[Driving] Homebuilt Locost 7 - from Motorcycle engine to Leaf power!

[Zieg]

Hi all,

As I am getting further in my project and coming here for more help/info, I thought I should also explain my build and present my work to date.


I built this car entirely on my own and completed it in 2019. It is based on a Lotus/Caterham 7, following the original 'book' dimensions for such replicas. That is to say, it's the smallest chassis size commonly used. Many are widened, stretched vertically and longitudinally, but not this one. I built it primarily for Autocross, but I also wanted it to be street legal. That meant some extra weight due to things like headlights and tail lights, a passenger seat, and a 16ga steel floor pan for safety. All told, with fluids and no driver it came in at a hair over 1200 lbs (550 kg). It's been fun to drive with the motorcycle engine (2004 Yamaha R1), but the bike engine is not so great for street use, and even at autocross its lack of low end torque made things difficult. Due to the high redline, I could go 110 KPH in first gear, but in 6th gear on the highway it was still over 5000 RPM - that gets old pretty quick. Also, even after repacking the muffler after every other event, I was still always on the edge of being over the noise limits. So, I asked myself, how can I make more low end torque and reduce noise at the same time?

Here's how the car looked at the end of last season:

I also designed the suspension myself (I'm a Mechanical Engineering Technologist by trade), so I already had CAD models to work with. I decided to use the Leaf motor and gen 2 Chevy Volt battery modules, and at first I looked at mounting the Leaf motor in the font and cramming the batteries around it, and in place of the fuel tank in the rear. That didn't look great.

Eventually, I decided my best course of action was to do away with the heavy (60 kg?) rear axle and use the Leaf motor together with its gearbox, with a new independent rear suspension. I'll need to shorten the half shafts, but that looks easy enough. I also had to mount the motor backwards. I know that people using unmodified inverters, this also requires swapping two phases to the motor and to the resolver. I'm not sure how they deal with the resolver offset being effectively backwards, but since I am using an Openinverter board I don't believe I will have to worry about any of that.

I've also modified the reducer for reverse operation, but only time will tell if it's enough. It's from I believe a 2015 model, so it has open bearings. I bought a BMW oil pump from eBay and made an adapter block for it. The plan is to pull oil from the lower drain plug hole and pump it back up to the upper reservoir (where it would be splashed to when running 'forward'). If anyone is interested I'd be happy to provide more details on this, but basically I drilled some holes in that reservoir and lightly ground some channels into the case in hopes of directing oil flow to the bearings that would not be splash lubricated in reverse operation.

Oil pump on block:

Example of holes in reservoir and grooves. The hope is that the flow from the pump and general chaos of oil flying around is enough to keep these bearings lubricated. I may switch to sealed bearings if I can find them, but for now they have proven pretty elusive. Thanks to another member here I have the part numbers for the two on the outer side of the gearbox, but that company doesn't list a bearing for the inside, so I'm not sure if the size was changed or not. It's very hard to find sealed bearings that can handle 10,000 RPM on the input shaft!

Next steps for me are to figure out exactly how much of the Chevy Volt battery I want to use, and how to fit them in the car. Post about all that to follow below...

[Zieg]

Okay, so for fitting the battery, I've been playing Battery Tetris for a while and have a few options. Some of them I like, others not so much. Given that I really don't need a lot of range, I would like to keep the weight of the battery to a minimum - but I also need enough voltage to reach a top speed of 130 KPH or so.

I'm currently off work on parental leave due to my wife and I having a baby late last year. While I'm off, I don't have access to Autodesk Inventor so I have been using Onshape instead. If anyone is looking for another option for free 3d CAD, it might be worth a look. I like that I can open and work on things with my tablet or my PC. The only catch is that anything you make is publicly available for anyone else to copy into their own workspace.

All that said, here's what I am looking at. The yellow outline is roughly the space I have to work with.

And for reference here is the pack with the corresponding cell groups numbered:

I have a steering rack right at the front that might be touching module 3 as it's shown here - it'd definitely be close either way. I also have the steering shaft running down the left side of module 1 but shouldn't be an issue as it runs higher than the battery in that area. The way I am showing module 2 will reduce passenger legroom by a bit, but I don't often have passengers, so if that's what I have to do, I might be okay with that. If it looks like I can do away with one group of cells (going down to 84s from 96s) that would make life a bit easier. Have to run more calcs.

Also on my todo list is work out a speedometer signal. I was using the motorcycle gauge cluster, but decided to swap that out with something more generic. I have a Dilithium BMS setup and got the LCD screen for all that fun stuff, but I still need a speedometer, odometer, and some indicator lights to keep it street legal. With that in mind, I bought a Koso EX-02s universal motorcycle cluster. It comes with a hall sensor and some magnets meant to be attached to the bike's front wheel. I have bench tested it with an Arduino outputting a Tone() and thankfully the square wave fools it nicely. Think I'm going to grab a Teensy and have it pull motor RPM over canbus, then translate it to a tone to feed the speedometer. Backup plan is to use one of the Leaf hubs that I already have, since it has a magnetic tone ring already for the ABS. All I'd need is the matching sensor and a bit more time on my mill.

This is the aforementioned Leaf hub on the new uprights I designed. I pressed the OEM studs out and replaced them with 1/2" studs to match what was on the axle I removed. The PCD was the same, thankfully, so no need to do anything with adapters or mismatched wheels. I even managed to get them to accept the Ford Mustang brakes that were already on the car.

[Zieg]

And as a side note, because I'm a 'necessity is the mother of invention' kind of guy, here is the ambulance that I use for transporting the car to and from the track. Trailering it was not fun and I do not have room to park a truck and a trailer, so I sold both and bought the ambulance. That also meant I didn't have to use the truck to commute to work and could replace it with a more sensible daily driver.

I stripped the interior and made a set of articulating ramps so I can winch the car in and then lower it down. Amazingly, the net weight of the ambulance plus car and ramps almost exactly matched the weight of the ambulance before I started gutting the interior. I also built a little sled that can ride on the ramps to make hauling things around easier. I've used this thing to move a lot of furniture for friends and family...

Here's an older video showing the car being loaded. I have since changed the front tow hook so I don't have to remove the nose cone anymore..



And for anyone interested, this is an Autocross run (Headphone warning - LOUD!):

[Pete9008]

Liking this a lot, especially those machined aluminium uprights!

I planned to do the same as you with the ambulance, got as far as buying a 3-axle Rohill minibus with air suspension to convert (accommodation in the front, car in the back) but got cold feet due to UK conversion/MOT regulation complexities, sold it and bought a trailer instead - still not sure which was the right descision. Nice to see a similar project completed

Looking forward to updates!

Edit - have you considered sacrificing the passenger seat completely and putting pack 3 there? As a track car the bettery weight would both balance the drivers weight and keep the weight close to the c of g for a low moment of inertia. The shift in weight to the back might also benefit getting a rwd car off the line too.

[Zieg]

Oh cool! Yes, MOT stuff can be a pain from what I have heard. These cars are way more common over there so a lot of my information (and a handful of parts) came straight from the UK. Here, I still did have an interesting conversation with my insurance company. They won't classify it as a motorhome unless it has a sleeping space and either a fridge or stove. I said I could put a hammock and a hot plate in it, but apparently that didn't fit the spirit of the rule.

And yeah I have considered deleting the passenger seat, but only as a last resort. It's nice to give the occasional ridealong, and I already knew it wouldn't be truly competitive while also being street legal. So, instead of building a formula-style car or A-mod, I built a Locost partly in so I could share the experience.

[Pete9008]

I seem to remember on mine I could convert to a camper no problem as long as there wasn't space for a car in the back. The moment it had a garage it would have problems going through the MOT. It was also right on the limit weight wise. I looked at ambulances but the UK ones ar3 too small for my car.

I have a soft spot for the seven style cars, used to have a Sylva Striker and still have a Sylva Fury (although it must be over 10years since it moved under its own power). Keep thinking about an EV conversion on it.

What about the transmission tunnel, could that be widened to fit a pack?

[Zieg]

Oh I'll bet the Fury would be a ton of fun with an electric conversion!

Trans tunnel is only 4" wide on the inside (6" on the outside due to the 1" square tube). There's only 17" of width on the passenger side for the seat, so widening it at all in the seat area would mean deleting the seat. With the way I have module 2 shown, it would already mean narrowing passenger legroom to 12" or less from about the knee forward. I could also try and cheat an inch or two into the driver side footwell since I don't need room for a clutch pedal anymore. That would be a fair bit more welding, though.

I think if I can drop the 12s group off module 3, I might be in good shape. Also on the todo list is to try laying them sideways and see what that looks like. Even staggering them vertically might be an option. Not great for the vertical CG, but it's so damn low anyway that it might not be the end of the world.

[Pete9008]

The way you have drawn it is making a lot of sense given the constraints. Don't like the idea of vertical stacking at all, the low c of g is one of the things that makes these cars handle so well.

Thinking about it the conversion to IRS wont do the unsprung weight any harm either, on a lightweight car a live axle isn't ideal (although having said that the tracks look pretty smooth). My Fury is an early live axle car and it's something I would like to change.

[Zieg]

Yup, my thoughts exactly. I was often asked why I went with a Ford 7.5" axle instead of the more common 8.8. My response was that the 7.5 was already too heavy for the car, and if I managed to break a small pickup truck axle with a motorcycle engine, well, I would be very impressed!

[bexander]

I too very much like this type of car, so simple and just bare miniumum of what you need. It crossed my mind many timed to build my own but have never got to it.

Not shure if this is what you are implying but if you remove a couple of cells from the #3 module and place it as far back as possible in the old engine bay it looks like it will fit sideways and then #1 and #2 also sidways in front of #3. Hope this makes sens? Tried edit your picture in GIMP but didn't manage to do what I wanted.

If you where to mount the Leaf powertrain in its original position (motor in front of diff), how far back would you have to move the rear axel (tires) to make it work? Just an idea since you anyway are building new suspension.

[Ruudi S]

Nice and simple car. I have one copy of this kind of car in my garage for ICE wiring work right now. It's a shame but it will not be an ev but maybe in the future, when I show to the car owner your electrified version driving. It's good that you chose Leaf motor and install it in the correct place. When you get it running you bring the Locost cars to this century maybe even start a new class?
Will you remove the gearbox tunnel? Then we can in the future make these cars with central seating positions and maybe put two modules to both sides of the driver?

[Pete9008]

Been thinking about this and it wouldn't surprise me if the electric version has a distinct advantage, no gears, instant throttle response, full torque from 0. Will probably be a little heavier but usable power to weight may actually be better.

How do the rules affect EVs, are they in a separate class or is there some form of handicapping system?

Also wondering what your plans are for battery current limits, what's the Volt battery capable of, are you going to keep the original battery liquid cooling circuits?

[SuperV8]

Okay, so for fitting the battery, I've been playing Battery Tetris for a while and have a few options. Some of them I like, others not so much. Given that I really don't need a lot of range, I would like to keep the weight of the battery to a minimum - but I also need enough voltage to reach a top speed of 130 KPH or so.

I'm currently off work on parental leave due to my wife and I having a baby late last year. While I'm off, I don't have access to Autodesk Inventor so I have been using Onshape instead. If anyone is looking for another option for free 3d CAD, it might be worth a look. I like that I can open and work on things with my tablet or my PC. The only catch is that anything you make is publicly available for anyone else to copy into their own workspace.

All that said, here's what I am looking at. The yellow outline is roughly the space I have to work with.
battery space.JPG

And for reference here is the pack with the corresponding cell groups numbered:
Volt battery pack.JPG

I have a steering rack right at the front that might be touching module 3 as it's shown here - it'd definitely be close either way. I also have the steering shaft running down the left side of module 1 but shouldn't be an issue as it runs higher than the battery in that area. The way I am showing module 2 will reduce passenger legroom by a bit, but I don't often have passengers, so if that's what I have to do, I might be okay with that. If it looks like I can do away with one group of cells (going down to 84s from 96s) that would make life a bit easier. Have to run more calcs.

Can you fit one or two modules behind the rear axle?
I have a Dax Rush and this is where my fuel tank is - quite a decent space.

[Zieg]

Oh wow, thank you for the kind words and suggestions everyone!

Not shure if this is what you are implying but if you remove a couple of cells from the #3 module and place it as far back as possible in the old engine bay it looks like it will fit sideways and then #1 and #2 also sidways in front of #3. Hope this makes sens? Tried edit your picture in GIMP but didn't manage to do what I wanted.

Hmm, that's a good idea. I think I can get away with ditching one 12s group. The orange one seen here is that module 3 with one 12s group removed (so just 2x 16s). I do need to allow for some space around the ends for the water cooling connections, and around the long sides for the holddown clamps. Then I plan to form some sheet metal into a protective box over the whole thing.

If you where to mount the Leaf powertrain in its original position (motor in front of diff), how far back would you have to move the rear axel (tires) to make it work? Just an idea since you anyway are building new suspension.

Ah, I can't show that right now without access to my other CAD software, but it was bad. I can't really move the axle more than maybe an inch or it causes a bit of a chain reaction with the wheelbase changing too much, fitment of the fenders over the rear of the car etc. I did test fit (in CAD) the Leaf transaxle facing forward, but the axle half shafts were swept forward at an uncomfortable angle. I know the outer CV joints are meant to work at extreme angles, but the inboard ones are just a simple 3 dogbone arrangement and I think it'd have all kinds of vibration and wear issues. But, I think addressing the oiling concern was easy enough. Hopefully it proves sufficient.

Will you remove the gearbox tunnel? Then we can in the future make these cars with central seating positions and maybe put two modules to both sides of the driver?

No, the tunnel adds a whole lot of strength to the chassis through the centre section. I also need it to run the HV wiring, some control wiring, coolant and brake lines etc.

How do the rules affect EVs, are they in a separate class or is there some form of handicapping system?

Ha, well, that's an interesting question. There are different classes based on the car and the modifications done. There is an EV class but it's for production cars only. I was in the D Mod class, but after the conversion might not really fit in that class anymore. The good news is the car was fast, yes, but not truly competitive in D Mod. The cars that are competitive have huge front and rear wings, no passenger seat, very small diameter racing slicks, and so on. My last car was a Mazda RX8 which was in a stock class and very competitive, but my days of placing top of the class died when I sold that car, haha.

Also, on that note, unfortunately for this season (and last), electric vehicles are not permitted at any motorsports events in Western Canada. It looks like the insurance company got scared and dropped a blanket ban, which is most unfortunate. I am hoping this gets lifted before the season is over, else I will be stuck driving on the street, and maybe the odd day at a private track that I may or may not be able to get onto through a friend.

Here's the message that went out to all clubs. It bugs me that a hybrid with more battery on board than my car would still be allowed, simply because it's not 'fully electric'. Maybe I should mount a 1cc model airplane engine with a propellor on the back and call it a hybrid range extender?

Also wondering what your plans are for battery current limits, what's the Volt battery capable of, are you going to keep the original battery liquid cooling circuits?

Well I have heard of people pulling 1000 amps from the gen 1 packs (admittedly those were arranged in 3P and mine are 2P). I kind of figured that if I wanted to make about 120kW then I'd be pulling somewhere around 350-400 amps. Here's the latest from my testing in the simulation software. I was meaning to ask if the Iq line in the controller currents chart represented actual current from the battery. If so, it looks like peak draw could be closer to 500 amps, although that's also while developing more torque than I could possibly put to the ground. In fact, if I think I am making more than twice as much torque as I could actually use in the low end, I might want to hit up Bratitude for a set of his 4.35:1 gears when they're ready.

I will be maintaining the liquid cooling. I grabbed an electric booster pump from Davies Craig and a reservoir from a Chevy Bolt. Planning to run the loop like this:

Can you fit one or two modules behind the rear axle?
I have a Dax Rush and this is where my fuel tank is - quite a decent space.

Yeah that's where my tank was too, but with the Leaf parts and IRS mounting points I don't think so. A mild concern would be balance of the car, and also I was strongly discouraged from having battery modules in both ends of the car. Something about HV safety - would need another contactor to break one or both ends of the connection between the front and rear cells, plus making sure that doing so wouldn't fry the BMS?

[Alibro]

Interesting project but have you considered using a hybrid car battery? Maybe a BMW one like Damien uses?
You'd get the same voltage (so speed) but would take less space.

[Pete9008]

Abdsolutely crazy on the motorsports ban, especially since hybrids still seem ok!! Hope it gets sorted.

New battery arrangement looks good.

Afraid Iq and battery current are two different things even though they can sometimes look similar. Afraid the only way at the moment is to work it out from power (I = P/batt volts). Best accuracy is to check the efficiency box which then shows electrical power into the motor as well as shaft power and use that for the calculation. Thinking about it probably ought to add a graph for dc current at some point.

[robertwa]

Interesting build! I just sold my homebuilt bike engined locost a few months ago to start my EV build. Great platform for a conversion!

[Zieg]

Not much of an update, but I have been delving into the idea of different batteries (Thanks Alibro for planting the idea of the BMW cells).

It looks like either Pacifica or BMW hybrid cells would fit well in my engine bay, negating the need to cut part of the frame out, and vastly simplifying the battery enclosure I'd need to build. So I have decided I want to sell my Volt battery and look for either a Pacifica battery or a 12kWh BMW pack (if I can satisfy myself that I should be ok to draw up to 140kW without harming either the cells or the wiring/connectors).

One benefit to the Pacifica pack is that it already has water cooled plates that the batteries sit on, but they are a bit bigger and of course a bit heavier. In either case I would probably look to stack the modules, which I think will still be okay in terms of CG. The Pacifica batteries will shift the CG a bit further forward in the car, but again considering how far back it was with the ICE components, that may not be a bad thing.

Pacifica modules...

BMW... Possibly a bit too close to the frame rails for comfort, may turn them 90 degrees to match how I'm showing the Pacifica modules.

[Pete9008]

Both look like they fit better than the Volt cells

140kW might be possible with the BMW. If you assume a 360V pack voltage and 20V of sag then you need 140,000/340 = 412A which is around 12C. Feels a bit high but for brief periods maybe?? Even half a 70 sec run is still well outside the normal pulse discharge time though and if you are using regen that will also heat the cells (although regen does increase the chance of locking the rear wheels so not sure I would on track). The only way to tell is to try it while carefully monitoring everything. I'm planning to limit at 100kW but then that's all my motor can do.

If you find evidence of the BMW packs being used at higher currents please post it as it would be great to find out what the limit is.

[Alibro]

What batteries was Damien using when he pulled 224hp in his BMW E39?
I had a vary quick Google and found this https://www.encycarpedia.com/bmw/19-330e-saloon
According to the website above the motor is 83kW max but you can be sure BMW will have a massive margin in that so the batteries will easily cope with much more/

[Pete9008]

Battery details are in a video somewhere, possibly the dyno video? Seem to remember it's an oldish Leaf pack in parallel with something newer (phev pack?).

I agree that there will be spare but how much? Plus the 83kW is short duration peak, so probably for just long enough to get to 60 (well under 10sec) which is different to working them hard for a minute. Would love a definitive answer on this!

Also interesting that they quote 11.15kW net out of a 12kW pack, that's not leaving much reserve at each end.

[Zieg]

Yeah I've watched Damien's videos but couldn't find any instances of a WOT pull on the BMW packs on their own (before he added the other cells in parallel). It may also come down to a question of availability - the Pacifica batteries seem a bit more common around here (and aren't subject to wreckers jacking the price up on anything that has a high end manufacturer's badge on it). I don't plan to use regen on track for both the heating and brake balancing reasons.

I also haven't really seen much info on the Pacifica water cooling plates being used in conversions, but I would be planning to do just that. If I do go that route, it would be nice to be able to provide more pictures and measurements of the cold plates for others to benefit from.

And coming soon: My Accelerator Pedal Position Sensor mount. Really annoying that Tilton has a bracket for mounting a hall sensor to their pedal boxes, but Wilwood does not. So I somewhat ripped off the design and am making my own. 3d printed a couple prototypes to refine the concept and enlisted the help of a friend with a basement CNC machine to bang out the basic shapes. He wasn't sure he could get a nice round hole for the bearings though, so I'll bore those out on my manual mill once I get my hands on 'em.

[Alibro]

That looks like something Superfast Matt's besties "Send Cut Send" could do very accurately.

[Zieg]

Yeah, you're right. I originally asked my friend because of the machined fit I need for the bearings, before he told me he wasn't sure how well that would turn out on his machine.

I will probably be using them for other stuff though, including my battery enclosure. My combat robotics club was actually sponsored by them last year and they made a bunch of parts for our arena.

[Pete9008]

You need some speed holes adding in that bracket