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

[johu]

Also verify that you get equal torque response in forward and reverse. That kind of indicates a correct syncofs

[Zieg]

Okay, I did not have the right syncofs. Everything is fine now, but I did have a rough few minutes getting it sorted out.

It was hard to tell with the car in the air but on the ground it was obvious, the torque response was very different in either direction. I tried a small change first and that felt normal, so I started driving but the first time I tried to accelerate slightly hard the car bucked and lost power. I had to cycle power to the inverter to get moving again. That was on a road with no place to pull over (stupid idea in hindsight) so I kept going, but the car tried to accelerate on its own again. This time I hit the brakes and it cut out again. I was not logging, but I wonder if it was tripping due to overcurrent? My one concern is that if it did trip and opened the contactors, could I have damaged the capacitor in the inverter? The car seems fine after adjusting the syncofs again, but I don't know how I would tell if the capacitor was blown.

Anyway, I finally pulled over to a safe place and started playing with the syncofs again, this time I tested by actually driving the car forward and backward until I found the sweet spot. Then I kept driving and it felt great. It pulled hard and there was no more uncommanded acceleration at higher speeds.

I even found a spot to try accelerating hard from a stop and discovered it's possible to spin the rear wheels. I can also tell from the video that the car is transferring a lot of weight to the back when accelerating this hard. I do think I am going to have to go with stiffer springs!

Lots of wind noise again, turn sound down. I think I was near 100% until the tires spun, then I was probably 75%...

[Zieg]

Wasn't able to go for another drive for a while because of work and weather, but I managed to sneak out for a bit today. I went to a nearby unmanned scale to check the car weight against the bathroom scales at home, and this scale said I was about 760kg with a 42%/58% front/rear balance. I'm hoping that's reading a bit high, because it would mean the car gained about 225 lbs of weight over when it was ICE, and I really can't see that. At home it was almost 700kg even with me inside. I did have some tools and such with me, but certainly not THAT much. The good news is, regardless of the absolute values, the weight distribution is better than it was with the ICE, so I'm happy. Going to try and get it weighed by a known calibrated scale when I have the chance.

Also, interesting anecdote, while coasting downhill at around 70 kph, I noticed that I was maintaining speed while completely off the accelerator. I still don't have any regen active, but it looks like on this particular hill, at that particular speed, wind resistance was enough to keep my speed constant. Google says it's about a 3.4% grade. Tough to tell the slope from this screenshot but here it is anyway:

This week I am planning to stiffen up my suspension a bit by changing the angle of the front shocks and new springs all around. I also need to replace some of my HV wire. I used 600v 2/0 welding wire in an orange mesh sleeve for my battery-inverter connection, but the SCCA rulebook says it has to be double insulated and shielded wire, so I had to spend almost $500 CAD on some 2/0 shielded wire from EV West. Ugh.

The other thing I need to do is replace the wiring to the charger and DC/DC converter. The rules state that 14awg and larger HV wire MUST be double insulated and shielded, regardless of how much current it carries. I used 12awg wire for the DCDC and charger, even though each set of wires sees 10 amps or less. So, I am just replacing it with regular 16awg wire. lol.

[jrbe]

I even found a spot to try accelerating hard from a stop and discovered it's possible to spin the rear wheels. I can also tell from the video that the car is transferring a lot of weight to the back when accelerating this hard. I do think I am going to have to go with stiffer springs!

If you're willing, do you have room to rework the rear suspension to add in some / more anti-squat?

[Zieg]

Hm, good point. It does have some antisquat built in, but I don't recall how much at the moment. If I wanted more I could just relocate the front brackets (or make them adjustable somehow), wouldn't be too hard. I'll see how it is with the new springs first though. What's on there now are just what I had laying around (275lb/in). I have a pair of 325# and a pair of 350# in the mail. Probably going to put the 350# on the rear first, then try moving the 275# to the front and see how that feels. Should put me in the range of 2.0-2.2 Hz depending on which scale is correct.

[EVSwap]

So cool, it definitely has way more torque than the motorcycle engine had off the line!

[Zieg]

Finally got the car out for more testing today. Now that I had the correct springs in place and new front shock mounting points, I really took my time with a proper wheel alignment. That turned out to be quite tedious, but it tracks straight and feels way more stable at speed so I'm happy. I 3d printed a handful of jigs and tools to help, and now that I've got a process figured out it should be faster next time.

I also decided to try a public charging station just to make sure it works. This one was free, but I still had to download the app. Thankfully it was pretty simple and worked as expected! I went inside long enough to buy a drink and when I came out the car had drawn a crowd, lol. So I spent the next 30 minutes answering questions, which was neat.

I still don't have a proper acceleration test on a full charge, but I did a quick test at 380v and noticed the car hooked up a lot better. I still rolled into it from a stop but didn't get wheel spin. It was also a much warmer day today (25c and full sun) so the tires and pavement were hotter. There's going to be a special 1/8 mile drag racing event held on a nearby runway later in the summer and I'd really like to get a few passes in if they'll let me. It's supposed to be 'heads up' racing with no timing, but if I'm able to run I'll probably buy a gps/accelerometer performance meter to see what my 0-60 is. I also have a 10hz gps and some accelerometer from an abandoned diy data logger project, maybe I'll see if I can get that finished too...

Edit: My gopro turned off for it but I was able to pull the dash cam footage of my 380v acceleration test. The dashcam GPS is slow to update and seems to do some averaging with the last reading, but at steady speed it does agree with the speedometer. It was recording at 29.97 fps and the speedometer seems to update every 10 frames. By counting frames from the first movement to 5 frames after the display refreshed to 99kph, I get 4.5 seconds exactly. The built in accelerometer registered about 1.3g up to 40kph which I think is around where field weakening starts. If this is all pretty close to correct then I'm going to be very happy. I have no idea what my 0-100kph was with the bike engine but I'm certain it wasn't even close, especially with how much I had to slip the clutch.

[Zieg]

Okay, I'm working on trying to fine tune things now. Eventually I'd like to get the thing on a dyno but not there yet. I have a copy of Pete's FOC calculator software up and running. Got a locked rotor datalog to find my Rs, but when I tried to do a driving datalog all 6 logs came back blank (columns for the parameters are there, but all the cells are empty except the time stamp). Might need to play with the 'samples per line' setting?

Anyway, in the meantime, I managed to get a good acceleration test where I pinned it from a standstill, got traction and did 0-60mph in 4.5 seconds (counting from the first video frame showing movement to the first video frame where the Gopro's 10Hz GPS shows 60mph). I'm pretty happy with that.

I then tried changing my throtcur from 5.5 to 6.0 A/%. I did another acceleration test and this time it was just a huge burnout. While in the range of peak power I went from pulling about 110kW to 126kW (according to my BMS). Checking the simulator, it seems like I could set throtcur to like 8.0 and that would put me around 130kW. But obviously now I have more torque than traction before getting into field weakening, so my question is this: Is there a way to limit the torque at lower speeds but not at higher speeds? That way I'd get a bit more top end power without having too much bottom end. I'm thinking maybe I can set idcmax to limit the low end power and then set throtcur to something high (maybe as high as 8)? Not sure where to ask this question, so I thought I'd try here first and then maybe go bump the FOC simulator thread if need be.

Here's a video of my successful acceleration test



And I also did a walkaround tour of the car, which got longer than it should be, but skipping around or watching it at 1.5-2.0x speed helps.



And finally, a screengrab from my cruise (while thinking i was getting a good datalog - d'oh!). I was musing to myself that all the guys with flashy expensive cars were probably downtown making noise to turn heads, while I was silently gliding through the coutryside. A few more curves in the road would have been nice, but otherwise it was paradise for me.

[johu]

Sounds exciting

Idcmax won't help as dc current increases with speed.
you could try playing with accelmax and accelflt. This is the acceleration rate limiter. accelflt is a filter between speed and rate limiter. The higher, the smoother but also slower reacting. Must be integer

[Zieg]

Oh, cool! I'll try that, thanks. Should be pretty easy to calculate a reasonable limit, too.

[Zieg]

So I've been trying to drive the car in various conditions every week over the summer, and I have to say it's SO much more fun than it was with the bike engine. As a result I'm driving it way more often, which is great.

I tried to register for a drag racing event happening in September, but they told me they can't allow EVs because of insurance. Sigh.

Over the winter I'm going to take a few things apart and ensure the car is fully compliant with the SCCA EV rulebook so that next year I could take it to Montana or British Columbia to run in some of their events.

I'll have to go through the rules with a fine toothed comb, but I already know I have to replace the HV wiring to the inverter with shielded wire, which I already have. I'll also need to replace the wiring to the charger and DCDC for the same reason. I also need an isolation monitor, which is purchased but not installed yet.

The funny thing about the dcdc wiring is that since I used 14 ga wire, it has to be shielded. But it's only got a 10 amp fuse so I could just replace it with 16 ga wire and then it wouldn't need to be shielded. Lol.

The charger is a little more complicated. Looks like I'd have to disassemble it to replace its wiring. But it's only 3kW and I'd been wanting to upgrade anyway, so I got my hands on a 10kW Tesla gen2 unit. I'm not exactly sure how I'm going to install it yet, but I'm toying with the idea of using it for offboard charging so I'm not carrying the extra weight on track.

So far I've installed the new control board and printed an adapter that will allow me to use 20mm cord grips in place of the old connectors which wouldn't be allowed. It would be kind of neat to give it its own small, self contained coolant loop, so all I'd have to do is connect an HV plug and a LV plug for 12v power and BMS communication. Any suggestions for a small water pump and radiator? I'm wondering if a PC water cooling system would be enough?

Adapter:

g2 charger cable gland adapter 20mm.stl

(245.98 KiB) Downloaded 43 times

Untested beyond checking the fitment

And this is the only wire I have found that seems compliant: https://www.iewc.com/catalog/wire-and-c ... 0032129622

[jrbe]

What's the efficiency of the Tesla charger? What's left to 100% is the heat the cooling system would have to deal with. I started making a peltier cooler with a large PC water cooled radiator that had 3 120mm fans. It had no problem keeping up with 200 watts of heat. It was limited to 8mm barbs from the small pump I was using. I'd guess you'll have to deal with 500-1000 watts of heat though for this.
You could likely use a heater core or 2 along with a variable Tesla / VW water pump and a 10-12" radiator fan to keep it cool.
Some of these onboard Tesla chargers were used in their standalone chargers too, no? I wonder how much info you could find as a reference from those?

[Zieg]

Yeah, I think you're probably bang on with that estimate. I managed to get it running and do a quick charging test on 120v one module at a time, everything seems ok, but as I've tried to set up a workable offboard charger I have realized it's probably going to be more hassle than it's worth. My home EVSE only goes up to 6kw and I can't physically fit the thing where I was hoping to. I think I'm actually going to sell it and focus on installing ccs with Foccci over the winter.

I've been trying to work on my tuning pretty much all summer, but with the weather turning cold now and my son getting sick, I don't think I'm going to have more than one or two more chances to drive it before winter.

Here's a little fail from yesterday that I hope someone else can at least get a laugh out of. I had used my kitchen oven and solder paste to solder an ESP32, but two pins got bridged from too much paste, so I manually cleared it and was giving it a quick reflow. Then my arm grazed the top rack and I dropped the stupid thing, which sent the ESP32 flying off the board and solder droplets all over the bottom of the oven. Unfortunately the ESP32 did not survive... glad I bought two.

So I'm compiling a todo list for the winter. Things that I have identified over the summer but didn't want to spend the time to fix until the car is parked for winter.

• Check SCCA rules and ensure full compliance so I can hopefully race next year.

• Drain and refill coolant under vacuum (no overheating issues observed, but I didn't do a vacuum fill initially)

• Install CCS/Foccci

• Relocate OBC to front of car

• Modify rear toe link geometry for easier adjustment

• Design some anti-sway bars and try to calculate nominal dimensions

• Replace HV wire to inverter with shielded wire (for SCCA compliance)

• Install insulation monitor (for SCCA compliance)

• Modify dead pedal for better comfort since I don't have to allow clearance for a clutch pedal anymore

• Finish designing, fabricate and install rear splitter for better aerodynamics

• Reinforce or replace ramps used to load the car into the hauler with something more sturdy

• Figure out why headlights and tail lights have a slight flicker when the DC/DC converter is running

• Look for any other aerodynamic improvements that I can make to improve highway range

[Zieg]

Okay, winter work has started. I got sick right at the end of the driving season so the car just sat for a bit, but now I've started taking things apart and planning all the winter work. First order of business was to get the tires into the house so the rubber won't be damaged by the incoming cold temperatures. Last year leaving the thing on jackstands was a bit annoying because I couldn't move it around, so this year I built a giant furniture dolley so I can move it around as needed and push it right against the wall when not working on it. I know Greg recently mentioned wanting to move his truck around for the same reason, although I don't think this is exactly what he had in mind.

For some reason I decided to work on the toe link extension first. This is the least critical and I'm not even sure I'll bother with it, but I wanted to get the design finished. The purpose of this part is to move the connection point further away from the lower ball joint, which will allow more precise adjustment when setting the rear toe angle. What I have is workable, just somewhat tedious and I was going for perfect. Originally I didn't make the mounting point further away because the uprights were already expensive and extending the tab further out would just add to the cost (larger stock size to start with, more time machining away all the unused material).

So this is what I'm thinking about doing. It will use the original toe link bolt hole as well as one more hole I will need to drill and tap. Thinking steel for the material - more strength and the weight is negligible.

Still iterating on the design, but now I also need to find a way to make it cost effective. One way to do that is to remove certain features that I know I can mill myself on my manual machine. I hate hogging out tons of material but if the savings are significant then it may be worth it. Also, maintaining a flat bottom surface and designing it such that the features can all be machined from above on a 3-axis mill can help, so I removed the hole that is perpendicular to the rest of the part - I can drill that myself.

Yikes - this is going to be expensive! Here, I left that perpendicular hole in, and left out some of the less precise cuts that I can hog out myself. Since the parts are mirror image though, no discount for quantity. Going to end up paying double this for both sides.

Okay, suppress that perpendicular hole, and what if I mirror the part myself so it's one long 'stick'. Add a bit in the middle for the saw kerf and I can order 'one' part and cut it in half. Okay, not double the price! Great!

Now, hogging out the middle is not going to be fun. Maybe I will replace that feature and see how much it goes up. Sorry this one is formatted differently, I had to go back to my history to screenshot it.

After that I noticed the height of the part is exactly 1". Width is just under 1.25", nothing I can do about that. But the height doesn't NEED to be exactly 1". They may think they need to start with a taller piece of stock so they can machine it DOWN to exactly 1", which adds material costs and the extra machining time to remove all that excess material. I can change it to be 0.985" tall with no impact to me, and that tells them they can start with a 1" bar and skim it down. Not a huge change, but a slight price drop.

That's where I am at now. Next step is to do some stress analysis and determine if it actually needs to be steel or if aluminum will do. Aluminum will be cheaper and easier for me to machine. The reason steel is more expensive it because it takes longer in the machine and is harder on tooling. So not a super consequential update but I home someone finds it interesting. It gives a small insight into the process of designing something like this for easy/cheap manufacturing.

[jrbe]

Do you have room in the lower ball joint area to do something like this?

It might need to flare around the ball joint for clearance / strength.
It could likely be the same for both sides. Looks like there's a lower ball joint boss to align on, can use the old ball joint hole too. This would require drilling your upright but I think those spots could handle it. Might be able to angle the tail section and angle the ball joint or flip it so it's aligned with the lower ball joint. Just an idea.

**Edit** might be able to use a single shear bolt with a safety washer in the tail section. C7 corvettes are like this.

Where's the quote from? Have you tried JLC's machining services yet? They made a couple intricate polypropylene parts for me with orings and BSP threads that run at 250psi. They did a great job.

[Zieg]

Ah, yeah no room to run it under the upright unfortunately, very little clearance to the control arm. I could probably run it on the top though. Still won't likely be symmetrical but I can try. I'll also have to jog it down or use a spacer where the link connects - gotta keep it on the same horizontal plane as the ball joint centerline or else it'll have bump steer. I'll put some more thought into going across the top though, thanks for the idea. Added benefit would be not having to take it all apart to drill that extra mounting hole in the bottom edge!

These quotes are from Protolabs, same place I got the uprights made. I haven't tried JLC but I will check it out, thanks.

[Zieg]

Been working on the junction box I need to add to the rear of the car, while also pulling out the old HV cable to replace it with shielded cable (required by the SCCA rulebook). I think I could use some opinions on where to locate my new CCS charge port. I think I have it narrowed down to two options but I'm not thrilled about either.

First, here's the new junction box. Basically the wires from the contactor box come in from the left side, terminate at the posts (which are at different heights), then exit through the bottom of the picture to go to the inverter. The contactors in this box are for the CCS port. The black box in the upper right is the insulation monitor. The cable from the OBC will enter through the cable gland at the bottom middle, where one leg will go through the fuse block and the other leg goes to the live side of the negative CCS contactor.

The random loose plug you see in the above image is for the insulation testing system. The regulations state that you need to bring your HV+ and HV- through some 10k resistors and terminate at a pair of 4mm banana jacks which have a separate cover over them. To test the system you need to connect a cable with one more resistor that will bridge either the HV +/- together or to chassis ground, and the insulation monitor must detect the fault and open the main contactors within a few seconds. Here you can see how I added those ports. I am thinking about 3d printing a cover that looks like a household electrical socket and then making the test cable look like a fork.

And here's a zoomed out photo of the whole boot, showing where the junction box is. The OBC gets mounted to the bracket just below it.

Now for the tricky part, where to mount my CCS port? I had the old J1772 port mounted at the base of the roll bar on the left (see old photo below) but I'm not sure I can fit the CCS port there. If I do, I will need to bring all the wiring across the back of the car and that would be a bit tight, but doable. I would also need to relocate the 12v battery, more work but totally doable. I'm also a bit worried about whether a big CCS plug would clear the roll bar brace. I may have to mount it at an odd angle.

So the other option I can think of is cutting a hole in the back of the car and mounting the port there (marked as option 2 below). The benefit to this is that both the OBC and the junction box are right beside it so wiring would be simple. The CCS port does have rubber plugs and I don't drive the car in the rain, but I still don't really like the idea of the port being visible all the time. I thought about maybe mouting it in the middle behind the license plate, but there's not enough space behind the motor. Also included in this shot is a sneak peek of the splitter I am designing, should help a lot to reduce drag.

Any suggestions would be welcome!