Paul & Sabrina's cheap DIY 144v motor controller

[taffy]

I am really liking this thread! I might make one of these before i splash out on buying a controller for my EV.
Will talk to a few friends of mine who are electrical engineers who work on power turbines and see if they can come up with a logical PCB layout or any suggestions.
Can someone point out the post with the material list, i cant find it again! :S

[MPaulHolmes]

Materials List! Yikes! I don't think there is a COMPLETELY up to date materials list. I need to do that this week.

I have a mostly updated controller parts list, but it's probably only 98% correct. I need to go through and make sure everything is right. I attached it anyways! ya!

EDIT: Dang it! The formatting is all messed up on it, since it wouldn't let me submit it as a rich text format. grrr.... I don't have word! Only wordpad.

[bennelson]

Paul,

We should try to get parts lists and schematics, etc. all together as a "sticky" or just attached to the very first post. Somewhere that it is all together, and we can keep it updated and correct.

[NiHaoMike]

Originally Posted by MPaulHolmes

EDIT: Dang it! The formatting is all messed up on it, since it wouldn't let me submit it as a rich text format. grrr.... I don't have word! Only wordpad.

Use OpenOffice.org - The Free and Open Productivity Suite .

[taffy]

Thanks for that paul, will look at it and see about formatting it + turning it into a PDF.

[taffy]

pdf version of the parts list

[GrahamMc]

Taffy-
I took Paul's advice and looked in on Ian Hooper's design of a 600 A motor controller

Zero Emission Vehicles Australia

It has some interesting design items, like the bus bars. I am still trying to understand Paul's schematic, but I do more mechanical than electrical and it is nto coming easy. Still wondering, Paul, are you using a PCB board for the bus bars with your 500 A controller? Ian replied to a question I had for him about an update and the current status of his controller. Here is what he said:

Hi Graham,

Ah yes it was great to see Paul's success with his motor controller
development - I know from experience how challenging it can be!

My latest prototype from December is still going strong, the only
change since then has been to add a larger heatsink plus cooling fans.
This eliminates the heating problem of course but personally I think
forced convection is a bit of a workaround for what is evidently a
somewhat inefficient design! Synchronous rectification would help a
lot (but is difficult to implement), and there may be ways to shorten
switching times (and hence reduce heat) as well.. I'm also concerned
with the lack of cycle-by-cycle current limiting; another thing to add.

So I guess in short I'm not particularly happy with the design yet,
and I don't think it's "bulletproof" enough for public consumption!
But hopefully it'll get there eventually..

Kind regards,

Ian Hooper


Any thoughts on his concerns about the heating and switching rates? Bulletproofing? Thanks for the component list and source info, and the schematic.

[MPaulHolmes]

Heating issues:

I used a big fat commercially pure (C110) copper heat spreader instead of aluminum. Also, my mosfets are more efficient than the ones Ian is using. Mine have an RdsON typical of 8 mOhms, and his is 17 mOhms. So, my mosfets will have 1/2 of the heat loss. He found the more efficient mosfets after already building his, so I used them instead.

In version 2, I have a way to bolt the heat sink to the copper heat spreader with much much more pressure (not tapping into the copper heat spreader), which will allow more efficient heat transfer from the heat spreader to heat sink.

I'll have larger switching losses because I'm using larger gate resistors at first, just to be safe, to keep down the voltage spikes. I can improve that later.

My design uses hardware current limiting, and so is bullet-proof in the sense of it being impossible for the current to get so high that the mosfets burn up.

[Roger Heuckeroth]

I've been thinking about what components would be on the logic board and what would be mounted on the power section of the controller.

Are the current sensor and the thermistor the only components that would be on the power section?

Using the modular approach we would need some I/O connectors to plug the two together. Does anyone have some suggestions? I guess we would need +5V, ground, current sensor, thermistor and PWM 12V. That's 5 wires, correct? I would think some standard 5 pin connector would work. Should it be a water-proof connector? Sheilded?

We also need a 12V power input connector, 2 wire throttle input connector and motor speed sensor input How about pack voltage? We discussed having the controller lower the current limit if battery voltage got too low. I think that was one of the reasons for stepping up to the ATMega16. I also suggested a tachometer driver option. That may need to be run through an opto-isolator to prevent ground fault. Also, let's not forget about the RS 232 connector as was suggested.

Details, details. I geuss the connectors don't need to be physically on the board. Just have solder points to connect the wire. I like the idea of keeping the logic board in its own seperate box. It would be nice to package the board and connectors all inside a nice water tight box. One of my pet peeves about the Zilla design is that it is really not water-proof. If you want your EV to be an all weather design, I think it should be designed to take a splash or two.

One other comment on the copper heat spreader and bus bars. I have heard it suggested that leaving copper bare will lead to corrosion problems down the road. Maybe plating the copper with a thin coat of zinc or gold (if money's no object) would be a good idea. The Zillas bus bars are plated copper. Not sure what they are plated with.

[Tango Charlie]

I've got your gold plating...