What is the micro controller in a MakeSkyBlue MPPT?

Clarification needed here-- there are only 2 transformers on the pcb.. the larger one is of course the switcher PS.. the little one--I referred to-- isjust a very small signal transf..4 pins..2 windings.. it just has switcher HF on the input and uses that as a signal that we have PS working onboard so let PV in..

I started here--

https://datasheet.octopart.com/TLP350(TP1,F)-Toshiba-datasheet-148055.pdf

ther are 2 TLP350s.. 2 (too-small) inductors and 2 sets of doubled up(parallel) FETs..

pin 5 of BOTH TLP350s  are dead short together..these are also dead short to the inductors..one side..

however, as u suspected, there is a separate supply for each pin8 (V+supply)of TLP350 from separate PS diodes from the switcher transf..on this board they are D17 and D16..but they are same polarity and same configuration..just redundant..so there is NOT isolation between both switching FETS, only redundant PS on pin8s.. 

I will have to break one apart and look at the bottom side to know anything more..

"Redundant power supplies on pin 8" is all that you need for running the FETs out of phase.  Betcha the TLP350 pin 5 individually go to the respective FET source leads, which then go to the separate inductors--and only on the OTHER side of the inductors will they be tied together at the battery output terminal.  They appear to be shorted together from a DC perspective--but from a switching perspective, they should be separate.

If you can confirm this on the PCB, then the unit has the hardware ability to be able to run the 2 FETs out of phase with each other.

OK, that would make a lot of sense.  Sounds like they're using it as an isolation device for the PV input high-side FET drive.  Clever if I say so myself 😉.

I cant believe they went to all that trouble, but maybe they had to..people connecting or switching the PV breaker first and causing smoke somehow..probably..

"Redundant power supplies on pin 8" is all that you need for running the FETs out of phase.  Betcha the TLP350 pin 5 individually go to the respective FET source leads, which then go to the separate inductors--and only on the OTHER side of the inductors will they be tied together at the battery output terminal.  They appear to be shorted together from a DC perspective--but from a switching perspective, they should be separate.

If you can confirm this on the PCB, then the unit has the hardware ability to be able to run the 2 FETs out of phase with each other.

I'll break this unit apart tmro and get back..

sunset... when it gets down to pip-squeak charging, the peak charging watts moves way up the voltage range.. but this only happens at a few amps charge or less.. 

sunset... when it gets down to pip-squeak charging, the peak charging watts moves way up the voltage range.. but this only happens at a few amps charge or less.

Very interesting  .         I do not know why .     

Ummm.....sounds like your "MPPT" algorithm needs some serious adjustment.

Solar panels are constant-current devices.  You can get 5A @ 72v or 5A @ 36v...and obviously, you get twice as much actual power at the higher voltage.

If you have a 80v VOC (volts open circuit) rated array, max power point isn't going to be much lower than 55v in full sun.  It definitely isn't a going to be a very suspicious 0.8-1.1v above your battery voltage.
...here all along, I thought you had a 40v panel array.  But if you really have an 80v array or something like that...then there's a lot of available power simply not getting captured.

Like I mentioned a bit earlier, if you're running the PWM frequency too low, you basically have a PWM charger--which will basically run the panels close to battery voltage.  Less heat is generated, but that's because no power conversion is taking place.

For reference, I have a ~120v open-circuit array on a Morningstar Tristar TS-MPPT-60.  It normally runs 85-96v when at max power point (i.e. when charging the batteries in the morning).  Right now in the definitely declining daylight, the panels are at 80v...where the max power point is.  The only time I ever see the MPPT input voltage near open-circuit voltage is when the MPPT is throttling back to avoid overvoltaging the batteries.

For reference, I have a ~120v open-circuit array on a Morningstar Tristar TS-MPPT-60.  It normally runs 85-96v when at max power point (i.e. when charging the batteries in the morning).  Right now in the definitely declining daylight, the panels are at 80v...where the max power point is.             

My  MSB  is  120 VOC  also and is doing almost exactly  like your Morningstar .     What a  surprise  for 100 dolars MPPT .      

 Ummm.....sounds like your "MPPT" algorithm needs some serious adjustment  ??    

results of msb proctology exam---

I removed the pcb from a 60A msb that has a bad switcher PS IC (uc3845B)..so had to remove the pcb anyway..

I made a general schematic of what it appears to be..

compare the inductor size to  another CC I have ..it is rated at only 30A yet it is much much larger and heavier than msb's 30A inductors..

Nice work on the schematic!

Strange...the "PV input" disconnect FETs are P-channel?  Those generally are more costly + higher Rds when compared to a similarly-rated N-FET.

I dont want to try and bend one up to get a part# might break..but PV in directly goes to the typical 'emitter' terminal of the to-220.. center 'collector' then goes on to the switching FETs.. emitter same as drain..i get drain and supply confused..so I just referred to them as emitter and collector for clarity..

they are FETs ..no diode junction showing..

Mr. Sid-- I was swetting trying to deal with boosting the 20khz to 40khz, but just now I seem to have done that with just one line of code..

void analogWrite20kHz(const uint32_t pin, uint32_t value)
{
  /*
    From the SAMD21 datasheet, page 642:

                  f{GCLK_TC}
      f{PWM_SS} = ----------
                  N(TOP + 1)

    Where:
    - f{PWM_SS} is the PWM frequency.
    - f{GCLK_TC} is the clock frequency (48MHz, in this case).
    - N is the prescaler value (16, in this case).
    - TOP is the max counter value (149, in this case).

    This gives f{PWM_SS} = 48MHz / (16(149 + 1)) = 20kHz.
  */

  //const static int TOP = 149;
  const static int TOP = 75;

  PinDescription pinDesc = g_APinDescription[pin];
  uint32_t attr = pinDesc.ulPinAttribute;
 

changinging the TOP value from 149 to 75 exactly doubled the pwm freq generator on my dso..

and I installed the new firmware on two msbs.. looking good.. somewhat better numbers..

I will monitor today and report back.. thanks!!

What kind of DSO do you have?  It would be pretty easy to see if the PWM frequency is too high/too low by checking input ripple.  If the inductor is saturating (i.e. frequency too low), you should be able to see that with sharp ripple on the solar input.

Seeing a bit more of a difference in numbers now, but still not quite high enough for true MPPT tracking, I don't think...