Great to hear from you. First off thanks for the comment, I think it’s really great that we are both working on our own custom trackers.

Yeah the Inductor is still the biggest problem efficiency wise with my tracker. I got rid of a lot of the switching loss in my REV 1 tracker with the new REV 2 tracker so now almost all of the loses left in the tracker are in the DCR and the core loss of the inductor. That’s what I plan on tackling next, most likely with a custom wound inductor. I’m really not sure what the new efficiency is because most of our good instruments are over in Australia.

We are going to be running my REV 1 trackers over in Australia during the race as I haven’t been able to fully test the REV 2, though the design for the REV 2 is more or less complete.

My REV 1 MPPT switches at 30KHz. I have found that as switching frequency is changed the slope of the boost factor vs efficiency curve changes. The curve is flatter at lower switching frequencies and steeper at higher switching frequencies. I chose 30KHz as the best compromise between peak efficiency and a flat boost/efficiency curve.

When the REV 2 is completely finished it will most likely actively switch between a set number of switching frequencies and this should allow it to run in efficiencies around 96-97% ((a guess) If I kept the 1140 inductor)) at low boost factors and still retain a pretty flat efficiency curve as boost factors go up. The best efficiency numbers I’ve gathered from my REV 1 tracker thus far are a little below 96% efficiency at 5 A in and at boost factor of 4 with the 1140 inductor. I haven’t bothered testing it at lower boost factors yet as I don’t expect us to see boost factors lower then 4 on Axiom.

http://calgarysolarteam.ca/blog/?p=52 you can check out my efficiency plot of my REV 1 MPPT here. You can see mine doesn’t peak as high as yours but is better at higher boost factors.

I did have switching problems with my initial designs a very long time ago due to trace inductance, but I’ve never experienced any radiated EMI problems like you describe. I’ve had a radio and computers nearby and have never noticed anything out of the ordinary. I’m not an RF guy by any means, but I imagine it’s probably due to the small size of my circuit. It’s about the size of a wallet so there isn’t really any trace long enough to throw off much EMI. It could also be that the switching frequency of my tracker is quite low for a SMPS, I’m really not sure.

-Tim

I’ve done some testing with those 1140 inductors from Bourns and I’ve never had great luck efficiency wise. I’m curious what kind of efficiency you can achieve with them, and what switching frequency you’re using with it. Have you had any EMI issues with them since they’re non-shielded? When I run it on my bench it will knock my PC keyboard off of the USB bus. I can get better then 99% with my custom toroids, but I could never get better then 94-96% in the same circuit with the 1140.

-Sam

Solar Tracker