Paul & Sabrina's cheap DIY 144v motor controller

[isaac_alaska]

i guess i shouldn't say "only"

but i think i could handle building a power board with the right equipment and cooling in it.

[MPaulHolmes]

Hi Isaac! I used to work in alaska, in Kasilof, cleaning fish. The hours were 8 on, 30 min off, 6 on 1 hour off, 6 on 6 off, repeat. aaahhh!

The parts and everything are somewhere. AND I'M BACK!!! i JUST EMAILED THE BOM TO MARK. Caps off now. I didn't mail a bomb to anyone though. I'm not the unibomber.

[isaac_alaska]

hi paul, great project you've got going here. i totally know where you're coming from with the fishing hours....i've got quite a few friends who've done it and it's always the same story. 48 hours on...maybe enough time off to grab a donut and cup of coffee from below deck. then come back up for another 24 hour shift..haha.

but yeah, excellent project. it looks way more promising for our project than buying a zilla, and from i can see, it looks like the one you're developing does everything a zilla will do. i'm pretty sure that 1000 amps is overkill for the motor we're running now (it's a netgain warP 7, rated for i think 495 amps peak for like 5 seconds) but i figure that i would rather err on the safe side and go bigger. that way i could program the controller to run around 500 amps and know i was in the safe zone. the logisystems controller we were using had a catastrophic failure and shot a nice flame out the side. later turned out that our pre-charge circuit wasn't doing what it was supposed to do.

a 400 volt pack is probably overkill as well, but i figured that losses would be lower by going with a higher pack voltage. also, contactors wouldn't have to be quite as beefy, since they would see a smaller current.

this is something i'd like to have up and running by about july of next year, so we can run tests on it and see what it can do. the competition (sae clean snowmobile challenge) isn't for another year from now, but i've learned it's a very good thing to plan ahead

[isaac_alaska]

hi paul, great project you've got going here. i totally know where you're coming from with the fishing hours....i've got quite a few friends who've done it and it's always the same story. 48 hours on...maybe enough time off to grab a donut and cup of coffee from below deck. then come back up for another 24 hour shift..haha.

but yeah, excellent project. it looks way more promising for our project than buying a zilla, and from i can see, it looks like the one you're developing does everything a zilla will do. i'm pretty sure that 1000 amps is overkill for the motor we're running now (it's a netgain warP 7, rated for i think 495 amps peak for like 5 seconds) but i figure that i would rather err on the safe side and go bigger. that way i could program the controller to run around 500 amps and know i was in the safe zone. the logisystems controller we were using had a catastrophic failure and shot a nice flame out the side. later turned out that our pre-charge circuit wasn't doing what it was supposed to do.

a 400 volt pack is probably overkill as well, but i figured that losses would be lower by going with a higher pack voltage. also, contactors wouldn't have to be quite as beefy, since they would see a smaller current.

this is something i'd like to have up and running by about july of next year, so we can run tests on it and see what it can do. the competition (sae clean snowmobile challenge) isn't for another year from now, but i've learned it's a very good thing to plan ahead

[sawickm]

Originally Posted by MPaulHolmes

The parts and everything are somewhere. i JUST EMAILED THE BOM TO MARK.

I just opened Paul's email, I have his files and will be uploading them to my file server, and modifying the Revolt wiki with the new links in the next few days.

-Mark

[isaac_alaska]

Originally Posted by adamj12b

Isaac,

Running 400V and reducing to 170 would be pointless unless you are using batteries that sag by 60%. You would most likely not need to go over 250V to always get 170 at the motor and nothing lower.

As for going higher then 250V, you will need to redo the power board at the minimum to increase the isolation gaps or else you will have problems with the high voltage humping the gaps.

If you want high voltage, you will be better off going with IGBT modules.

Are you riding the sled long range or doing something like grass drags?

-Adam

i was planning igbt modules, as i like having a few large components onboard instead of 30 to 50 small transistors. we are riding the sled both long range, and doing drags with it. we're competing in the sae clean snowmobile challenge, and one of the events is longest range (therefore efficiency is important.)

is there any reason you can think of that lower voltage would be more efficient? (maybe igbt's are more efficient at lower voltages?) but from evereything i've done, higher voltage is always more efficient. i haven't gotten into power electronics in very much depth in my learning career, but i'm almost done with my bachelors in electrical engineering (studied power and controls, but not a lot of power electronics)

running more cells in series would also mean having fewer cells tied together in parallel. i'm thinking this would be a good thing, as a cell that develops an internal short is going to bring down all of the other cells in parallel with it. the other option is to simply parallel the strings and leave the individual cells alone, but this leads to bms nightmares.

the sled also competes in a maximum pulling power event, where it's attached to a rolling contraption and the brakes on the contraption are applied until the machine won't pull it anymore. another event is acceleration under load, where you run 500 feet as fast as possible, pulling a 500 lb load.

The batteries we're using have very low internal resistance and a very high discharge rate, so voltage sag isn't an issue, higher voltage would simply be for higher efficiency and the ability to run fewer cells in parallel with each other.

what are your thoughts?

[MPaulHolmes]

I think higher voltage is definitely better. IGBT modules... You'll need a good way to drive them. The control board as it is can drive some OK, but can't do negative voltages for turn off. You can definitely make something work that will stomp the competition. hahaha ya!

[isaac_alaska]

can you pm me the current schematic of the control board and power board? i didn't realize that igbt's required a negative voltage to turn them off, but i'm envisioning a small circuit so that the nominally zero voltage from the control board will go through a transistor and be re-biased to whatever negative voltage is required to turn the igbt's off. i guess one of the first steps would be to decide what igbt's to use, and pore over the data sheets for them to see what kind of control signal they need.

happy to hear that you think higher voltage is better as well, kind of felt like i was climbing out on a limb there for a bit

*edit*

from the data sheet i'm looking at right now, it looks like the igbt needs anything over six volts to turn it on....maybe this is just a special one though? it's hard to find the igbt's i have in mind on digikey though...they have so much stock you pretty much have to know exactly what you need. The data sheet i'm looking at though shows it drawn as essentially a mosfet. quite small though..i'd way rather go with something large and beefy, if only for heat dissipation purposes

[MPaulHolmes]

Ya, the standard seems to be about -8v off, 15v on. We made a driver on here that does just that. It requires the vla500-01 though, which is out of stock until May.

Ebay might be a good place to find some really big ones nice and cheap:
PRX IGBT CM600HA-24H Power Module 1200Volt 600Amp - eBay (item 350362896821 end time Mar-31-11 12:44:35 PDT)

[isaac_alaska]

Originally Posted by MPaulHolmes

Ya, the standard seems to be about -8v off, 15v on. We made a driver on here that does just that. It requires the vla500-01 though, which is out of stock until May.

Ebay might be a good place to find some really big ones nice and cheap:
PRX IGBT CM600HA-24H Power Module 1200Volt 600Amp - eBay (item 350362896821 end time Mar-31-11 12:44:35 PDT)

that's definitely what i had in mind...somewhere i had read that the higher voltage ratings were less efficient, because of the larger collector-to-emitter voltage drop when the device was conducting. (maybe someone can let me know if i'm wrong and misread that) i think i've seen quite a few 600 volt, 400 amp ones on ebay though, and i'll probably go with three of them. 1200 amps seems like a nice round number, and if i ever want to drive an electric shuttle bus with the controller, it should handle it no problem

also, this looks like the part you mentioned...digikey is out of stock but they seem to be available in other places on the internet.
VLA500-01 - POWEREX DC-DC CONV & IGBT DRIVER