Paul & Sabrina's cheap DIY 144v motor controller

[MPaulHolmes]

Oh ya, and I ony drill about 1 inch per minute.

Yes, I love this shirt. I always wear it to school, and the students always ask me if I'm going to get a new one.

[squiggles]

You are supposed to sharpen drills different for drilling copper, can't remember the detail, it was 35 years ago in trade school when I learned it!

Less sharp cutting edge I think.

[jackbauer]

Fran and Paul kindly put together some software for me to read and limit motor rpm. Seems to have a little weird side effect of causing the limit function to kick in at very low rpm if the serial communcation lead is left connected

http://www.youtube.com/watch?v=dwPaF2jC2nY

[MPaulHolmes]

Is this the most recent code with the "rpm too high for at least 100ms" requirement? If it is, then I have a different idea. The rpm is probably going from around 100 to over 3000 probably instantly. It may even be the higher bits that are just plain impossible. Like it might go from 0b0000000011111111 (255 rpm) to 0b1000000011111111. Some noise in the high bits. Those high bits are irrelevant, since the rpm will never be something ridiculous like 50000. So, I can make sure the high bits are always set to zero before checking that the rpm is too high. Or if any one of those bits is 1, then you can know the poor thing is out to lunch again, and have it disabled.

Also, the rpm is being checked at around 1kHz, so you shouldn't see an increase in rpm from one read to the next of more than say 100. I could just ignore the fault if the change in rpm is too big?

[jackbauer]

Paul , Its not a problem with the serial lead disconnected or with the laptop hooked up. I just thought you'd like to see the results. Yes this is the 100ms code. The other code would have rattled the car apart

[squiggles]

An un-terminated serial lead causes spurious errors....
Is it
a) Working as an antenna and introducing noise to the ground plane? or
b) Introducing noise to the serial comms data lines?

We are talking about an electrically noisy environment after all.

If it is (a) then maybe some decoupling is required around the electronics, a quick look at the schematic shows me that there is some decoupling, a good thing particularly around logic chips.

If it is (b) parity checking or crc check coding should eliminate errors, that is any message that fails an error check should be discarded.....

It is a curly problem, anyone else got some ideas?
I would be looking for a solution, if it happens with the lead attached there is every chance that it can happen under some other circumstance.

[MPaulHolmes]

I'll add some error checking that might eliminate the problem, but won't affect valid rpms. It's not going to go from 100 rpm to 16000rpm in 0.001 seconds.

The mill is working! hurray! It starts and stops so quietly now. It used to make really loud popping noises when it stopped, but I thought that was just some weird contactor. That means it had been broken for a really long time. The capacitor had a big dent in it when I took it out. Go Harbor Freight!

Here's a message from PCB Cart about the driver board:

Dear Michael Holmes,
Your order 110218-194810-769 is confirmed Mar 4, 2011 18:00 and schedule to deliver on Mar 8, 2011.
We will send you delivery details as delivered. For part number 1000amp Driver Board, we are sorry there is some shipping deay because of test failure, we arrange to remake and new shipping date is Mar 8, 2011. Sorry for the inconvenience. Please let us know if had any questions, thank you.

Those stinkers! I got's da get me one o' dem karate chop sandwiches with peanut butter and honey and PCBCart on it!

Each piece of gcode is written as I make the item, so the list is quite long at this point, but all the milling and drilling is ALMOST done!!! Then it's 50 hours of soldering. My wife had a good idea for a modification to the enclosure (easy for the DIY type), so I think I'll do that. Right now the end caps just sort of slide over the bus bars and sit there minding their own business, but I was going to flare out (or in) the left and right part of the end so there's a face to bolt the endcaps to. Then, with a little caulking or something, it will be easy to make the dang thing pretty much water proof.

[MPaulHolmes]

OK, for 3 of the 4 prototype controllers, all metals drilled and nice and rounded on all the edges, and milled super flat and shiny. Now to send out that PCB for the current sensor. I tested it with 3 different bus bar dimensions, and 0.125" spacing is perfect for all of them. So I can use that for anything from 500 amps up to about 1200amps, and if you add one more PCB spacer, it would be good to around 2000 amps. There's no limit to the number of amps it can monitor. The pcb I'm about to order is going to be 0.125" thick, so you just bolt that sucker on, and away you go! The cost of parts is around $5.50 for everything including the PCB, which is way cheaper than $27.50 for a LEM!

You do need to be able to solder a few surface mount components, but it's not too bad. You also need to use two #6 nylon screws. It also takes up less space than the LEM.

The bus bar needs to be < = 1" wide, due to the fact that the holes where you connect the wires would be too close to the bus bar if it were much wider than that.

[sawickm]

Originally Posted by MPaulHolmes

PCB for the current sensor. I tested it with 3 different bus bar dimensions, and 0.125" spacing is perfect for all of them. So I can use that for anything from 500 amps up to about 1200amps, and if you add one more PCB spacer, it would be good to around 2000 amps. There's no limit to the number of amps it can monitor.

Hi Paul,

A $6.00 1000amp current sensor, that sounds great !!!

Would a smaller 0.062" PCB work with your 300amp per phase AC Controller ? (maybe a 300amp - 500amp range) ?

I would be very interested a couple of those !!!

Regards,

-Mark

[MPaulHolmes]

Yes. In fact, it's what Brusa uses on their 3 phase controller I believe, and what Fran is going to use in his (he's switching from the LEM too).