Paul & Sabrina's cheap DIY 144v motor controller

[isaac_alaska]

good to see you back! Horatio, if i'm not mistaken? I'm nearing completion on my motor controller, using the VLA's i got from you last year. Realized i needed the 15V dc-dc and an additional 5v to power them, so those parts are on the way. i'll be posting more pics once i start on the power stage

[jackbauer]

Thought i'd drop in with a little update on the liquid cooled controller. Its been running just great over the past few months. No issues at all. Easilly pulls 500amp accelerations and cruises at 100-150amps with ease. I've done motorway (60-70mph) and town driving with no temp issues at all. No fan used. Just ram air while driving. Typically it only rises a degree or two above ambient. Biggest problem i have is bugs in the radiator

[isaac_alaska]

Is there any benefit to laying the bus bars flat on the igbt terminals, instead of standing them on edge? Standing them on edge gives more isolation between them, and a larger space at the end: more room to bolt terminals to them

[adamj12b]

Originally Posted by isaac_alaska

Is there any benefit to laying the bus bars flat on the igbt terminals, instead of standing them on edge? Standing them on edge gives more isolation between them, and a larger space at the end: more room to bolt terminals to them

How would you stand them on their side?

-Adam

[mrbigh]

Originally Posted by adamj12b

How would you stand them on their side?

-Adam

Yeah, give us a picture of your concept......

[isaac_alaska]

I was thinking something like this, it seemed that with really high currents it would be hard to cram enough copper into a 1/4 inch thick bar that was laying flat, without coming pretty close to each other. Looking at it now though it looks like it will all fit.



i was thinking of making the bars only as wide as the terminals on the IGBT's, and then making them taller to make up for the reduced width and still get the same copper cross sectional area. probably just added weight though...

[isaac_alaska]

Would someone be so kind as to describe how this circuit works, or perhaps draw me a schematic? I'm pretty sure it's a pre-charge circuit but i'm not real sure at all where it all connects to or how it works



Also, does the board's 12v+ need to be connected when programming using the header pins?

[MPaulHolmes]

Hello! Yes you need to hook up 12v power to program the microcontroller unless you are using the stk500.

Purplish wire leading out goes to battery pack +. Blue resistor can handle huge energy surges. Left 2 posts of orange relay are the switch. One side of the switch connects to the precharge resistor. The other side of the switch connects to the capacitor bank +. The other side is the 12v control of the relay. Upper right post of relay goes to 12v battery ground. lower right post goes to the output of the leftmost (in the picture above) output of the PNP transistor (or p channel mosfet. I forgot). That transistor goes from 0v up to 12v when the microcontroller enables it. ya!

[adamj12b]

Originally Posted by isaac_alaska

I was thinking something like this, it seemed that with really high currents it would be hard to cram enough copper into a 1/4 inch thick bar that was laying flat, without coming pretty close to each other. Looking at it now though it looks like it will all fit.



i was thinking of making the bars only as wide as the terminals on the IGBT's, and then making them taller to make up for the reduced width and still get the same copper cross sectional area. probably just added weight though...

A 1/4" x 3/4" bus bar is good for about 35A continuous with a 30 degree C rise in temp. The same size bar is good for 500A continuous with a 65 degree C rise.

Even the 1kA controller will not run at more then 350A continuous, and peaks like acceleration will not melt the bars or anything.

Here's a chart that might help a bit. Ampacity Chart for Copper Bus Bar

-Adam

[DJBecker]

Originally Posted by adamj12b

A 1/4" x 3/4" bus bar is good for about 35A continuous with a 30 degree C rise in temp. The same size bar is good for 500A continuous with a 65 degree C rise.
...

Here's a chart that might help a bit. Ampacity Chart for Copper Bus Bar

Keep in mind that this type of chart is of minimal usefulness when designing a motor controller.

The ampacity rating of a bare bus bar is typically rated in still air, with no other heat input. Even there, the rating leaves many cooling parameters unlisted (vertical, horizontal, or on-edge has a big effect on convective cooling). The rating of a wire is usually based on being in a conduit, with long-term heat build-up taken into account.

A motor controller environment is much different. There is a major heat input from the devices, combined with active cooling. You typically design around the device cooling rather than bus bar ampacity. In many cases short wires are significantly thinner than charts would suggest, with the heat being draw out by the surrounding structures.

Modern thermal design has moved from massive heatsinks to immediate dissipation. Fansinks use just enough copper to spread the heat to thin fins, where it's immediately transferred to moving air. Having extra thermal mass isn't bad in a motor controller, where there is often 10 seconds of full-current acceleration, followed by a much lower thermal load. But you still can't design around only thermal mass or you'll overheat on hills.