What is the micro controller in a MakeSkyBlue MPPT?

the 7 to 9A is what is reported by Dickson-- un-modified msb.. I have modified 2 30A msbs --wound my own toroid-- FT-240..bifillar so just one toroid ..and I get much much amps..I have both hanging on the wall and the un-modified one runs much cooler but is limited in max current to around what Dicckson reports..7 or 8 amps.. my FT-240 roars along 12amps, 15amps...I had to add a throttle to the mppt routine to cut back on power because it will overheat and/or fry the FETs if left unregulated..

 if(degHOT>32)PWM=peakPWM/4;
 if(degHOT>34)PWM=2;
 if(WATTS>maxChargeWatts)PWM=PWM-(PWM/8);

 if(mode=="MPPT"){driveFETs(f_idx,PWM);}

I think what msb has done was actually quite clever-- the reason they state the reduced VOCs is simply to mask the fact they have used too-small inductors.. so they openly state so without saying that..this limits the heat tremendously and likely saves many units in the field that get hooked up to too large a solar array that would otherwise cause damage..this is clever, actually..

to run my FT-240s, I have added finned heatsinks to the back side--not easy to do-- and made the msbs to each fit into the base of a chimney the width of the msb..

the 7 to 9A is what is reported by Dickson-- un-modified msb.. I have modified 2 30A msbs --wound my own toroid-- FT-240..bifillar so just one toroid ..and I get much much amps.

Your inductor is huge .       I use the MSB 60A  as it can get  66 vDC  as most MPPT can not do .   Most of the  time it is 5 Amps    but  12 hours of sunlight  helps .      

so they openly state so without saying that..this limits the heat tremendously and likely saves many units in the field that get hooked up to too large a solar array that would otherwise cause damage..this is clever, actually..     You are right about the heat  so I  only   connect  2 solar panels  instead  of  four .        Thank you  for your hard work .    

 I have 3 msbs running.. the left one is original toroid.. the middle and right units have a FT-240 I wound..

left unit maxxes out at 7 or 8 amps.. center unit does 12 to 15 no problem..just runs hotter  both msb have exactly same solar panels.. 2 each in series.. same brand,same size,same lot..

the original msb (at least the 30A model) is designed to have an under-weight inductor apparently to limit maximum heat/stress on FETs, etc.. on my modified msb, before I throttled the watts I saw it read out 700+ watts..and the heat was up to 83 degF OVER when the fan comes on..it was set at 90 then..it was hot enuff to cook an egg.. smelled a little toasty.. so I added throttling.. working nicely so far..

what first caught my eye on these msbs was that in full sunlight with plenty of solar panel wattage available, they wouldjust plod along at a few hundred watts.. so what I did was simply direct short between PV+ and BATT+ and watched my external amp meter on my battery stack.. the amps jumped up!!! just a few amps, but regardless, this couldnt happen unless the msb was choking the solar power available.. and it does..my FT-240 replacement proves this.. even thought the original msb appears to be mppt-ing, it is also limiting the upper limit of amperage..which is OK in a way, but I need to harvest all the solar I can within the safe limits of my equipment..so I cheat..I am gonna wind a whopper of a toroidand connect it externally from the msb..not enuff room inside for ahuge toroid.. it may work better and produce less heat and transfer alot more amps..

.I am gonna wind a whopper of a toroidand connect it externally from the msb..not enuff room inside for ahuge toroid.. it may work better and produce less heat and transfer alot more amps..

Maybe add a  big external  fan  also ?            

if I mount it to the bottom of the msb, it will be in the 'chimney'.. no need for a fan (I hope)..this bad-boy is hefty..I have no idea what the material is.. just experimenting with what I have around..

this large toroid produces more than the FT-240s do..considerably more amps/watts..

I doubled up the windings--thus reducing the DC resistance to half of that of the FT-240s..used same wire,just made double(parallel) windings for both FETs..

it produces a nice power pwm curve at the same freq as the FT-240s, et all..(40Khz)..I havent tried other freqs.. dont really need to..its working just fine at 40Khz..

I think in conclusion of all my toroid madness is this-- the larger the toroid, the greater the power handling transfer and possibly higher efficiency..I havent done enuff detailed study to prove the real efficiencies for each toroid, however, the increased power transfer as a simple function of size is more than obvious..

the 60A msbs already have much larger toroids than the 30A model.. so I dont plan on messing with them..they should be alright the way they are..I just want to run my software on them..I have one right now I am converting....  

That's looking a LOT more like an MPPT curve...though not falling off as quickly as I'd expect on the top end.

Remember, there's 2 ways to transfer more "stuff" (energy)...either move the bucket faster, or use a bigger bucket.  Replacing the inductor serves the latter purpose.

And who knows, the stock one might well be undersized.  I just haven't dug into it.

I looked into moving the bucket faster.. this proved out to be not valuable with toroid inductors, at least of the types I have used..the toroids greatest claim-to-fame is that they are wide-band frequency tolerant inductors..great for blocking broad-band noise,etc.. well trying switching frequencies from 60khz in various steps down to 5 khz what I found was this: the switching freq doesnt correlate to a bucket-rate.. there must be more factors involved, of which I dont pretend to exactly understand..except that 'resonance' is not achievable and therefore 'dip-the-plate' thinking is not appropriate to toroid inductance.. gets kinda confusing..I wrote three procedures to generate

three different freq ranges for the toroid switching freq.. 25 to 60khz, 12 to 20khz, and 5 to 9khz..

took me a couple weeks or more to ramble thru all that.. and make it reliable to do on-the-fly..

turned out in the end to be a rabbit trail..

msb is the very simplest mppt.. asynch buck.. I noticed that efficiency appears to be much higher at low amps, on all inductors I used.. when the current gets heavy-- over 6 amps or so, efficiency goes down from the great efficiency before at lower amps..frustrating..if I couldmake it resonant I might keep the efficiency high..havent been able to do that..I appreciate ur thoughts..it may require synchronous active rectification to do any better..thats beyond where msb or I will go with msb hardware..

regardless, I am pleased with FT-240 performance.. FT-240 cores are about 10bucks.. so affordable for the boost in performance and pretty easy to wind..otherwise, just buy the 60A msb..what amazedme was that msb apparently even in their latest software versions do NOT save user variables to flash.. umbelievable.. turn breakers off, backon..u get to cycle thru all the setup again, over and over each time u killpower.. thats what causedme to write my own crap.. 

 turn breakers off, backon..u get to cycle thru all the setup again, over and over each time u killpower.. thats what caused me to write my own crap.. 

My  breaker to the battery  is  always on  once the setup is done .       I  can switch the  power to the solar panel  off and  on  if work need to be done .    Have to  make sure the battery  never get  discharged  or bad thing happen .   

Following this with interest. Don't know the firmware version you have but I have 3 off v118 units and they maintain settings through power cycle. Will have to try and look at their efficiency one of these days.

I started out with 3 30A msbs on the wall..dont know the software version of any of them..I never saw anywhere I could display that info..but what happened was every morning 2 of them would refuse to start a new day of charging.. fault icon blinking, stuff like that-- no charging.. so had to reboot.. reboot and lost all settings.. had to clunk clunk thru a really slow interface..very frustrating..

finally just gave up and did what I did.. very glad I did.. now I dont have to kick-start the dumb things in the morning and they put out twice as much 'juice' overall..lotta work but was worth it to me..

my software is free to all for starting point.. my pc board will be available on oshpark.com soon..

or I may offer to do the mod for people who would rather have it done for them..I still want to test for a while longer and arrive at a stable point..

On 7/17/2022 at 4:38 PM, dickson said:

 

  https://oshpark.com/shared_projects

<a class="ipsAttachLink" data-fileext="zip" data-fileid="1558" href="/applications/core/interface/file/attachment.php?id=1558&key=314d6873d525553a8b57f4ab542d4e6c" rel="">makeSkyBluer-master.zip

I hope MakeSkyBlue will take my initial work and just use it  and go add the rest of the needed stuff -- like logging, wifi, etc.. they are using basically the same chip--a derivative of cortex M0, just with a slowed down 5v wrapper on it..they could jump-start a new version of msb and only need to get rid of that horrible out-dated display and replace it with a nice readable OLED 1.5"..

Adafruit QT Py has a wifi version..and underside of the QT PY has placement for a 2M ram for data collection..GD25Q16

Hi Guys, new here...

I have three of these 96v 30 amp MSB controllers, two are dead, and all three have totally different circuit boards in them. That makes tracing out three different schematics a real pain, as all the component designators and locations are different.

Also, the mppt software is a bit suspect, as has already been noted here, it sometimes derails and locks up into some funny mode where solar voltage is only a very few volts higher than battery voltage. And that cannot be right.

Have to agree, the MSB chokes are very small, no way the wire gauge used can carry the rated thirty amps without almost being on fire.  So have abandoned fixing these three solar controllers for now, but may revisit the whole disaster later on.

About suitable alternative chokes. 

At least for the higher 96 volt units, if you wish to use off the shelf chokes, commercial light dimmer chokes of suitable current rating work fine for me in this application in my own home brew mppt controller.  These chokes are normally used in series with high power SCR light dimmers to reduce conducted EMI. 

They use sensibly rated thick wire, and have a high turns count which reduces high frequency core heating.  120 volt rated dimmer chokes are usually 0.5mH, and 230 volt rated dimmer chokes 1mH which is very high for a buck regulator.

That  hugely reduces ripple current in the electrolytics. It would be very difficult to hand wind chokes as neatly as the machine wound dimmer chokes with such heavy wire. They are very large though, but work well for a higher voltage pwm buck type solar controller.

Curious what switching frequency you're running on your homebrew MPPT?

The only disadvantage I'm aware of with the larger chokes would be a higher DC resistance...but the increased inductance makes everything else easier!  And I wouldn't be surprised if said chokes had a LOWER DC resistance than the MSB chokes...

FWIW while just thinking about MPPT/buck methodology (but not actually testing anything), I've wondered if the MPPT frequency actually would be determined by the voltage drop between input and output?  Maximum pulse width kinda would be limited by the saturation rate of the inductor--meaning that the only way to move more power would be to increase the frequency (up to a point, obviously!)  In other words, MPPT frequency would not be a constant, but rather a dynamic variable...

And....pretty sure, but just want to know...you're the same Warpspeed from TheBackShed forums?  Should be quite beneficial having some more brains on board, even if I might have to flip everything right-side up before I can understand it 🤣🤪.  (Sorry, I just had to...)

Admittedly, this isn't too uncommon.  Sean has been testing some AmpInvt 120A MPPTs...and in low light, they tend to do exactly just that.  Which of course is a double-whammy for cloudy weather: not only are the panels not generating much power, the MPPT is throwing away over half of THAT simply by holding the panel voltage too low.

From a software standpoint, this pretty much means that the panels are just straight-through to the battery, less only the forward diode on the MSBs (as they aren't synchronous buck converters).

I personally have been using 2 Epever Tracer 8420AN units (in addition to a venerable but pricey Morningstar Tristar TS-MPPT-60).  I did some very low-light tracking tests on the Tracers, and they correctly "found" the max-power voltage each time...sweeping from Voc down to battery over 2-3 seconds, and then "dropping" back to the calculated max power voltage.  Conversely, the AmpInvts would "sweep" the whole array voltage range (Voc down to battery voltage), and then just "leave" the panels at the battery voltage.