Paul & Sabrina's cheap DIY 144v motor controller

[mora]

Is it possible to limit older design of 1000A controller down to, let's say 700 amps? In hardware protection limit I mean. I got the PCBs and almost all the necessary components for it but what I've read it is not be stable. I thought it might work with some downgrading. Or should I just get new PCBs when those become available?

[Greg Fordyce]

Originally Posted by Anaerin

I can think of a good reason why it won't - that old Sesame Street favourite: B-A-L-A-N-C-E.

You would have to ensure that both controllers output the same voltage/wattage all the time.

I still don't see a problem. Since the controllers would be connected in parallel, voltage is going to be the same. You would need to make sure both controllers get an identical throttle signal and then each controllers output will be matched. That is both controllers will be working to provide the same amp output and so should stay in balance.

The advantage of doing this, if it works, is that the 500 amp controller is a proven design. If it could be shown that 2 controllers can be run together then you have 1000 amps available now. When all the bugs are worked out of the 1000 amp controller then you have an easy upgrade path to 2000 amps.

[MPaulHolmes]

Originally Posted by mora

Is it possible to limit older design of 1000A controller down to, let's say 700 amps? In hardware protection limit I mean. I got the PCBs and almost all the necessary components for it but what I've read it is not be stable. I thought it might work with some downgrading. Or should I just get new PCBs when those become available?

It does seem a bit unpredictable at the moment. I have the one that was behaving erratically and I'm going to see what's going on with it. He said it started to act a little weird after a really hard acceleration. If I had to guess, I would say that the super high impedance pwm signals from the control board to the driver board is part of the problem. It seems to be working for another guy though. I wouldn't make it if I were you, but I think derating would probably work fine. It's just quite an expensive list of parts if you derate. One think I woul do is change to no thicker than a 1/4" thick M- bus bar. The 3/8" bar made it so the mosfet legs barely stuck out the other side of the pcb, which caused me to be concerned about how well they could be soldered.

[sawickm]

Originally Posted by Greg Fordyce

Since the controllers would be connected in parallel, voltage is going to be the same. You would need to make sure both controllers get an identical throttle signal and then each controllers output will be matched. That is both controllers will be working to provide the same amp output and so should stay in balance. When all the bugs are worked out of the 1000 amp controller then you have an easy upgrade path to 2000 amps.

You might want to check out Adam Brunette's EV Corvette Conversion:

http://ecomodder.com/forum/117833-post28.html and http://ecomodder.com/forum/119036-post31.html.

The ReVolt system is modular. You only need one (1) controller, for single or multiple driver PCBs and Power PCBs.

-Mark

[voltmatic automaton]

So a while back i finished building the controller and when I put it in it didn't work. So i shelved the project for a while. Today i hooked the controller up to a 12volt power supply (for the board) , a small compressor motor (12v) and a deep cycle battery also a 12 volt.

everything powered up ok and when i pulled the lever (on the pot) the motor spun!

so on a small test scale the controller seems to work fine.

put it in the car with the 120volt pack and nothing...not a sausage!

what could be the problem?

[Greg Fordyce]

Wiring problem on the car? What does rtd explorer show?

[jyanof]

Originally Posted by Greg Fordyce

I still don't see a problem. Since the controllers would be connected in parallel, voltage is going to be the same. You would need to make sure both controllers get an identical throttle signal and then each controllers output will be matched. That is both controllers will be working to provide the same amp output and so should stay in balance...

The power components (mosfets/igbts) will need to be switched identically in sync.

A rough hypothetical: Imagine that the two 1000A controllers are at full output, there is 2000A flowing in the motor loop, and the mosfets are on. If one controller's set of mosfets switch off before the other set, there will still be 2000A flowing, but now it's all flowing through one controller. poof! (The diodes only conduct when the mosfets are off and allows voltage at the drain pin to rise above B+)

Not that it can't be done, but I don't think it's necessarily as simple as wiring them up in parallel.

[pamcmills]

Hi all. I've been reading this forum for a while now and thought it time to chime in. On the subject of parallel controllers, the only way it would work is by a sync line tying them together. One would be designated as the master and the other as a slave. The elegant way would be to send a sync signal to an IO port and and add lines off code to establish a master / slave relationship between the two controllers. Another way would be to send the PWM signal of the master to the driver section of the slave and disable the micro-controller of the slave. That's my 10 cents worth anyway.

By the way, I'm the crazy person that Paul had mentioned a while back who tried to run the controller at 192 volts. Worked good for 300 miles before frying.

Paul. I have over 2000 miles since rebuilding the controller and going back to 144 volts.
Even with the 115F temp we've had here in Arizona It's been working great. The highest temp I have recorded at the heat sink has been 135F.

The last thing I want say is thanks to Paul and everyone involved with this open source project. I would not be as far along as I am on my conversion without the help of all of you folks out there.

[Greg Fordyce]

Originally Posted by jyanof

The power components (mosfets/igbts) will need to be switched identically in sync.

A rough hypothetical: Imagine that the two 1000A controllers are at full output, there is 2000A flowing in the motor loop, and the mosfets are on. If one controller's set of mosfets switch off before the other set, there will still be 2000A flowing, but now it's all flowing through one controller. poof! (The diodes only conduct when the mosfets are off and allows voltage at the drain pin to rise above B+)

Not that it can't be done, but I don't think it's necessarily as simple as wiring them up in parallel.

I think there could be another way to run 2 controllers in parallel. The 2000A flowing through one controller is actually just peak amps, the average amps through each controller is still 1000A in the above example. So if one of the controllers is seeing 2000 amps at any one point in time then the other is seeing 0 amps, and then they swap so the first one is seeing 0 amps and the second is seeing 2000 amps. If 2 appropriately sized inductors are added between the output of each controller and the motor (one inductor per controller) then the current will be smoothed out and keep the diodes happy in both controllers.

The controller and motor form a buck converter. Buck converter - Wikipedia, the free encyclopedia Fig 1 on the Wikipedia entry shows the basic layout of a buck converter. The field coils in the motor are used as an inductor, but there is nothing stopping you from using an external inductor. In my last EV I used an Agni PM brushed DC motor, and in the instructions it stated that if you used the motor with a Curtis controller (the ones that whine at low speed) you need to add an external inductor. The reason for this was the whine is caused by the low switching speed (about 1.5 khz) of the controller and since the permanent magnet motor had no field coils it could only use the armature coils to complete the buck converter circuit. At higher frequencies the inductance of the armature was sufficient. This suggests that the required inductance wouldn't need to be very great for running parallel controllers at 16khz, but I wouldn't know where to start with the calculations.

Buck converter - Wikipedia, the free encyclopedia Look at the layout of the multi-phase synchronous buck converter, one inductor per phase. In this example they evenly space the phases, that is they intentionally don't keep them in phase.

So I still think it is possible to use 2 standard controllers to run one motor, provided you add an inductor between each controller and the motor. The advantages would be using 2 lower powered, proven controllers to give you more power, and at a later stage you could sell one or both of the controllers when you move onto the next project, or a bigger controller. Remember, this crazy idea of mine was started by Jim and Darin's tractor project and their need of more amps. That got me thinking, they already have one good controller, shame they couldn't just build another one the same and connect it up.

Greg

P.S. Since getting my latest EV on the road in May, I have done over 2,000 miles with my rev 2C controller.

[sawickm]

Originally Posted by sawickm

The ReVolt system is modular. You only need one (1) controller, for single or multiple driver PCBs and Power PCBs.

Hey guys,

Just to clarify my point of a modular ReVolt system:

(1) 1k controller pcb, (1) 1k driver pcb, (1) 1k power pcb = 1000A

(1) 1k controller pcb, (2) 1k driver pcb, (2) 1k power pcb = 2000A

(1) 1k controller pcb, (3) 1k driver pcb, (3) 1k power pcb = 3000A

etc.......

-Mark