Paul & Sabrina's cheap DIY 144v motor controller

[Bela]

Alright, one more post and I can add pictures:

What did the father buffalo say to the son buffalo when he left for work?

Bison

[Bela]

Well I think my post count is high enough that I can add pictures. The pictures show the base plate with most of the parts arranged although since the picture was taken I decided to use a smaller relay for precharging instead of the GX12.


The picture with the copper plates show the precharger relay and a second relay for controlling a cooling pump.


One problem I had was that my base plate isn't long enough so I had to overlap the Open Revolt board with the IGBTs. To do this I bent some wires that go from the high voltage side of the control board to the igbt connectors. I don't have any experience with EMF, especially at these voltages and currents so I hope this works OK.


On the bright side, by elevating the controller board, I have room under it for a raspberry pi and for a new board with some voltage monitoring logic for battery and motor voltage. More details to come when I actually get the board working.

The base plate is 1/2 inch aluminum which was thick enough to allow for cooling grooves to be cut directly into it. Thanks goes to Micah and his electric porsche blog for the idea.


Hopefully in a couple weeks I can post a finished controller!

[64jeep]

Hey, nice looking water plate! The rest of the controller looks good too. I'm just surprised to hear someone read my blog!

[Bela]

Would the following modification work for using a hall effect throttle controller? It is hard to see in the picture but the idea is to cut the trace where the red X is (looks like a red blob on this picture) and then swap +5 and Gnd as compared to Paul's original picture. Doing this would make my pinouts a little cleaner.

Looking at the schematic it looks like if the resister R46 was removed then the 2 pads for +5 and gnd would not be connected to anything so their polarity wouldn't matter.

[MPaulHolmes]

I'm a little confused. If you get rid of R46, you can use the left pad of R46 for 5v. Ground must have a connection on the board that's connected to the ground plane. I'm having trouble seeing how you are getting a ground connection on the board by doing what you are suggesting.

[Bela]

The idea is to run a wire from the test point GND to the left pad of jumper J1. Then cut the trace between the J1 left pad and the R46 right pad. Then get 5V from the right pad of R46.

It sounds messy but would allow me to put a 3 pin connector on the board for connecting to the hall effect throttle.

Hope this is right because I already made the changes! Here are some pictures to help clarify.

Before powering on the board I wanted to double check that the J1 left pad is not connected to anything once the trace is cut.

Here is the ground wire:


And the 5V wire (The bottom of the left most pin on the 3 pin connector is bent and soldered to the red wire):

[MPaulHolmes]

Oh! yes!

[Articus]

Just put together the controller board. Is it testable without the powerboard? Right now I have controller board, 0-5K pot, and current sensor plugged in. My PWM output at R17 is flatline. The PWM signal on pin15 of the Atmega 168 is also flatline 0V. I've gone through and checked the supply voltages at the 6W converter, the 12V in, and the LM7805. All are correct voltages at 12V and 5V for the latter.

Also I am getting errors trying to download RTD explorer. Says something about not being able to find bobert...Can anyone help?

Thanks!

[Articus]

I also just get a solid green light. Yellow LED does not come on.

[jedsmd]

Originally Posted by Articus

Just put together the controller board. Is it testable without the powerboard? Right now I have controller board, 0-5K pot, and current sensor plugged in. My PWM output at R17 is flatline. The PWM signal on pin15 of the Atmega 168 is also flatline 0V. I've gone through and checked the supply voltages at the 6W converter, the 12V in, and the LM7805. All are correct voltages at 12V and 5V for the latter.

Also I am getting errors trying to download RTD explorer. Says something about not being able to find bobert...Can anyone help?

Thanks!

Are you using a preprogramed chip from Paul? If not, have you successfully loaded the unified firmware?

Index of /firmware/cougar-v1.11b/hexfiles-m168

Also when loading firmware for the first time the fusebits need to be set, see mora's post:

http://ecomodder.com/forum/showthrea...tml#post399312