Filed under: Charge Controller.
My request for help to the Free Charge Controller mailing list was answered by many generous people. Donald H, a member of the group, pointed out that C2 was being overworked. Below are the results of trying different values for C2.
The gate droop issue that was plaguing the free charge controller turned out to be due to C2 being too small of a capacitance. Tim Nolan’s original design showed a C2 value of 0.1 uF, which I copied. You can see C2 in the drive circuit here:
I want thank the many people in the group who contributed suggestions to help me fix this bug in the charge controller. A few of the people who offered suggestions were Karolina Lindqvist, Donald H, Boon, and Troy Benjegerdes.
After changing the values of C2, I took a scope image of the gate voltages on M2 and M3. Here you can see the gate droop on M2 with a 50 KHz signal:
The signal above should be a 50Khz square wave with a 75% duty cycle. However, the ‘square’ wave takes a nose dive right away.
Here is the signal again after upgrading C2 from 0.1 uF (micro farads) to 0.43 uF:
A significant improvement! There is still a significant droop on the square wave, but I was very encouraged by the improvement in the signal.
The capacitors I grabbed were ceramic capacitors. Luckily, I had another 0.33 uF cap on hand that I was able to put in parallel, to bring the capacitance of C2 up to 0.76 uF:
There is still a rounded corner on the edge, but this circuit now appears to be functional. In the future, I think the magic number for C2 is 1 uF.
This next screen shot shows the output of the charge controller. You can see that it’s outputting a steady 12.4 volts from an 18 volt input. Again, the drive PWM signal is 50Khz with a 75% duty cycle. Increasing the duty cycle would increase the output voltage.
Again, thank you to everyone who contributed to helping me debug this circuit. I know it’s difficult to help someone with a circuit when all you have is a schematic to go off of.
I’m going to continue to test the v4 prototype and report the results. I’m also working to finalize the BOM so that Jameco can begin to offer it as a kit.