Paul & Sabrina's cheap DIY 144v motor controller - Page 198 - Fuel Economy, Hypermiling, EcoModding News and Forum

MPaulHolmes · 07-10-2009, 06:52 PM

There's a different Tamura that wouldn't require major redesign. It takes 5v input. The only problem is the aperture for this Tamura 600 amp sensor is too small:

http://www.tamuracorp.com/clientuplo...L01ZXXXS05.pdf

I have been using 0.75" wide bus bars, but I could just shave them down a bit??? I have been thinking hard about that for awhile.

TheSGC · 07-10-2009, 08:23 PM

I actually have two of those 600 amp Tamura sensors and they are actually the same current sensor that is used for the PakTrakr.

I haven't used them yet, but they do have a small hole that needs a real small buss bar.

steiner · 07-10-2009, 09:30 PM

I had previously spoke to an applications engineer about the LEM current sensors. He did tell me that the voltage increases linearly with the same slope once you exceed the continuous rating of the sensor up to the max limit which I believe is +/-900 amps for the 300, 500 and 600 amp sensor.

I am running my LEM500 with a small PIC chip and LCD display. I haven't had any problems. However, I am not running your controller (yet!!!) and maybe it is not producing as much EMI. It is also a stand alone system and not part of the controller. I plan on incorporating it into my BMS in the next few days.

Rick

jyanof · 07-10-2009, 09:37 PM

Wow!

There was a good sized package at my door when I got home! Xmas in July! The controller is beautifully constructed and has a very 'solid' feel to it.

I went out to my car and held it above the curtis 1231c and the dimensional similarities are uncanny. It's just an inch or so shorter in length and roughly the same height compared to the curtis. the width of the vertical portion is identical - the mounting flanges stick out past the width of the curtis.

Paul, I saw that M- is in the middle with the current sensor on it. When looking AT the controller from the front (forward looking aft), is it B- M- B+ from left to right? My guess is based on the cap markings...

If that's right, it's perfect!

on another note, my curtis heatsink temp got up to 55C while meeting some people for lunch in the middle of the day (~44C ambient). Lotsa time in current limit mode...

I'll probably charge my batteries and get started on installing this beauty tomorrow.

MPaulHolmes · 07-10-2009, 11:06 PM

Hey Joe! Yes. B-, M-, B+. I'm glad it made it there! I can't wait for us to hunt down that dang blasted cause of the micro latch up. I have some ideas. When my son goes to sleep, I shall continue. It makes me sort of sad that it wasn't perfect when sent out, but that's OK. We'll see what happens.

For now, Joe, be careful. Don't drive it in places where it is critical that it doesn't latch up. I guess they don't call it beta testing for nothing!

MPaulHolmes · 07-10-2009, 11:46 PM

I am much closer to finding the problem! The microcontroller ISN'T latching up! Instead, the mosfet driver is getting disabled by the hardware overcurrent shutdown, and the microcontroller isn't clearing the overcurrent event! This is fixable! It's not some vague voltage spike problem.

In fact, I believe there's a temporary work-around. I'll try it when it's light outside. After the overcurrent event happens, just have it enter an infinite loop, which will cause the watchdog to reset. Set the hardware overcurrent trip point pretty high (maybe 650 amps?), and tune a PI loop so that it never gets tripped (software keeps current limit around 500 amps or so), and if it is tripped by some miracle, Joe and Adrian and Ben will just have to lift their foot for a second, and keep driving (the high pedal lockout will be engaged again, since the program would have restarted).

Even better than than, is to figure out why the micro isn't clearing the overcurrent event. hmm! Oh the mystery of it all!

MPaulHolmes · 07-11-2009, 02:54 AM

Alternatively, maybe the overcurrent flag isn't getting to the micro, or the micro isn't clearing flip flop. I can't wait to test which one it is (or maybe something different entirely!).

blowncurtice · 07-11-2009, 05:16 AM

If only I were 40 or 50 years younger . I don't know how paul does it but after this he should not have to work as a substitute teacher, should be able to get a job problem solving in a tech organisation. I had better get to sleep now so I can get up at one AM here in AUS to catch Paul's test results from Seattle at about eight AM.
My total Parts list proce is AU$303.86 = US$237.47 at todays rate.

MPaulHolmes · 07-11-2009, 03:09 PM

I believe the flip-flop isn't resetting OR it's related to the undervoltage lockout.

My next test will be to remove R14, and put in a LARGER resistor, thus lowering the overcurrent trip point. Then I'll take a drive, and if it trips almost right away, I'll know the overcurrent shutdown is coming on, and not getting turned back off.

TheSGC · 07-11-2009, 04:04 PM

Quote:

Originally Posted by MPaulHolmes

I believe the flip-flop isn't resetting OR it's related to the undervoltage lockout.

My next test will be to remove R14, and put in a LARGER resistor, thus lowering the overcurrent trip point. Then I'll take a drive, and if it trips almost right away, I'll know the overcurrent shutdown is coming on, and not getting turned back off.

Maybe I am reading the wrong code, but in the Wiki the code I looked at doesn't seem to run the Reset Current command, it appears to just stop after going into hardware current limit and setting pwmDuty to 0. Which code are you using?