Last week we tried out the "ReVolt" open-source controller for the first time.  At the time, we only had one MOSFET and diode installed.  We were able to spin the small motor at 12V, but when we tried the kart motor at 48V, something popped so we stopped for the evening. This week, with some modifications and a more cautious approach, we were ready to try it again.

Cause of Last Week's "Failure"

First, some commentary on what went wrong last week.  When Bill took apart the power section, he really could not find anything wrong.  At the time we thought that perhaps there was a current spike through the diode as we applied power to the big motor.  But now we think that maybe it was the silver epoxy.  Perhaps there was a high resistance path somewhere that overheated when we started to flow some real current through it.


New configuration for the power circuit
New configuration for the power circuit

That brings me to the modifications for this week.  Thanks to Bill for doing most of this work.  Now we have 5 MOSFETs and diode pairs in parallel so we can handle more current.  Also, we got rid of the silver epoxy and used crimp lugs to attach the leads to the busbars with screws.  I think this is more secure and makes a better electrical connection.  Here is a photo showing this configuration.  The diodes are on the bottom.

Small Motor Test

Our procedure was pretty much the same as last week - start with a small DC motor and make sure that works first.  We didn't have the 12V bench supply this time so we used one 12V battery from the kart.  The controller  board was powered from the same battery.

We were able to spin the small motor from no power to full power - no surprise here.  We checked the temperature on the power electronics by feeling them.  We want to eventually install temperature sensors and add a temperature display to the controller screen.  Next we installed a panel current meter in series so we could observe the current in the motor (at least the average current).  We noticed a couple of things.  If you are a motor guy then I am sure this is not news to you but it is nice to see it in reality: first, before the motor starts turning, the current continues to rise (up to ~1A in this case).  Then as soon as the motor starts to turn there is a significant drop in current.  Second, when there is no load on the motor, the current does not seem to increase no matter how the PWM is varied.  We saw the meter stay about the same as I changed the PWM from near zero to near 100% duty cycle.

Big Motor Test

When I say "big motor" I am talking about the kart motor.  It is not as big as will be used in the car but it is non-trivial and can pull a lot of current.  See here for info about the kart.

Preparing for the motor test
Preparing for the motor test

We decided to start with 12V this time, instead of going straight to 48V.  As I started to apply power, Brian turned the kart wheel so the motor would not be initially stalled.  We did this because we suspected that last week the stalled motor pulled a lot of current until it started and maybe that caused us to damage some electronics.  In hindsight I don't think that is what happened last week and it was really not necessary to turn the kart wheel.  But it didn't hurt.

We used a clamp meter to observe the current in the motor.  At first we only increase the PWM to about 10% and then stopped to check on the power electronics and make sure nothing was getting hot.  Then we slowly increased the PWM until the motor started to turn at about 18% duty cycle.  This was further increased to about 25% and we let it turn at that speed for a couple of minutes, keeping an eye on the temperature of the power board.  There was some mechanical binding in the chain so we stopped for a moment to make a mechanical adjustment before continuing.  As the motor ran, wee also applied the kart brake causing the current to jump up to about 30-35A and held that for a minute.  Now we noticed that the busbar the diodes were mounted on were starting to get warm.  After that we continued increasing the PWM duty cycle until it was at 50%.

Now, for 24V

Doodling about the power circuit
Doodling about the power circuit

We stopped the motor and discussed things for a bit, then reconfigured for 24V.  The motor was supplied by two of the 12V cells in series, and the controller board by the lower 12V cell.  This was pretty similar to the 12V test except that the motor started turning at about 10% duty cycle, and when we got up to about 30%, the motor was probably turning faster than it was with 12V @ 50% (we didn't have a tachometer - add it to the list).  We ended up running it at a maximum duty cycle of 40% when we decided to stop.  As before the diodes were getting warm.  The busbar with the MOSFETs was also somewhat above ambient but not nearly as warm as the diodes.

We thought about trying 36 or 48V next but decided to stop because it was getting late.  Also we want to change the design of the power circuit and we didn't want to risk blowing anything up at this point.

Future design changes

We are going to try synchronous rectification next week.  Bill will replace the diodes with MOSFETs and I am going to modify the controller board to use a half-bridge driver.  The motor will go across the low side MOSFET and the high side will be used to supply voltage to the motor.  The low side will be turned on when the high side is off, acting as a low resistance in place of the diode.  Since the voltage drop will be less than the diode, less power should be wasted here.  At least that is the theory.

This week's episode ...

And here is a short video (~8 min, quicktime) summarizing the activities of the evening.

Controller Test 2


There are no comments yet.

Add a comment
You can use markdown in your comment. No personal information is collected or retained. Your email address will not be posted. This site uses static comments via email. That means there is no third party tracking you. I will get an email from you and I will see whatever email address you sent it from, which I don't care about. The email contents will then be converted to a static comment and appear here.