Weber and Coulomb's MX-5

[acmotor]

Hmmmmm,
TIC
I get three images. Two outer ones with enough steel to keep BHP happy and a centre one of unobtainium weighing less than half the weight.

[Canberra32]

Hey what war are you in?
The tank like battery box is quite the panza
I would be taking to that with a drill and lightening the load bebeh.
One thing I have learnt from race cars is that we often under estimate the strength of the material we use and the load it can handle.

I made a battery box from 25x25 angle and drilled the hell out of it with a 13mm bit and it holds 14x 180 cells that weigh 5.6 kg each and it didn't twist or distort under pressure.
And while tig is awesome and its all I use sometimes you just can't get the torch in there so u mig the crap out of it

You got be excited now your so close to the finish line

[Johny]

Instructions on how to view cross-eyed 3D images

Damn - now I can't see properly... (Just joking - but there is a warning on the site.)

[weber]

Yes, a ridiculous amount of steel for just seven 40 Ah cells. 14 kg total mass, 2.8 kg box.

Here is another cross-eyed stereo pair to delight those who can "get" them and frustrate those who can't. This time one of the two completed front mudguard battery boxes.

How to view cross-eyed 3D images

[Canberra32]

How did you seal the top of the box to the cells? I see the silicon around the windo but curious how it was sealed down

[coulomb]

How did you seal the top of the box to the cells?

I somehow wasn't present for when the other covers were made. So when I cut the clear PVC, I didn't take heed of the curl. (We have some polycarbonate, but can't get any more, so we keep that for where impact resistance is needed.)

Bzzzt. The inside of the tops aren't flat; there are pieces of angle iron that sit on top of each other, so parts of the lid are 3 mm higher than the other. It's important to get the curl to help with that unevenness. So we had to re-cut some of the pieces I had done.

The secret, I'm told, is lots of silicone.

But that's just the clear lid. We also have foam rubber (36 mm x 10 mm) inside the lid. There is one layer stuck inside the lid, and we put another layer at the end of each row of cells, siliconed to the tops of the end cells. These have cut-outs for fibre optic connectors, some larger components, the edge of the BMU printed circuit board, and cables that exit the box. The foam rubber on the cells ends up inside the foam rubber on the lids. Hopefully, this makes a pretty good seal.

We also have plastic under the cells at the bottom of the boxes. We don't seal these, so that any water that makes its way into the boxes can drain out, but the main stream of run-off from tyres and the road will hopefully be deflected by the bottom plastic.

When cutting the lid (e.g. to allow straps or cables from one row of cells to the next), we found it is important to protect the foam rubber from flying iron particles. These fly off red hot, and make holes in the foam rubber, rendering it useless and also unattractive.

[ Edit: spelling, size of foam rubber. Plastic is PVC, not acrylic. Also, added this image, where you can just barely see the foam rubber: ]



See also this post, for details of the foam rubber at the ends of the cell rows.

[weber]

How did you seal the top of the box to the cells? I see the silicon around the windo but curious how it was sealed down

I see Coulomb has answered your question admirably while I was composing the following. I've posted it anyway, for the few additional details it contains.

The grey stuff you can see is 36 x 10 mm self-adhesive polyethylene foam from Clark Rubber, stuck to the vertical faces on the inside of the lid and siliconed where it meets at the corners. You have to kneel on the lid to compress this foam, to get the lid bolts started.

The power cable and optic fibre connector exit vertically between two layers of this foam, at the row ends you can't see in the images above. The inner layer of foam is siliconed to the top of the cell, after being cut to fit around the right-angle power-cable lug and optic fibre connector.

[Edit: BTW, We brush talcum powder onto the remaining exposed self-adhesive to inactivate it.]

Many thanks to everyone who helped get these two battery boxes designed and built in record time: Coulomb (Dr Mike Van Emmerik), Jeff Owen, David Keenan senior (my father) and Mark Aylott.

[Nevilleh]

The trials and tribulations of the EV maker! Wondrous to behold - speaking of which, that 3D business completely escapes me and all I get is blurred vision! Are you sure you're not having us on?

[weber]

The trials and tribulations of the EV maker! Wondrous to behold - speaking of which, that 3D business completely escapes me and all I get is blurred vision! Are you sure you're not having us on?

Thanks Neville. I must admit I have only heard from one other person who can see these, Jeff Owen. It may help to be farsighted, since what you have to do is to converge your eyes on a point that is between you and the screen, but without focussing there. Your plane of focus still has to be at the screen. In normal seeing these two things are not separate, so it can take a while to learn to separate them.

I made the images larger this time. This makes it harder to get, but more rewarding when you do. But to get started, I suggest you zoom way out so corresponding points on the two images are only about 30 to 35 mm apart (about half the distance between your eyes). In Firefox you can do this by typing Ctrl minus-sign many times. Ctrl zero will get it back to normal.

Sit square-on to a point on the screen between the two images. Then make a circle with your thumb and a finger (vitaka mudra) and hold it so you can look through it, at the region between the two images. This hand aperture should be about half as far from the screen as it is from your eyes (so 1/3 : 2/3). This is the same as the ratio of image-spacing to eye-spacing.

Start by looking at your hand, but notice the blurry overlapping images through the aperture. You have to get the frames of those two blurry images to align both vertically and horizontally. The horizontal alignment is achieved by moving your hand slightly towards or away from the screen. The vertical alignment is achieved by tiny adjustments to the tilt of your head. Once you've locked in the alignment, hopefully the focus will gradually become clearer without losing the alignment.

If you've ever "got" one of those random-dot stereograms, I'm sorry to have to tell you that this is completely different, in fact the opposite. With those you have to converge at a point _behind_ the image. i.e. They required wall-eyed viewing, not cross-eyed. You could view the above images wall-eyed if you swapped them left for right and kept them small. If you view them wall-eyed without swapping them, the object will appear paradoxically inside-out and with parts missing, which sounds like what acmotor described as "a centre one of unobtainium weighing less than half the weight"

[weber]

The photo below shows the base, and the other faces you can't see in the above images. You can understand how tight a fit this is when you see that we have not only cut away the corners of the steel base where the plastic bumper curves under, but we have cut away part of a cell!



In an earlier post you will see that we thought we were only going to get 6 cells in each of these spaces, with only 24 mm minimum clearance from the tyre. In fact we achieved 7 cells and 34 mm clearance (which is the same as the existing minimum clearance from tyre to wheel-arch).

NCOP 14 requires that batteries must not break free of the vehicle in a crash with
Front impact – 20 g (i.e. 20 times the battery weight);
Side impact – 15 g;
Rear impact – 10 g; and
Vertical (rollover) impact – 10 g.

The largest mounting tab (lower right in this image), carries most of the responsibility for this, with a single class-8.8 M10 bolt attaching it to a captive nut inside a sloping reinforced section on the underside of the chassis rail, where a tie-down loop used to be bolted. The other two mounting points are just to stabilise it in normal operation and are only M6, one of which is indicated by a hole in the side of the lid in the above image. All 3 mounting points make use of existing bolts.

The lower outer corner of the plastic bumper used to have a 5 mm dia steel bracing strut that passed through the space that the battery box now occupies. A shortened strut now braces the bumper to the battery box. You can see the M5 hole in the base that it threads into.

[coulomb]

In the above photo, you can also see that we put a bit of thought and design into the clamping for the two cells that are at right angles to the other 5. (This is the first battery box to be this way).

We're constrained to using the bottom of the third and the top of the fifth "grooves" (inter-rib spaces) of the cell for running the threaded rods. So as you can see, the threaded rods appear to intersect. We managed this by welding the threaded rod to the steel channels at the back of the set of two cells.

[ Edit: clarity; had conflated two types of ribs. ]

[weber]

Here's a photo with the lid off, showing how the cells and BMUs connect. It also shows a better view of the doubly-gusseted main mounting tab (50 x 50 x 3 mm) with its M10 bolt hole. Most of the box is made from 25 x 25 x 3 mm mild steel angle. The three lid bolts are M8 x 12 mm (class 8.8). The clamping rods are M5 stainless. The cell terminal bolts are M6 stainless with a belleville washer and a flat washer.



The next photo shows the morror-imaging of the two boxes, and shows the small brackets that are used for the outer mounting because the existing M6 bolt that is used is inaccessible when the box is in place. So the bracket is bolted to the existing M6 and then when the box is in place the box lid is bolted to the bracket with another M6.



All this effort for only 14 cells! But it does get us back to the MX-5's original 50:50 weight distribution while only losing 10 cells from our previous design. Well OK the original MX-5 had 50:50 with driver and passenger. We only have 50:50 without them.

The plan now is for 218 cells instead of 228, so 720 V instead of 750 V. And we want to allow for a 200 kg payload (80 kg driver, 80 kg passenger, 40 kg luggage) distributed 143 kg rear 57 kg front. So the masses are:

             Total    Rear    Front   Distribution Rear:Front
Unladen      1380 kg  692 kg  688 kg  50:50
Driver only  1460 kg  741 kg  719 kg  51:49
Fully laden  1580 kg  835 kg  745 kg  53:47

There are severe constraints on what numbers of cells we can have, due to the requirement to have two equal half-packs (for the dual chargers, heaters and DC-DC converters) and the desire to not split any boxes between half-packs or remove any cells from any box except the rearmost. But we do have a fallback position with 202 cells, where we remove another 16 cells from the rear (we have already removed 24 from the rear). These are worth about 2 kg per cell including steelwork and ancillary electricals.

The above figures have been OKed by our engineer, but are contingent on passing braking and emergency-lane-change tests. We have already upgraded the brakes. The suspension is next. I think I want variable everything (ride height, spring constant and damping factor) so it can be tuned by trial and error based on actual driving, after it's registered. And to allow for the 202 cell fallback position.

[weber]

Here's the new 218 cell battery layout diagram for the MX-5.



Here's the old back-heavy 228 cell version for comparison.

[BjBlaster]

I like it

[7circle]

Great build description W & C.
Just wondering if Charger A Pos should link to Controller Pos.

Not sure on parallel charge links.

Take care
7C

[weber]

Great build description W & C.
Just wondering if Charger A Pos should link to Controller Pos.

You are so right! Well spotted. Fixed now. Thanks, 7circle!

[coulomb]

To replace a deleted post, some actual EV content. Progress is creeping forward:



You can see

* 99% completed contactor boxes (just needs some cables siliconed) under where the controller normally sits. These control the charger, DC/DC, and heater for both half packs.

* One of two completed wheel-well packs (bottom left corner of the photo)

* DC/DCs mounted and HV end wired (one is visible between an empty plastic box and the controller)

* DC/DC output cable has been routed but not yet connected (pair of red wires draping over the side of the car

* B side cable running to the Charger/DC-DC/heater contactor box has been routed but not connected (top right of the bonnet)

* The empty contactor box is where we intend to put some wiring to control the contactors (possibly with some relay logic). Two split conduits draped over the side of the car at the bottom of the picture have 8 wires for the various contactors in the Charger/DC-DC/heater boxes; other contactor connections will likely connect to there.

* Not visible in the photo, the optical connections are all complete, and all that "just worked" first go. However, there is a problem with slightly higher than expected "link voltages" that we need to sort out one day.

* Also not visible in the above, we re-jigged one of the A half pack contactor boxes to become straight through, removing a contactor. This removes the 19-cell rollbar box from the A half pack; it will become part of the B half-pack. It is replaced by the new 14-cell wheel-well boxes, this losing 5 cells from the half-pack. We've had to lose 10 cells overall (5 from each half pack) to get our weight to manageable levels.

* Also not visible in the above, the under-boot battery box has had two cells added to it. It's part of the "B half pack" which will be finished later; we're concentrating on the A half-pack initially.

So still to be done:

* All contactor boxes for the B half-pack; there are 7 contactors in that.
* Finish the contactor control logic.
* Finish the DC/DC wiring, connect up and test.
* Finish charger wiring, and install chargers. Connect up and test. This will finally allow our long-suffering auxiliary battery to get charged without having to clip on a mains charger.
* Suspension upgrades.
At this point, the car should be drivable and registrable.
* Finish off the B side contactor boxes.
* Upgrade the motor controller wiring from 16 mm^2 to 35 mm^2. We temporarily made 16 mm^2 connections ages ago, just to get the car running.
* All the myriad of miscellaneous jobs that are needed to finish off an EV.

[ Edit: added upgrade of motor controller wiring, miscellaneous. Edit2: 99% refers to the contactor box, not the whole car. ]

[Canberra32]

Are you worried electricity will be obsolete by the time you finish :p ?

[acmotor]

There is still time, given the slow uptake of EVs in Oz !
What, there is less than 1,000 EVs in the whole of Oz.

[Renard]

* 99% completed

Ahh that last 1%.

That is a delicious photo. It looks like my wife's suitcase before a trip -- just before I'm asked to sit on it to get it to close.
More strength to your backside!

[Johny]

It looks like one day you'll look at each other and pronounce it finished and there won't be enough room to stow the bonnet stay.

[weber]

* 99% completed

Ahh that last 1%.

Hee hee. Coulomb's wording was misleading. It was only a pair of contactor boxes that was 99% completed. Not the whole car. Sigh. At least those two contactor boxes are now 100% completed.

It looks like one day you'll look at each other and pronounce it finished and there won't be enough room to stow the bonnet stay.

Ha! It does look like that, but in fact the bonnet stay is a cold-bending work of art by Jeff Owen. He bent and rebent it until it had the standard (for this project) 0.5 mm clearance from all obstacles, in both the upright and folded positions (and yes, along the path in between, tortuous though it may be). It fits neatly between the battery box and the WaveSculptor when folded.

Last Friday we noticed a faint scraping noise coming from the front of the motor. It turned out to be an idler pulley rubbing against the head of a bolt. We probably hadn't allowed for the thickness of the second coat of paint on the raised lettering on the bolt-head. So I held a file against the pulley as it turned, until the pulley stopped scraping on the bolt. Not even 0.5 mm clearance on that one.

[weber]

Quite a milestone for the MX-<lightning bolt> today. For the first time we have smooth control of the motor-controller's battery current limit by the battery monitoring system (BMS). What does that mean in real life? (I hear you ask)

It means that with half our cells installed (109 x 40 Ah), we turned up the Whinge-factor(TM) on our cell-top battery monitoring units (BMUs) to simulate cells having several times their actual internal resistance, and we set the Tritium WaveSculptor's maximum battery current up to 180 A at 360 Vdc. And when I floored the accelerator from a standing start in first gear on a short stretch of bitumen at the front of my place, the system responded to the simulated undervoltage distress of the worst cells by winding the motor power and hence the battery current down to whatever it took to make their pain go away. It did this smoothly and within one second. It overshot to about 120 A, chirping the tyres once, before backing off to about 60 A. Then I yanked my foot off the accelerator and it performed the same trick with the regen braking, again chirping the tyres once. But this time the distress was caused by overvoltage. Max regen torque was set to 60% of max drive torque -- a setting for "experienced" accelerator-regen drivers.

CAN bus logging confirmed that it was being limited by dynamically-reducing battery-current-setpoints in both cases. Hoorah!

Earlier attempts had the gains set too high in the PID controller running in our BMS master (a modified Tritium Driver Controls unit) which made the MX-ϟ hop like a kangaroo as the battery current limit slammed from 180 A to zero and back to 180 A, 2.5 times a second, chirping the tyres each time.

[coulomb]

High impedance inputs are bad

At times, we have referred to our MX-5 conversion as a test bed for the Battery Management System (BMS). BMS design (two main ones, analogue and digital, many revisions of each), testing, and building has consumed some three years (at the rate of about 1 day per week for two people).

One feature that we were quite proud of, because no-one else seems to have thought of it, is our ability to measure the voltage drop across the battery links. This allows us to find out in advance when a link is developing (relatively) high resistance. High resistance is bad, because it limits power, and causes local heating. In severe cases, it can melt battery terminals or even start a fire. This is relatively less serious with a higher voltage pack (and more to the point, the consequent lower current draw). But it's still important.

It's a simple enough idea. You have a spare analogue to digital converter port (the MSP430 chips we are using have 8 I/O lines that can be configured as analogue inputs), and you measure the voltage across the link. One end of the link is the negative terminal of your Battery Management Unit (BMU) (you could also do it on the positive end). So that means one extra wire to get to the other end of the link. With the squiggle joins, these are mostly "for free".

The problem comes about if the link is ever very high resistance, or worse, completely open circuit. This even happens deliberately at battery breakup contactors, but you can just choose not to try to measure the voltage across those links. But unexpected high resistance or open circuit links cause the entire pack voltage to appear across the link. We were designing for a peak pack voltage when charging of 820 V, so the circuit has to be able to withstand (for a short time at least) that voltage, while normally measuring a few millivolts. The protection diodes in the microprocessor are capable of handling only about 2 mA. That implies a high resistance; ohms law say R = E/I = 820 / 2e-3 = 410 k. The power in that resistor is E*I = 820 x 2e-3 = 1.6 W, way more than a small surface mount resistor can handle.

There are also difficulties attempting to contain 820 V safely on a very small printed circuit board, where clearances are necessarily small. A single 0805 resistor, for example, won't withstand more than a few hundred volts; it's just too small.

We ended up with a megohm of resistance, split over 6 0805 resistors, limiting the current to 820 uA (= 0.82 mA), so the highest value resistors (180k) end up dissipating a power of I^2.R = (.82e-3)^2 x 180e3 = 120 mW. They are rated at 125 mW, so they won't burn the house down if the link opens circuit in the middle of the night while charging.

The circuit we have has a voltage divider made up of this 1 megohm resistor with another megohm resistor to ground. So the input impedance is half a megohm.

But half a megohm of impedance is substantial. A really high link resistance might cause 100 mV of voltage drop, so that might be the highest limit you'd want to set. The current needed to cause 100 mV into the voltage divider is I = E/R = 100e-3 / 500e3 = 0.2 uA, or 200 pico-amps. Is it possible that electromagnetic interference (EMI) from the motor controller might induce that sort of current? It turns out that it is, especially for battery boxes near the motor controller and motor. We now find that every link that needs an explicit wire longer than about 75 mm (basically, at the end of every row of cells, and in other boxes where there are sets of cells split to clear various interferences), we get nuisance stress from the BMUs.

We thought we tested the BMUs pretty thoroughly; progress on the conversion stalled for many weeks when we were chasing noise. It's still not obvious to me how we missed this. I guess at the time of testing, we weren't looking at the link voltages much, except for a few spot checks. Now we are using them to contribute towards cell stress, which cuts back the motor current.

We also don't know what impedance would be acceptable; all we know is that half a megohm is too high. We'd like to "fix" this problem some time, but it's also not obvious how to do this without needing a board full of resistors to handle the power dissipation, which will only be seen under fault conditions.

So maybe the reason that no-one else's BMS (that we know of) has link voltage measurement is that it's harder than it seems?

In summary, we failed to consider how sensitive to noise our circuit would become as a result of high resistor values. Perhaps also, if no-one else implements an apparently useful feature, maybe there's a good reason for it.

Edit: moved here unchanged from the "Some don'ts" topic, to keep technical discussion out of that topic.

[woody]

possibly a simpler and almost as good measure is comparing the sum of the battery internal resistances to the motor controller internal resistance. When a link goes bad this difference should get wider, the problem then is to find out which link...

Could be used with your system as a "only check for high resistance links when it looks like there is one"...