openinverter forum

I'd like an electric motorbike, but I can't afford one. Plus, I want to learn how EVs work. So, here's my conversion project: I've always quite liked the Moto Morini Dart. Despite being 30 years old, I think it looks futuristic. It's small, quite low-powered and covered in fibreglass, which in my mind makes it an ideal candidate for conversion. They're also rare as hens teeth, so my conversion is sure to wind up the purists

Keeping to the spirit of openinverter.org, I'll be documenting progress here and sharing all design files & software. Please weigh in with (helpful) thoughts/opinions/obervations. I have a lot to learn...

Project links:
Blog http://electricdart.net
Repository https://github.com/clanger9/electricdart
Twitter @electric_dart

Great project

Years ago, I bought several examples of the Cagiva Freccia 125 (mostly C9 versions) to use as parts donors for my two Morini Darts. The Darts eventually moved on to new owners, but I still had quite a few Freccia bits. One was converted into an EV;

Some pictures of the work and completed bike;
https://photos.app.goo.gl/YAMNs7HdaeahVTso6

Another is on the bench now;


Personally, I wouldn't do a Dart because of the rarity and the fragile nature of the aging ABS bodywork. Freccias on the other hand were made in much larger numbers and the later C10/C12 bodywork is made of a much tougher plastic.

The dart is nowhere as aerodynamic as you might think. The front mudguard is pretty good, but one of my Darts went off to the USA to be run at Bonneville, setting four class records with the engine I'd put together from spares. It ran slightly faster with the bodywork (other than front mudguard) removed.

Thanks for that video - very interesting!

It's amazing how much motor & battery tech has moved on in recent years. I can go direct drive (i.e. conventional chain) with an 4,000 RPM motor and I'm looking at a potential 35kW from a mere 3.5kg. An 8-12kWh battery comes in under 100kg, which makes everything so much easier...

Out of respect for the Dart's rarity, one of the rules I'm working to is "no cutting or welding any original parts". I'm not a total philistine

Way to go, keep clanging and no cutting. Hope the 35Kw motor works out for you
BTW, thanks for the ride from Dublin last week

So far I have no motor, no battery, no charger and no VCU.
But I do have two of these. Gotta start somewhere!

So, I've been having a few thoughts on different battery/motor/inverter/charger/VCU combinations for my bike. Everything is inter-related, so it's tricky. Here's where my mind is at currently:

• Battery: 8x battery packs from a 40kWh Leaf. This will give me 12.8kWh at 128V from 70kg

• Motor: MP15470 drone outrunner. Claimed specs are outrageously optimistic, but in theory this will give me ~35kW. It's tiny - only 3.5kg! No idea how you're supposed to cool it. Very easy to mount & fit chain sprocket. Surprisingly low speed/high torque. It appears the 27KV windings give a mere 3,240RPM at 120V: so 85MPH with standard chain gearing. Feels about right for a road bike.

• Charger: Elcon TC HK-J 3.3kW. Small, but expensive & slow. I'd really like to have rapid DC charging, but that's not possible at 128V DC...

• Inverter: Curtis Model 1239E. Expensive and bulky, but gives me a nice programmable VCU as part of the deal

Thoughts?

Please be kind, I'm still figuring this stuff out...

Been doing a bit of reverse engineering on a Peugeot throttle position sensor. These are cable operated, so ideal for a bike throttle. They're also cheap (only £16 new). Part number is 9643365680. Google for "Peugeot 306 HDi throttle position sensor" Rather than a true potentiometer throttle, it appears these use a contactless sensor to output a pair of analogue signals.
Using a 5V supply and a 1kΩ pull-up resistor, I got the following readings:
0% throttle white: 0.419V yellow: 0.209V
100% throttle white: 3.780V yellow: 1.871V I may use two: one on the throttle cable and one on the clutch cable (for regen control )

clanger9 wrote: ↑Wed Mar 25, 2020 10:47 pm "Peugeot 306 HDi throttle position sensor"

Good choice of throttle pot.
I've had one on my 1900TDI diesel bike for about three years.

Spot on!
Did you need to lighten the return spring, or is it OK on a twist-grip?

I cant actuality remember. I know I took it apart but can't remember if I messed with the spring in it, but there's a chance it had two springs and I removed one. Sorry to be vague but I was sleep deprived at the time and I can't find any photos I took of that part of the project.

I know I tried several different cable operated ones before settling on this one. The others such as a VW thrpottle pot had too strong return spring for a twist grip and no way of dismantling it for adjustement.

I also crudely fitted a microswitch to override the throttle pot at rest and give a separate digital input for throttle at rest ie a redundancy in throttle safety.

Everyone likes a bit of wee-oo-wee-oo, so I had a go at making an Arduino motor controller to work with the Peugeot HDi throttle position sensor.


It interesting to see how the two analogue signals make the controller fault tolerant: both signals are different, meaning you can detect open circuits, short circuits, failed sensors, etc. I did a bit of signal smoothing and averaging as well as some basic error handling. Presumably this is how it's done on real controllers?

You can find my embarrassing source code on GitHub here: https://github.com/clanger9/electricdar ... sition.ino

Thoughts & suggestions very welcome. I'm particularly keen to hear if you find ways my code could fail with the throttle stuck open.
The most obvious failure mode is if the Arduino crashes...

Today's rethink...

I've had a go at measuring up the space for the various components and it appears that 40kWh Leaf modules are NOT the way to go. They're just an awkward shape for a bike and don't really fit anywhere. I think what I need is a battery pack made up of discrete cells. Which brings me to my latest idea:

2x battery modules from a 120Ah BMW i3.

This will give me 10.6kWh at from a mere 56kg - but only 90V. The beauty of the i3 pack is that it's a 12s arrangement, so the cells form a single string.
The advantage of using complete modules electrically is I think I can retain the existing BMW BMS (while unpacking and spreading the cells out according to where I can find space).

Can I get away with such a low voltage? Unfortunately a standard 27KV motor would spin to only 2,400RPM at 90V, which only gives a a top speed of 60MPH (using the standard 14/39 sprockets). I could change the front sprocket to 23 teeth and that would give me 100MPH. However, the motor is also available with 55KV windings, which presumably would give a theoretical 120MPH on standard gearing. Is this a better approach?

Thoughts?

clanger9 wrote: ↑Wed Apr 01, 2020 4:08 pm Can I get away with such a low voltage?

I have a quad bike that uses a 72V battery and has a top speed ~60mph It should be easy to achieve highway speeds with some simple gearing (chain and sprocket on a bike).

Thanks, Kevin. A 23 tooth front sprocket isn't too extreme, and that should gear it up nicely...

Nothing wrong with creating your own pack from 18650's. Exactly the shape and location you want and you can easily balance voltage and power. nkon.nl is a decent source for them and keenlab.de sells an amazing spot welder for this task.

Hmm.

I admit I was spooked by the idea of using 18650s given a) the need to design/build my own BMS and b) their tendency to go off like firecrackers if you get something wrong...

On the other hand, 18650s would give me all the power & performance I need in a flexible, lightweight package. I like your thinking.

Something to ponder.

Gen3 inverter measurements: 22x22x35cm + brackets, cooling ports etc here and there a few cm.

I used to have a bottomless source of 18650s. But it dried up last year.

I have enough of them to build my electric car, but probably not enough of them anymore to also finish my electric motorbike. Otherwise I would be done.

You can buy clip together 4x5 grids for 18650s, and then start to plan out how many and where you can fit them on your frame.



I wouldn't be too worried about the lack of BMS. Especially if it's only 100v. Build it capacity-matched, throw on a Bat-Bridge as a warning light, and then manually balance once a season.

Unrelated, circa 1969 forklift SCR speed controller spinning the motor up:

More ideas from today's workshop session. It appears BMW i3 94Ah cells would be ideal!
https://pushevs.com/2018/04/05/samsung- ... fications/

30 cells would give me 10.5kWh at 111V - in only 60kg. Here's an approximation of 6 of them stacked together: Even so, it'll be a bloody tight fit! The battery box and cabling will be {ahem} interesting... Unfortunately, there isn't space for a Prius 3 inverter, no matter what I try.

So I'll need an Elcon TC HK-J 3.3kW charger plus either a (pricey, closed source) Curtis Model 1239E inverter or a (scarily) light & compact Chinese drone inverter. Specs are impressive, but it looks like it'd destroy itself if you looked at it funny. https://fliermodel.en.alibaba.com/produ ... A_esc.html

If I go with the Chinese 500A thing, it'll fit under the seat, which is ideal for air cooling. In my diagram, I can swap the position of batteries 1 & 2 with the inverter.

Anyone tried one?

clanger9 wrote: ↑Wed Apr 08, 2020 4:34 pm It appears BMW i3 94Ah cells would be ideal!

Just a heads up that the BMW i3 packs can not be split because they are bonded together in the aluminium frame with some sort of structural glue.
If you go looking several people have tried failed.
Sorry for being a party pooper.

Thanks for the heads-up. Of course, I do love a challenge.

Aren’t they just standard Samsung SDI cells? Maybe I can get individual cells from another source...

Probably, I've never seen one up close to compare. However the i3 packs contain 12 cells firmly bonded together in the aluminium chassis.

There's some good youtube videos of the factory assembly of them which show both the glue going on as well as the plasma / electrostatic preparation of the surface got best adhesion. It's worth watching for the laser welding bit.


It does look like the glue goes down onto a plastic film on the cell case but I couldn’t find anyone who has successfully separated them. I found a couple of forum posts of people who had tried and only succeed in damaging the cells trying. If you have any ideas I'd be up for discussing them but unfortunately don’t have the funds to risk my pack or buy a spare one for experimentation. As they'd be ideal for repacking into a small inner city bike rather than the mile muncher I have planned.

They also have aluminium terminals and wire which will be fun to join up when I get that far with mine

Great video, thanks Mouse. That explains why they're so difficult to get apart.
That laser welding is truly amazing!

Individual cells available here, but I'll need to wait for the price to drop somewhat...
https://shop.gwl.eu/Samsung-SDI94-NCM/L ... g-NCM.html

Well, here's the answer: apparently those Chinese "500A" inverters are pure garbage.



Great video, by the way. Time for another rethink!