Paul & Sabrina's cheap DIY 144v motor controller - Page 565 - Fuel Economy, Hypermiling, EcoModding News and Forum

DJBecker · 03-30-2012, 12:55 AM

Quote:

Originally Posted by Alcibiades Ramos

Here is a 3 phase that delivers 132Kw for $100.00 plus shipping.

You've mis-read that. $100 gets you an under-1KW motor.

MPaulHolmes · 03-31-2012, 06:22 AM

I finally finished making and ordering the BOM for the new "1000 amp" high side controller. It should have some flexibility to go over 144v too. I've ordered 250v mosfets, and 400v 120amp diodes. I'm sticking with the 200v capacitors, but they are rated to 250v surges, so I figure they'll be OK. The metals are also ordered. The one I just finished in the garage is a low side, and seems to work fine, but we'll see.

sawickm · 03-31-2012, 11:44 AM

Quote:

Originally Posted by MPaulHolmes

I'm sticking with the 200v capacitors, but they are rated to 250v surges, so I figure they'll be OK.

Hey Paul, I have a large quantity of "NEW" Cornell Dubilier Electronics (CDE) electrolytic capacitors, and might sell or swap some of them. - CAP ALUM 1000UF 200V 20% SNAP (381L102M200A052) series 381L lower ESR. They are made to handle the tough switching of motor-drive bus capacitor applications. PM me if your interested.

-Mark

MPaulHolmes · 04-01-2012, 03:11 AM

Does anyone know what causes a continuous break signal with the serial communications? I'm using realterm, and I'm getting BREAK always set. It says it means "rxd line is broken" I think. But as far as I can tell, the rxd line isn't broken. Are there suggestions on debugging serial communications? I have an oscilloscope I can hook up to rxd to ground, or txd to ground. I never thought to check the serial communications on this board. I verified that it's wired the same as previous boards that actually work.

The RX line on the working one I believe was -6.6v or so. On the nonworking one, it was -13.2v If I remember right. And the working TX was -9.2 and the nonworking was -0.2v. But I can't remember now, I'm too depressed and tired. haha

MPaulHolmes · 04-01-2012, 01:31 PM

Never mind. I had rotated the 3 hole interface for the serial plug on the new board on accident, and so my cable wires were mirrored. I made a new cable and now everything is fine. hurray! I thought I essentially flushed $2000 down the toilet. haha.

Alcibiades Ramos · 04-01-2012, 05:03 PM

Every now and then when I'm winding a guitar pickup I can't help but to have this micro/macro vision of guitarrist David Gilmour playing on a concert for the whole world and then breaking the awg 43 wire that runs through pickup. Something so tiny and fine can be the determining factor between failure and success.

MPaulHolmes · 04-03-2012, 01:24 PM

April 14th is the day the boards will ship to me. hurray! I had done a booboo (a minor one) on the DC driver boards, and tried to ask them if I could fix it, but I was too late, so I had to pay like $65 for them to start over.

mrbigh · 04-03-2012, 09:36 PM

Quote:

Originally Posted by MPaulHolmes

April 14th is the day the boards will ship to me. hurray! I had done a booboo (a minor one) on the DC driver boards, and tried to ask them if I could fix it, but I was too late, so I had to pay like $65 for them to start over.

Measure 10 times, cut only once !!!!!...

dave koller · 04-03-2012, 10:54 PM

Paul:

Hate when I do that !!

Forgot one silly little thing arrrrrrrrggggggggggh...

I sometimes print out the board and then turn it upside down and walk thru it to help spot some obvious goof. It is one trick that has caught some boo boos ..

I do not use schematics and auto-route ... I do fairly complex boards by just years of laying them out with paper tape and stick-ons on ye olde light tables.
So using CAD and putting down traces comes easy.... BUT one still can reverse one stupid trace - or NOT!!! Hey but I can do mazes faster then anybody - you know the kind they put on menus to keep kids busy whilst the junk-food is being prepared...

MPaulHolmes · 04-07-2012, 04:26 AM

All the previous features now work on the "1000 amp controller". All the trouble was the addition of the programmable max current. Everything else had to change too, since so much of it assumed that max_current was about 500 amps. Now there's about 45% of the computing time available, down from 60 some %. The low side prototype is going out tomorrow. 5 will get you 10 that it blows up in the first day. haha

The components for the new high side approach, that I'm planning on actually committing to, arrived today. Now I just have to write all the gcode and wait until the 16th for the boards to arrive.