DO NOT connect them directly! The MCU cannot tolerate their output voltages.
You will need a preconditioning circuit.
3 Likes
hahahaha that could have ended badly, probably not worth bothering at this point as using vesc isn’t the designed use case here
1 Like
I wonder if just adding a sensor would work better.
Maybe an AS5047 sensor if you can access the end of a shaft?
1 Like
I could totally get access to a shaft, again it’s probably not worth installing a sensor just to confirm that this works with vesc, unless @Deodand and Ben want more data points for their firmware. I’m happy to do it for science if actually needed
2 Likes
Oh I forgot you weren’t actually going to use that in there 
2 Likes
soo did hfi work on the mg1 or nah?
1 Like
Haven’t had a chance to take it insdie yet. Just rearranged my study so I have more desk space, floor space and drawer space. Also moving a 100kg transmission up a few stairs on your ain’t easy 
a couple hours and I should have results
3 Likes
I can confirm that HFI does not work on this motor. The ld lq diff value was too close to the actual inductance. In an ideal scenario, I would have the ESC properly wired into the cable connectors and not just shoved and taped into the hole. This motor starts up without any stuttering at all really. Still might try the encoder…
4 Likes
@Gamer43 no issues here. Very interesting, the motor cogs like no other in FOC, but I think that’s due to hardware because the vesc faults as soon as I try and throttle. In bldc it starts up fairly smoothly, but there is some stuttering. Braking is also basically nonexistent. This can be seen in yesterdays videos. In this video, with the resolver, you can see that there is smooth and instant start up, and the same goes for braking. This is great
4 Likes
Managed to blow the 4.12 doing a detection on the 6.6. I think the bullets were touching something. I tried removed the parking pawl gear and it was still in park so MG2 is out of action until I can get the entire thing open.
At this stage in the project the only things I’m waiting on is money to enter my bank account. The main stuff that needs to be bought is Damian’s board, which just saw a $90AUD increase today with a new board release; and the batteries. This is gonna be the most expensive part of the entire build. I think I will buy a 150v 1P battery from someone here on the forum to use for testing, and save up a bunch of cash so I can go out and buy a heap of batteries in bulk. Unless someone wants to be my sugar daddy…
3 Likes
Did you check into the freight forwarding option that the dude from BCH suggested? I used to use one when I was living in Malaysia. It was not exactly cheap, so I shudder to think what it would cost for the amount of cells you would need, but it might be your cheapest option. That said, since you are closer to Asia than most if us in the states, have you looked into buying direct-from-manufacture? I know nothing about that other than it can be done.
If you can find me a company that will ship by sea and will ship lithium batteries I’ll do it. I have looked extensively for batteries on Aliexpress but the issue is buying new ones really isn’t gonna be economical at all for what I’d consider a ‘budget’ EV conversion. JeffWu suggested I go for the A123 NMC 20Ah cells, which are honestly great cells for an awesome price. Unfortunately that price is ~$50 per cell, so even just for one parallel group it would cost me $2500 while only giving me a range of 40-50km. Batteries is a tricky situation because there is so many aspects to consider
3 Likes
These look interesting. For $700aud shipped I can get 50 lifepo4 cells, gonna keep looking but keep this one in mind
those pouch LiFePO4 cells are great
can’t say the same for pouch li-ion cells
1 Like
The specs of the SPIM08s specify a cycle life of 2000 cycles. While I don’t believe this, NMC as a chemistry is being continually developed and there is always more room for improvement. Another factor is the fact that a lot of higher end batteries will never reach the hands of us lowly consumers as they are being producer specifically for certain larger customers. I’m really interested to see the NMC pouch cells that will be coming in the MIC Model 3 at some point in the future, because if they can show a well designed product there it could be useful for me, especially when it comes to things like heating and cooling the battery pack
3 Likes
Here’s the inside of the gear selector mechanism. We have a pancake style 3ph Inrunner will a very large gear of ~70 teeth. The gear is slightly offset so that only a few teeth are engaged at a time. I tried hooking this thing up to my vesc to try getting it spinning but it cogs like crazy in both BLDC and FOC. There are another 5 pins on the connector, but unfortunately the service manual didn’t seem to have anything about these extra pins. Might be unused here, but I believe there is sensors.
On another note, here is an installation video from Damian for his inverter board. This video isn’t for the faint of heart, being 90 minutes long. This is because Damian is trying to cover as much ground as possible so that he doesn’t get asked 10000 questions. There are a lot of tech illiterate people who think building an EV would be cool and then come to Damian expecting customer service. He does this stuff full time living off Patreon and money from the boards he sells and also has a number of health issues so his time is very valuable. The video is very dry, but his Irish snarkyness keeps it interesting.
3 Likes
For those that prefer visuals over words, I’m going into work tomorrow to pick up a bunch of cameras and mics and live streaming gear so I can start making some videos on this stuff. They won’t be super high quality, but will contain the information you need. Wanna stream the teardown of the transaxle as well. Will post the date and time for that here
4 Likes
@Gamer43 I assume that RB, RA and RVC are my sensor pins? I can confirm that the shift control actuator has hall effect sensors, with only a single reference in 4500 pages of documents…
This image is also the only image in 4500 pages of the existence of those sensor pins, thank god for Ctrl+F
1 Like
This fucking thing is so weird. I have wired up the hall sensors, ensured that I have the correct phases wired up (managed to do them upside down…). During the motor detection they work perfectly, no harsh cogging and clean start up. As soon as I try and run them outside detection, sometimes they work, sometimes they don’t. BLDC simply doesn’t work. The hall sensors seem to work but I don’t really know what’s up with it. I’m sure having multiple lots of jumper cable doesn’t help. Giving up for tonight
2 Likes