Modding my EB236 150V controller the same as my EB318 12kW

zombiess said:
Only bad news is I put a scope on the gate of an FET and did some measurements to see the driver performance... really poor. See my other tech thread for details on that
Dude when you get as far as I have been you will realize why its good to stop with the china controllers 100% and move on to something better. We have a few DIY controllers on the forum and we are getting other options that vary from fro board to sevcon to kelly. I just can justify spending the time to build a 24 or 36 fet china controller for 8-12 hours anymore when I can build the entire thing from scratch with lebowskies chip.
Dont get me wrong dude you did some great work here and for a hub motor it should rip but im not suprised the driver performance is lame! What thread is it on? I cant find it :(
 
Arlo1 said:
zombiess said:
Only bad news is I put a scope on the gate of an FET and did some measurements to see the driver performance... really poor. See my other tech thread for details on that
Dude when you get as far as I have been you will realize why its good to stop with the china controllers 100% and move on to something better. We have a few DIY controllers on the forum and we are getting other options that vary from fro board to sevcon to kelly. I just can justify spending the time to build a 24 or 36 fet china controller for 8-12 hours anymore when I can build the entire thing from scratch with lebowskies chip.
Dont get me wrong dude you did some great work here and for a hub motor it should rip but im not suprised the driver performance is lame! What thread is it on? I cant find it :(

It is what it is. Not going to stop me from using it after I figure out if I can improve it with some simple tweaks. Lots of people using these controllers with crappy drivers to power their high powered bikes and I'm no exception. I would like better, but this is what I got for the moment. Surprises me no one has posted this information before. Was pretty easy to gather, took me all of about 45 mins of playing around with a scope, battery, ebike and camera. I'll see if I can update this thread tomorrow after I experiment some more with different resistor values or higher voltage to allow more current to the on board SMPS and check for ripple on the 15V regulator. I have a feeling it's struggling to charge the caps on the gate drivers with my low voltage and high 510 ohm R115 resistor.

I have one of Lebowskis chips, now I need the time to build it and my power stage stuff, but before I do those projects I'm working on my throttle interface so I can hopefully have a more controlled high powered ebike as soon as I wrap this project up and install it in it's case.
 
Zombiess,
Great work! This huge controller build is quite entertaining.... I found myself laughing out loud when in the photo shoot on page 1 of this thread, the assembled 36 fet controller is resting peacefully on a pamphlet titled:

"Helping You Take Control of Your Money and Credit" :D

I'm so there.

Mark.
 
Just got back from the first test ride, controller works good in the two miles I rode it and tried to abuse it and I couldn't even get the case to not be cold to the touch. Set it for 70A battery 200A phase on a partially charged pack that was at 115 Volts. I'm going to need a new swing arm to really turn it up beyond where I ride it now. All I need to do is the R12 mod to allow regen to work at high voltage and it should be a done project.

Hard part was mounting it. If I installed it where the 18 FET goes there would only b 1-2mm clearance between the crank arms because the case is so wide. I ended up installing it on an old seat post rack since I'm just testing it out to make sure it's OK. I'm going to put the 18 FET controller back on this bike since it's all it needs unless I decide to go hill climbing.

Building one of these from the board up was a great learning experience, but quite time consuming. next one I build should go quite a bit faster now that I know a bunch of tricks. I guess this project wasn't too bad, took me 24 days from start to final test and I only worked on it a little at a time + waiting for parts, not working on it for several days at a time. I'm guessing I'm into this controller for about 40 hours of labor mainly because of the copper buss bars and not having a hot enough soldering iron. Once I got the right iron everything went faster.
 
I already posted this info in another thread, but I wanted to complete this one for anyone searching. Some hard core testing I did
-------------------------------------------------
Just got back from a 16 mile test ride, controller settings were 100A battery (the max the software went to) and 180A phase, 110% throttle in position 3. Battery setup is 30S3P (added on 5 additional 5AH 20C in parallel with the 25C Nano Techs). On the last 6 miles of my ride I was basically riding it like I stole it and even managed to get the motor case warm enough that I could only hold my wussy hands on it for a few seconds after it sat for 5 minutes for the first time WOOT! Not hot enough to reduce any performance though so probably not much over 100C if even that, I need to install a temp probe in the windings eventually, just need a type K with a long lead.

My CA is not calibrated to this shunt, it's setup for my 0.5 mOhm shunt in my 18 FET, I'm guessing these 5 shunts in parallel put it some where around 0.7 mOhms, but it's just a guess based on previously dealing with several of these shunts (3 in parallel ended up being 1.25mOhm).

Anyways the real point to this post is I managed to pop my 120A fuse from current draw on the 110% throttle setting and the controller was just slightly above ambient temp after lots of hard acceleration due to varying my speed a lot on the last 6 miles. Good news is I always carry a spare, but it was only an 80A and it popped about 1/4 mile from home after I went 57 MPH with the battery sitting at 115V at rest and was messing around doing acceleration pulls. Block time is still set at 1.0 sec. I need to figure out the shunt value on this controller and start using XPD with it.

Pretty awesome the controller is doing just fine and I melted a 120A fuse just from pure power. I'm guessing that means I was pulling about 150A for a while for the fuse to get hot enough to pop. I have some 150A fuses I can install which should also help. Was a really fun ride. Turning up the battery amps really gives the bike awesome acceleration, much more than my 18 FET was giving it but it was much more controllable and smoother on this controller.

Looks like this 36 FET I built is going to be good for at least 150A with no sweat. Even tried climbing some lower grade hills (maybe 4-5% at 40% throttle for about 2 miles no stop) and the controller didn't get warm.
 
Back
Top