Page 54 of 68

Weber and Coulomb's MX-5

Posted: Thu, 03 Oct 2013, 15:33
by Adverse Effects
awsome to see

grats guys

you are going to have to change your tag line in your Sig from

Working with Coulomb on converting an MX-5 to an MX-ϟ

to some thing like

finished converting a MX-5 to a MX-ϟ with Coulomb

or

Converted a MX-5 to a MX-ϟ with Coulomb


Weber and Coulomb's MX-5

Posted: Thu, 03 Oct 2013, 23:23
by woody
Congrats Guys, are you driving to/from Canberra? Stop by in Mt Colah on the way home and say G'day and have a cuppa.

Weber and Coulomb's MX-5

Posted: Thu, 03 Oct 2013, 23:56
by a4x4kiwi
Well done guys. That,s brilliant. all 54 pages of updates. :)
Enjoy the EV grin.

Would one of your other SI's like to start a car with my Azure Dynamics kit that is for sale viewtopic.php?title=set-of-azure-dynami ... ale&t=3834

Weber and Coulomb's MX-5

Posted: Fri, 04 Oct 2013, 00:05
by jonescg
Mal, give them a break! After all this time not driving EVs, they need at least a good two weeks before starting the next project Image

Weber and Coulomb's MX-5

Posted: Fri, 04 Oct 2013, 00:41
by coulomb
jonescg wrote: ... next project

Next project? This one isn't quite finished! It's finished just enough to be registered. It is running on only one half-pack; we still have two contactor boxes to wire in, a second charger to wire up and bolt in, tidy up the charger inlet wiring, charge up the B half-pack, finally put them in parallel.

It would be nice to get the second half-pack going for Canberra.

Weber and Coulomb's MX-5

Posted: Fri, 04 Oct 2013, 02:04
by weber
Thanks Johny, Woody, Renard, PlanB, BigMouse, CometBoy, AdverseEffects, a4x4kiwi and jonescg. And thanks for all your help and support over the years.

Woody, we'd love to drop in if we can. We will be towing the MX-5 on a trailer to Canberra and back. PM me your address and phone numbers.

We'll have to change our sigs eventually, but she's not finished yet. Not exactly an engineering masterpiece yet. We've done the minimum required to get her registered. This will make it soooo much easier to debug. We are still only running on half the batteries, although the other half is all there with its Anderson connectors dangling, waiting to be wired in parallel.

It's really fairly unpleasant to drive at the moment (now that the novelty of being allowed to drive it at all has worn off) since it seems the firmware in the WaveSculptor has never before been tested in the part-speed-control/part-torque-control mode we use to simulate ICE accelerator-pedal response with regen "engine braking", and it is refusing to allow the induction motor to go into the overspeed or field-weakening region. Since we're currently running on half the voltage we originally designed for, this means we bounce off a rev limit at about 3000 rpm, and if we're asking for more torque by going up a hill or accelerating hard, this barrier can come down as low as 2000 rpm. So you really have to watch the tacho and make far more use of the gears than any EV should have to. Far more use of the gears even than with the original ICE. You really have to plan ahead for gear changes. It's like driving a chugging old diesel.

Tritium have reproduced the problem on their dyno and we will take the MX-5 to them just as soon as the relevant personnel get back from the Darwin to Adelaide World Solar Challenge. Hopefully they can easily figure out how to make it work as expected. But I suspect that's cutting things a bit fine since we want to head off for Canberra early on Thu the 10th of October (2013). So we might have to revert to plain torque-control with no regen for now.

Yeah. Give us a break, Mal. Image

I got the MX-5 registered today. First I took the Safety Certificate, in the Leaf, to the RACQ at Indooroopilly, and obtained a CTP Insurance Certificate so I could legally drive the MX-5 to the Department of Transport office at Sherwood. But I was still as nervous as hell driving there in case something went wrong with the car, or the cops pulled me over. I figured a car with no number plates must stick out like a sore thumb, so I left the toy plates on for the trip, but hoped they were obviously toys up close (with that lightning bolt) so I wouldn't be accused of trying to counterfeit personalised plates.

At the DoT, after a long wait, not one but two young female staff came outside and shook their heads at the strangeness of what they saw when I lifted the bonnet. They asked me a few questions, like did it have a fuel tank or any kind of backup and what sort of range did it have, then they checked the VIN and Mod Plate numbers. Fortunately they didn't ask to see the Engine number as they would have needed an endoscope. They apparently were happy to take the Engine number as read from the Mod Certificate or the Safety Certificate, and the guys who wrote those were happy to take it from a photo.

Back inside the DoT office, as one of the young women nonchalantly slid the number plates across the counter to me I said, with a grin from ear to ear, "I want you to know what a momentous thing you just did. I've been waiting four and a half years for that." Then she was grinning as she told me I had 24 hours to fit the plates. But 24 hours be damned, I went back outside and fitted them there in the inspection bay, so I didn't have to worry about the cops pulling me over on the way home.

Image

Weber and Coulomb's MX-5

Posted: Fri, 04 Oct 2013, 02:23
by coulomb
Adverse Effects wrote: you are going to have to change your tag line in your Sig ...

Yes, I suppose so. Here it is updated partway.

[ Edit: you can see the lie in my avatar photo; the emblem clearly says "mazda", not "ELECTRIC". I think we intend to get an "Electric" emblem of some sort eventually. ]

Weber and Coulomb's MX-5

Posted: Fri, 04 Oct 2013, 04:13
by PlanB
I look like having to drive to Albury Sat 12th, might try to detour via Canberra for a look.
Hey Woody, my place to Mt. Colah 54 km, within Leaf round trip range.

Weber and Coulomb's MX-5

Posted: Sat, 05 Oct 2013, 12:55
by peskanov
Congratulations! I am sure this car was well worth the effort. Your conversion is certainly awesome, especially the BMS part.
In a few months I will have to legalise my project car in Spain, a country with close to zero experience in EV conversions. I am sure is going to be a long ride Image

Weber and Coulomb's MX-5

Posted: Sun, 06 Oct 2013, 15:29
by coulomb
After a marathon day yesterday, all the HazV wiring for the B half-pack is completed. There is still the 12 V wiring for the contactors in the B2 box, then we have the ability to charge and drive with the B half-pack. The A half-pack is fully charged, so after the B half-pack is charged, we can let the packs parallel themselves. (We can choose one or other half-pack or both by undoing an Anderson connector; thanks to Jeff Owen for that suggestion.)

We still need to finish the second charger (needs the hazV DC connector installed and the charger bolted in).

Edit: the B2 contactor box isn't so much a can of worms as a box of orange snakes, all twisting around each other and packed solid in places.

Weber and Coulomb's MX-5

Posted: Sun, 06 Oct 2013, 16:48
by marty11
Yes ! Its registered !   Congrats from marty down south, been following your posts for the journey, and will continue to do so.

Interested when all is going well to see results for real world range , driver satisfaction etc... well done guys Image

Weber and Coulomb's MX-5

Posted: Wed, 09 Oct 2013, 19:30
by Renard
weber wrote: here are the two mad professors celebrating, and toasting absent friends


I have just had a request from my wife, who has been accidentally exposed to all your posts, to pass on her congratulations. She is happy to put faces to the two names which she has heard for so long.
I had to point out to her that you represent two basic units...

Weber and Coulomb's MX-5

Posted: Wed, 09 Oct 2013, 20:32
by weber
Renard wrote:I have just had a request from my wife, who has been accidentally exposed to all your posts, to pass on her congratulations. She is happy to put faces to the two names which she has heard for so long.
I had to point out to her that you represent two basic units...

Thanks, Mrs Renard! And by the way I don't really have cat's ears (or are they devil's horns). That was just my son very cleverly lining up the tops of two chairs behind my head when he took the photo. Actually, I believe him when he says it was an accident. :-)

We're still ridiculously trying to get all the software and electronics working for the second half pack (the "B" pack as we call it). But it doesn't look like it's gonna happen. Software is like that. But we can still drive it on the A-pack to meet up with the trailer and tow vehicle.

So for those who are coming to the Canberra festival: A reminder that we really rushed through a bare minimum of functionality so we could get the MX-5 registered, and therby easier to test, but it is still very much a work in progress and not yet living up to its "lightning" moniker. So I'm afraid there will be no joy rides.

But the good news from Tritium is that they should have a high voltage WaveSculptor for us to try soon after we get back from Canberra. Woohoo!

Weber and Coulomb's MX-5

Posted: Wed, 09 Oct 2013, 22:54
by PlanB
The 700v 150A WS, nice, I can smell the ozone from here!

Weber and Coulomb's MX-5

Posted: Wed, 09 Oct 2013, 23:26
by jonescg
I hope 150 A is the continuous rating. It would be great to see an Australian competitor to the Rinehart or the Sevcon.

Weber and Coulomb's MX-5

Posted: Thu, 10 Oct 2013, 01:44
by Renard
jonescg wrote: I hope 150 A is the continuous rating. It would be great to see an Australian competitor to the Rinehart or the Sevcon.

Continuous?

The motor would be dissipating about 10kW continuously?

Weber and Coulomb's MX-5

Posted: Thu, 10 Oct 2013, 01:59
by jonescg
Yeah that sounds about right - the Rinehart PM150DZ inverter will happily dish out 225 amps at 700 V - At 95% efficiency that's 8 kW of heat to dissipate through the liquid cooling loop.


Weber and Coulomb's MX-5

Posted: Thu, 10 Oct 2013, 02:08
by coulomb
weber wrote: We're still ridiculously trying to get all the software and electronics working for the second half pack (the "B" pack as we call it). But it doesn't look like it's gonna happen.

Ah, but Coulomb the master debugger was having lunch in the vicinity, and dropped by to work around the problem.
Software is like that.
Software is like nothing else in the universe.

It looks like we had two problems:
1) An unknown hardware error on a BMU; replacing a string of 8 BMUs seems to have fixed that. Fortunately, it was in an easy-to-get-to battery box.
2) A weird timing issue, whereby the software uploading new code to the string of BMUs just started running faster than it used to. Alas, I didn't have access to my laptop at the Weber Wigwam, so in place of the proprietary C++ compiler I bodgied up some C code using timing loops. I checked the timing on Weber's DSO.

But we still have mysteries unsolved:
1) What was the hardware error? Will something like it happen again where we can't get ready access?
2) Why did the software suddenly seem to speed up? Actually, the speed-up may possibly be subjective; perhaps we were a bit skinny on a delay which we got away with because the netbook's hard drive was fragged, or some such.
3) Why didn't the TestICal software work when Weber JTAGd it into some BMUs?

But the important thing is that the B half-pack is back to balancing again. We'll run it all night, and attempt to balance it more as we have access to power on the way down. Unfortunately, it probably won't make it, but we can say we gave it our best shot. Initial balances (regardless of top or bottom balancing, with or without a BMS) just take a long time, but hopefully once balanced, we should never have to do it again.

It was satisfying to at least have a work around, and we didn't have to remove any battery boxes (just two accessible battery box lids).

Now to do a rather rushed pack for the trip to Canberra...

Weber and Coulomb's MX-5

Posted: Thu, 10 Oct 2013, 06:55
by weber
Coulomb did indeed perform legendary feats of debugging and coding today (Wednesday) to get the BMS working again on the B-pack without having to remove any battery boxes.

Ya gotta love these guys (Coulomb and Newton (Jeff Owen)). They had already put in ridiculous numbers of hours on the MX-5 in the past week, particularly Coulomb at this stage, because most of the problems were software and electronics. And I did not expect to see either of them today. But they both just happened to be in the area, didn't they, and so they dropped in to see if they could help. You guys are amazing. Thanks.

Weber and Coulomb's MX-5

Posted: Mon, 14 Oct 2013, 02:07
by EV2Go
Congrats guys been a long time coming. Really glad to see those pates on there :)

Weber and Coulomb's MX-5

Posted: Sat, 19 Oct 2013, 04:40
by weber
Coulomb started a new thread to report on our trip taking the MX-5 and other EVs to the 2013 EV Festival and AEVA AGM in Canberra, entitled The Queenslanders' Excellent Adventure

Weber and Coulomb's MX-5

Posted: Wed, 23 Oct 2013, 12:52
by Nevilleh
Well done, guys!
I haven't looked on here for quite a while as I sold my BMW to a very happy new owner and haven't done any EV stuff for some time, but I thought I'd better get an update on the MX-5 and - lo and behold! - I see its registered and driveable. Not to say "finished" because, like mine, it'll never be REALLY finished!
Anyway, my deepest congratulations, I know what its like.

Weber and Coulomb's MX-5

Posted: Mon, 28 Oct 2013, 02:09
by coulomb
Last EV day the B half-pack was fully debugged and finally balanced. The amber LEDs mostly stay on steady, as they used to and are supposed to; they were blinking on and off way too much. This was due to badly soldered bypass resistors on one BMU. We also found and fixed a BMU with high quiescent current (7 mA where it should be <= 1.5 mA); this was due to a bad MOSFET transistor. Well, it went away when we replaced the transistor, and on paper the transistor should be fully off when the gate drive is zero volts. We checked quiescent current as part of our BMU testing, so either we missed this one or it became leaky after testing.

The A half pack, while not fully debugged, has been successfully balanced and charged, so there was no reason not to attempt to parallel the two half packs. Of course we measured the two half packs first; both measured the same to three significant digits (I think it was 386 V). They paralleled without any apparent fuss; another milestone. (Actually driving with the two half packs paralleled would be a bigger milestone, but that will have to wait till after the under boot box is re-installed.) We later realised that we could measure the difference between the two half packs with more precision by measuring the voltage across the contactor that parallels them. The highest we measured was from memory 360 mV.

[ Edit: bypass resistors bad on one BMU only; high quiescent current was in a BMU, not a cell. ]

Weber and Coulomb's MX-5

Posted: Mon, 28 Oct 2013, 03:51
by jonescg
So the contactor could double as a "dump shunt" Image

Weber and Coulomb's MX-5

Posted: Mon, 28 Oct 2013, 13:04
by coulomb
jonescg wrote: So the contactor could double as a "dump shunt" Image

I assume you mean that it can act to connect a full pack across a depleted pack, charging the depleted one from the full one. The resulting uncontrolled surge of current could end up damaging both packs, depending on how high the peak current is, how long it lasts, and how often it is done.

In fact, if we are careless enough to charge one half pack more than the other, then this will indeed happen. It highlights the danger of running paralleled half packs, and we really only intend to run it this way temporarily.