150V 45A MPPT - roll your own

Author Message


Joined: 21/06/2018
Location: Australia
Posts: 1066
Posted: 10:28pm 30 Mar 2020      

Their is essentially no difference between the power stage of half (one leg) of an inverter and the power stage of a buck PWM used for MPPT. However their is quite a difference when returning inductor energy during the PWM off time.between a synchronous PWM buck and a non synchronous PWM buck (flywheel diode).

Efficiency should rise from ~ 90% up to >95% and hopefully closer to 97+%. The Power stage I am using for my inverter with the cross coupled LED/Opto drive that Warpspeed has advocated for some time, does work a treat. With no changes to my PCBs I can either have a 1 x 8 parallel Mosfet PWM controller that should be good for output currents to 100+A or by cutting a couple of tracks 2 channels of 50+A that could be applied to 2 different input strings combining their outputs to a common battery.  Their is a large range of Mosfets to choose from in the 100 - 150V range with sub 10mOhm on resistance down to ~ 3mOhms depending on your budget and acceptable losses.

Using the HY5110 that Ben first posted on here with the power PCB configured as 8 in parallel would yield losses of 2.6W at 100Amps standing, so switching could be ~ 5W total loss for strings under 100V @100A out. ($3 - $5ea on Ali Express)

Using a FDH055N15A configured as 8 in parallel would yield losses of ~7.5W at 100Amps standing so switching could be ~ 15W total loss for strings >100V but < 150V @100A out. ($7.30 - $8ea on Digikey)

Assume a switching rate of 20kHz, a good starting point for experimenting would be as below.

For strings of Vin 60 - 100V & Vout 55V using 8 x 100V Mosfets in Parallel could either use a common larger choke of ~ 30uH  or 2 smaller chokes ~60uH.

For strings of Vin 100 - 150V & Vout 55V using 8 x 150V Mosfets in Parallel could either use a common larger choke of ~ 45uH  or 2 smaller chokes ~90uH.

Using the topology I adopted, galvanically isolates the analog section from the Power section using isolated supplies and opto coupling the drive waveforms from the nano.
Edited 2020-03-31 08:32 by wiseguy
If at first you dont succeed, I suggest you avoid sky diving....
Cheers Mike