Paul & Sabrina's cheap DIY 144v motor controller

[Tango Charlie]

Yay, 101 pages!

and binary 9 posts!

Yes, I am a child.

[esoneson]

Paul,

Digi just sent me a notice that my SDK500 development kit has been shipped from the Manufacturer (China) and is due in at Digi in about a week. So I will have my development environment set up in about two weeks to assist/test in code development for the controller. Thanks again for the lead.

Eric

[Roger Heuckeroth]

Paul,

I noticed you normally buy from Mouser and Digikey. Ever buy from Newark? I got a 15% off my next purchase coupon good until the end of the month. I was wondering if it would make sense to buy a bulk order of caps, MOSFETs of diodes from Newark. I figure it might be a good way to get the cost down on some of these components. What do you think?

Roger

[edown]

I just orded my mosfets IRFP4668PBF from AVNET for $6.18 each. They have no minimum or handling charge.. Newark has a 100 piece minimum. Digikey has none in stock.. And IRF.com has a $30 handling charge.. Arrow is $4.18 each if you back order 100 pieces..

[MPaulHolmes]

Eric: That would be hugely awesome getting you in on the programming. I'm not a professional programmer.

Roger:
That would be great, but I checked, and the mosfets and diodes are more expensive (even with the discount), and they don't carry the capacitors. Oh well! That was a great idea!

Wait, they are the ONLY place I've found that carry one of the new components that will allow us to have the contactor AND the pre-charge resistor completely handled without the human in the loop at all. The component is $6 each I think. I have 3 of them right now from them. I think we would need 2 per controller at this point.

Part Number: PVN012PBF
They are $5.88 each for 1-9. $4.71 for 10-99. That would be about $2 off per controller with the 15% discount, so not a HUGE savings, but better than nothing?

Alternatively, I've been eyeing the filter capacitors on digikey. They *only* have 186 left right now. To get a price of $1.52 each, we have to get 200 of them, which would be a savings to the kit purchaser of about $25, which is more significant. Also, they need to order more, and their lead times can cause problems! June 7th is the expected next shipment I think. That's probably fine, but I would like to do a couple practice controllers to try some different things. Ben's and maybe another couple.

Also, 30 sample diodes are on their way. What a difference being associated with Ecomodder, and Darin's vouching for the project makes! This company is going to get us a VERY good deal on the diodes. I'd like to buy one or 2 hundred of those when the time comes. Those would then be $2 each, instead of $4.21 each from Digikey. That would save about another $20-$25 for kit buyers. That can really add up. I would also like to do a bulk purchase at some point of about 100 mosfets, which would represent a savings to the individual buyer of around $30-$35 off the price of parts. Cooperating like this will make this a much cheaper controller. I need to figure out the new total cost based on the new components that are adding robustness and safety. I'll get to that soon, once the BOM is done on version 2.

[MazdaMatt]

Wow... big savings from 1-off to 100.

As for that opto solid state relay thing... what's it being used for? It looks like an expensive version of a optoisolator and a mosfet. If you are just using it to turn on a relay... scrap it and buy an optoisolator and a little mosfet!

[Roger Heuckeroth]

The offer of 15% off at Newark is good until 4/30/09, so we have some time to comparison shop. What do you think the best way to do this is? I know there are all types of people interested in converting ICE to EV. Some are experts at electronic assembly, and the kit set up is ideal for them. However, we don't want to discriminate against those with little or no electronic assembly skills. And then there are people like me that have the skills, but would rather pay someone to do it for them.

Paul, you mentioned that you would be glad to put together some controllers over the summer, and Matt generously offered to solder up the Beta run of logic boards. This is all great, and will really work out well for the time being. I'm just wondering if you can make a real business out of this? Have you thought about that?

Say if the parts come to $300, and it takes a day to make a controller from scratch. If you turned around and sold that controller for $600, a lot of people would be happy to pay that price or more for a decent 500 A, 144V controller. You would end up making $300/day building those controllers. Would that be something your into? It would get pretty tedious after a while, IMO.

Another thing to consider is liability. We all joked about Ben's days being numbered when he became the guinea pig for the first 144V test. Well all joking aside, if you go into business building controllers for others, product liability is something to consider. Wherewolf mentioned setting up a nonprofit corp. That and some good insurance may be necessary to protect you.

I've started up several successful businesses, and I would be glad to help you in any way possible.

[GrahamMc]

Hey a friend of mine told me about octopart.com. Lists several suppliers and inventories. Might want to check them out.

[MPaulHolmes]

The controller arrived! Ben sent it back really carefully! I bet that cost a lot. I'm going to save all the stuff when I send version 2 to him. OK, I've had the chance to do some analysis. Here's what I've done:

The free-wheel diodes are all good. The 3 mosfets closest to the B- post all had their legs which were connected to B- blown off. The worst damage was the mosfet closest to the B- post, and then 2nd worst was the 2nd, and then 3rd worst was 3rd. All other mosfets still had all their legs.


For those 3 mosfets, let's name the one closest to B- post MOSFET A, the next one, MOSFET B, and the last one MOSFET C.

For those 3 mosfets the resistors from gate to source had burned up on the source side of each one.

Mosfet A couldn't be checked if it had shorted from drain to source, since there was no source leg. It measured 170 Ohms from gate to drain.

Mosfet B is not failed shorted from drain to source, or gate to source, or gate to drain. It might still be good, but the source leg is mostly burned off.

Mosfet C IS failed shorted from drain to source, and gate to source measured 8 Ohms.

The 7 unharmed mosfets are not failed shorted. They are still good.

The control board seems completely unharmed. I removed the control board and powered it up on the oscilloscope, and it is behaving normally. I was able to control the pwm signal with a pot just fine.

MOSFET A

ANOTHER PICTURE OF MOSFET A

MOSFET B

MOSFET C

[Christ]

Dude - you really shouldn't trim your fingernails back that far. :P Check your PM's!