh's BMW

Post by weber »

Hi Neville. Thanks for the cautionary tale. The misalignment and the too-rigid coupling would be enough to explain it, except in that case I'd expect the bearing nearest the break to be shot as well. Did you check it? If it was still good, the smaller shaft diameter may have been a significant contributing factor. Torque capability of a solid shaft is proportional to its diameter squared. So the 7/8" shaft has half the capability of the 1 1/8" shaft.

Hi Renard. Thanks for the ZATAOMM story. So true.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
jonescg
Senior Member
Posts: 3104
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Nevilleh's BMW

Post by jonescg »

Zen is an awesome book, one everyone must read.

On the subject of mating two motors, if my Agni motors taught me anything, it's

a) don't do it unless there is no choice
b) if you have no choice, use a very solid coupling with minimal flex and align perfectly, and
c) run your DC motors in series to share the current evenly. If you;re running AC, just use one big motor.
AEVA National Secretary, WA branch chair.

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

Nevilleh's BMW

Post by Nevilleh »

weber wrote: Hi Neville. Thanks for the cautionary tale. The misalignment and the too-rigid coupling would be enough to explain it, except in that case I'd expect the bearing nearest the break to be shot as well. Did you check it? If it was still good, the smaller shaft diameter may have been a significant contributing factor. Torque capability of a solid shaft is proportional to its diameter squared. So the 7/8" shaft has half the capability of the 1 1/8" shaft.

Hi Renard. Thanks for the ZATAOMM story. So true.


The bearing was fine. The shaft actually steps down to the smaller size just outside the bearing. Why on earth they did that, I have no idea. Would've been easier to leave it the same size all through.

User avatar
weber
Site Admin
Posts: 2643
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Nevilleh's BMW

Post by weber »

Hi Neville,

I think Section 1.6 (pages 21-25) of this document might explain what happened with your shaft. The problem is not the reduced diameter, but the abrupt transition between diameters that concentrated the stress caused by the misalignment and the too-rigid coupling.
http://www.engineersaustralia.org.au/si ... rt%205.pdf

Of course the non-drive-end (NDE) shaft of an induction motor is typically only intended to drive the motor-cooling fan, so this step change of diameter wouldn't normally matter. We are kind of making the same mistake as you. We are using our NDE shaft (which is 25 mm versus the 38 mm of our DE shaft) to drive the aircon compressor, with a permanent side-load due to the pulley and belt. But hopefully this is way less stress than what happened in your case.
jonescg wrote: Zen is an awesome book, one everyone must read.

On the subject of mating two motors, if my Agni motors taught me anything, it's

a) don't do it unless there is no choice
b) if you have no choice, use a very solid coupling with minimal flex and align perfectly, and
c) run your DC motors in series to share the current evenly. If you;re running AC, just use one big motor.

The only part I don't agree with is (b). Sure you try for perfect alignment, but perfection belongs only to the realm of the Platonic Forms. Here in the real universe, there are limits to precision, and variations with age, temperature and accellerations ("gee forces").

So what makes you recommend a solid coupling?

In theory you could have a solid coupling if the bodies of the two motors had independent flexible mountings, but I'm guessing their inertia probably means this would be of limited value.

We do have a rigid coupling between two shafts in the MX-5, with independent flexible mounting of the bodies. But this is between our induction motor and its shaft-encoder which only weighs a few grams.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

HeadsUp
Groupie
Posts: 265
Joined: Sat, 28 Feb 2009, 09:13
Real Name: Mark W
Location: Sydney

Nevilleh's BMW

Post by HeadsUp »


i think i warned you about this when you first posted pics of the motor frame a couple of years ago ?

your aluminium motor mounting frame is not rigid

if you try to connect two motors without a flexible coupling you will break shafts.

end of sermon . nothing more to discuss

User avatar
weber
Site Admin
Posts: 2643
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Nevilleh's BMW

Post by weber »

HeadsUp wrote:i think i warned you about this when you first posted pics of the motor frame a couple of years ago ?

your aluminium motor mounting frame is not rigid
Hi HeadsUp. I don't see where you said that exactly, but you made me go back and look at Neville's cradle photos.

Image

Image

Neville, It's true that for the same thickness, steel is 3 times stiffer than Aluminium. And for the same weight, stiffness is the same but Aluminium costs more and takes up 3 times as much space.

But I think the biggest problem is the asymmetry of the cradle connecting the motor bodies. The way it is, more torque will cause more parallel offset misalignment. It wouldn't matter if the motor bodies were connected with lasagne (high tensile) provided it was arranged symmetrically around the circle so the relative rotation of the two motor bodies was centered on their shafts.

It may also flex causing angular misalignment when you go over bumps.

The solution seems simple. Add a bolt-in "hat" like the one in Zeva's cradle, shown here.

Image

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

Wellsey
Groupie
Posts: 43
Joined: Wed, 30 Jul 2008, 07:29
Real Name: Brad Wells
Location: Canberra

Nevilleh's BMW

Post by Wellsey »

Have you seen this Australian designed "Thompson Couplings"?

http://www.thompsoncouplings.com/site/p ... videos.php

The alignment eliminator is supposed to accommodate up to 3 deg of mis allignment and I'm sure you guys are more accurate than that. Alternatively you could use the Constant velocity coupling.

Brad

User avatar
jonescg
Senior Member
Posts: 3104
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Nevilleh's BMW

Post by jonescg »

weber wrote:
jonescg wrote: Zen is an awesome book, one everyone must read.

On the subject of mating two motors, if my Agni motors taught me anything, it's

a) don't do it unless there is no choice
b) if you have no choice, use a very solid coupling with minimal flex and align perfectly, and
c) run your DC motors in series to share the current evenly. If you;re running AC, just use one big motor.

The only part I don't agree with is (b). Sure you try for perfect alignment, but perfection belongs only to the realm of the Platonic Forms. Here in the real universe, there are limits to precision, and variations with age, temperature and accellerations ("gee forces").

So what makes you recommend a solid coupling?


I'd say this would be my main reason:
Image

I had two couplings, one was the drive sprocket, the other was a sleeve. Under a high torque scenario, it just popped out.

So I instead went for a single solid shaft with a drive sprocket at one end:

Image

This one needed the odd brass shim behind the bolts to make the motors spin fairly freely, but nothing came loose or even vibrated unpleasantly. Given the nature of the setup, I can't think where you could possibly mount a flexible shaft coupling.

Granted in Neville's situation, a flexible coupling would work better than in my situation - as all of the torque was being applied to the actual coupling via the chain.
AEVA National Secretary, WA branch chair.

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

Nevilleh's BMW

Post by Nevilleh »

Thanks for a lot of interesting comments guys.
I THOUGHT I had a flexible coupling! I THINK the one I have now is much more flexible than the last one. Certainly it is possible to flex the rubber disk by hand, something not possible with the old one - which came from a wrecker, by the way. I think that if it happens again I will revert to a single motor and go back to a 1000A controller to regain the lost torque. That'll stress everything lots more though.
Last edited by Nevilleh on Mon, 07 Jan 2013, 00:40, edited 1 time in total.

User avatar
jonescg
Senior Member
Posts: 3104
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Nevilleh's BMW

Post by jonescg »

For those who don't quite see what I had going on with my old coupling. It wasn't a 'flexible coupling' as such. It was a coupling which had flex, when it shouldn't have. The points of flex were the drive sprocket coupling and the sleeve coupling. As I said, it's not really appropriate to compare it to Neville's situation, since the torsional forces due to the chain are strongly radial. The solid shaft joining the two output shafts of the motors was much better.
AEVA National Secretary, WA branch chair.

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

Nevilleh's BMW

Post by Nevilleh »

I think I would have to say that, unlike heads, motors are best kept single.
Some figures from my car taken over the last 1000 kms:
Average amp-hrs/km = 1.25
With 45 cells at say 3.2 V, that's about 176 watt-hrs per km. I find the AH/km most useful as I can quickly estimate range etc just from my odometer.
I have a storage capacity of nominally 120 AH, but I measured it as 140 taking the volts down to 2.5, so allowing for 80% discharge, I have 112 usable AH and if desperate, can go to 125 (90% discharge). I suppose if REALLY desperate, I could risk the battery by going to 140 AH.

Anyway, 112 AH can take me 90 kms, 125 Ah will do 100 kms and 140 would be 112 so I watch my trip meter - I reset it every charge - and know I can go 90 kms if I have to with maybe a wee reserve. I tend to charge it every time I use it and most days I seem to only do maybe 25 kms or up to twice that, so I always have plenty in hand.

I've started top balancing the battery to see if it makes any difference, but it is taking a long time to get the cells starting to come into line at the top end. Not surprising as they were well bottom balanced before. One thing I've noticed already though, is that the fully charged voltage is higher ie more cells are getting up to close to 3.6 when the first one gets there, so the car is actually a bit faster when freshly charged! It easily does 110 kph now when hot off the charger.

Interesting stuff.

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

Nevilleh's BMW

Post by Johny »

Interesting stuff Neville. I tend to "think" in AH as well. Even though it must change over the charge (lower cell voltage means more amps to go a certain speed), it's so much easier to figure charging time etc. when you know what has come out (AH) at a glance.
BTW I have 192 cell pairs and 3.2V/cell over that many is about right as well. I see 610-630 VDC most of the time during driving except for heavy load (and heavy regen) and initial few km. (Admittedly I haven't driven it much yet).

a4x4kiwi
Senior Member
Posts: 772
Joined: Thu, 03 Jan 2008, 19:04
Real Name: Malcolm Faed
Location: Australia
Contact:

Nevilleh's BMW

Post by a4x4kiwi »

Thanks for the info. That helps with my estimated range calculations for my E30.
Silicon is just sand with attitude.

Blog: http://malfunction.faed.name

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

Nevilleh's BMW

Post by Nevilleh »

One other thing to note is that if I drive it at a steady 100 kph the consumption rises to about 1.8 AH/km. Brick aerodynamics!

I have a guilty secret to confess:
I've received notification that my new car is on its way and its NOT an EV, its powered by a 6.3 litre, gas guzzling, V8. (Hides head in shame).
Last edited by Nevilleh on Tue, 08 Jan 2013, 01:10, edited 1 time in total.

Canberra32
Groupie
Posts: 238
Joined: Wed, 13 Jun 2012, 04:25
Real Name: Damian butcher
Location: Canberra

Nevilleh's BMW

Post by Canberra32 »

And you should hang your head!!! There are bigger V8s out there you settled for 6.3 shame shame shame!!
You had better force some induction so your not a sissy pants girls blouse!

Post Reply