jonescg's eCRX pre-build thread!

[99t4]

What's the D-sub port for?

 

[jonescg]

What's the D-sub port for?

It's the RS232 port for programming the Rinehart inverter. Keen to do any tuning from the comfort of my driver's seat, and not under the hood.

 

[jonescg]

Just need to get that direction control PCB up on a stand and re-route the HV cables.

 

[jonescg]

A little more progress to report - the stereo has been fitted and it works! Digital radio and bluetooth for the phone.

And today I managed to get the Everything Board working! This serves to prevent drive-off while plugged in and charging, but it also manages BMS decisions (like cut power to the drive, or the chargers for a low or high cell event respectively) and theoretically, charge rate as determined by the control pilot signal. We'll need a battery to charge first before I can know that's working.

I've managed to connect to the PM100DXR to my laptop via RS232, and everything is showing up! There will be some tuning involved and that could take days, weeks or months... Even got a temperature sensor showing some weird numbers.

But the point is, I can see the Accel_V going up and down as I press the pedal, the brake switch goes high and low when I press the brake, the regen-off switch goes high and low when I ground it out... it's all coming up Milhouse.

I reckon there will be some messing around trying to get a baseline for the motor to start to turn, so I'm all ears if anyone has ever tuned a PM100DXR to a Leaf motor.

 

[jonescg]

These little 1 mm wide, 8 mm long slots will make soldering the copper-nickel tabs a whole lot easier, and stronger.

 

[jonescg]

I jury-rigged this 180 V battery to power up the inverter... and it worked!

The Remy HV motor program is a good starting point, but I have a bunch of other stuff to mess with before it behaves itself. Still, turning a wheel for the first time is a solid milestone!

 

[jonescg]

I have finally tuned the resolver to the rotor, making all future tuning accurate and reliable. So far things are coming up Milhouse!

 

[Zambam]

These little 1 mm wide, 8 mm long slots will make soldering the copper-nickel tabs a whole lot easier, and stronger.

Smart to solder the Ni tabs instead of spot welding them to the 0.8mm thick Cu busbar (expensive /powerful welder needed)! What thickness is the Ni tabs? What spot welder you using?

 

[jonescg]

0.15 mm nickel, 0.07 mm copper
I'm using a transformer based Sunkko 709 to do the welding.

 

[Zambam]

0.07 mm copper for the busbar? That's paper thin (0.0028"), not much cross sectional area to carry the current.

 

[jonescg]

The 0.07 mm copper is on the nickel-copper tab sandwich. The copper busplate is 0.8 mm thick.

 

[Zambam]

The 0.07 mm copper is on the nickel-copper tab sandwich. The copper busplate is 0.8 mm thick.

You didn't show the Ni-Cu tab sandwich in your video, which got me confused. Are the tabs also cut with a fiber laser? What power level is needed on your Sunkko 709 to weld the 0.15mm Ni/ 0.07mm Cu tab sandwich to the cells?

 

[The Toecutter]

This is a good donor chassis. Outside of a 1st gen Honda Insight, a Lotus Europa from the 1960s, 1st gen Lotus Elite, pre-wedge Saab Sonett, 1st gen Toyota MR2, a Triumph GT6/Spitfire, and certain kit cars, you won't find anything readily available for under $100k USD that is more efficient regarding CdA value that is also under 1,000 kg.

 

[jonescg]

You didn't show the Ni-Cu tab sandwich in your video, which got me confused. Are the tabs also cut with a fiber laser? What power level is needed on your Sunkko 709 to weld the 0.15mm Ni/ 0.07mm Cu tab sandwich to the cells?

I wish. Unfortunately I have the high tech approach of rolling it out 1 m at a time, sticking the adhesive copper tape on, and cutting it into 25 mm long sections with my snips (and cutting a tiny corner off each one of these, as the tolerance on the busplate slots was too tight!)

I'm at the point now where I have the resolver aligned with the rotor, but I have no torque settings in mind. Either I go for a drive down the street, or try to load up the wheels with a dyno of some persuasion.

Unfortunately the Leaf motor, being an IPM rotor, means the efficiency won't be consistent across the full range of speeds and torques like a surface mounted PMSM. This will take some tuning, particularly at high power I suspect.

Open to any expert advice on tuning a PM100DXR to a Leaf motor.

 

[jonescg]

So $3800 later, the capture plates for my battery modules are made. These were CNC routed from 9.5 mm polycarbonate and are designed to fit the 21700 format cells I hope to use, much like I did for Voltron's latest battery.

The laser cut copper busplates are also carefully organised, and the next major task is to solder the nickel tabs to the plates. Now, I have demonstrated that with the copper foil underneath, I can get a reliable connection which will pass over 60 amps per cell. But in all honesty, 20 amps per cell will be more than enough. I guess any loss is a loss worth avoiding, but it's just going to add weeks to the build.

The 20 mm PVC tubes are designed to fill in the blank cell spot where there's no cell, but critically it allows me to put a threaded rod right through the stack of modules and cooling plates. There's 12 holes in the modules, but I'll only be using 6 of them.

Still yet to buy the cells, but there's about $8k in Samsung goodness. Final arrangement will be 96s20p as 8 modules of 12s and 90 Ah. So about 30 kWh on a warm day.