PLEASE NOTE: If you had an account with the previous forum, it has been ported to the new Genetry website!
You will need to reset the password to access the new forum. Click Log In → Forgot Password → enter your username or forum email address → click Email Reset Link.
Once again...I'm seeing the same issue on the readings: solar array voltage within 1-2v of battery voltage (instead of being 2-3x battery voltage). Depending on your solar array setup, this means you are easily losing over half of the available power.
On most of those photos, the PWM throttle is maxed out--meaning that the "MPPT" is just doing a straight pass-thru of the power, with no conversion being done. (Peeked in the "wiring_analog_extras.cpp" to see that the basic library of course has a lot of "inefficient conveniences" in it...so the "value" is MAPped to the "TOP" scale. Makes it easier for users to understand--but you'd better believe that MAP is a slow function!)
I think my final design will be to run mppt with multiple pwm freqs ONLY up to a specific BATTi current then admit the torroids are saturated and just use minimum pwm freq and basically run pwm.. thats where I am headed..
When you run PWM, you're cutting the available power in more than half (assuming 80v solar input and 30v battery)--so of course things will run cooler.
If you remove some of the abstraction layers (including the above "MAP"), it is actually much easier to control PWM with varied frequency--and keeping the pulse width from exceeding the saturation value (i.e. Value = pulse width--code limited to the maximum allowable for the voltage range, while Top can be separately adjusted to change the frequency without changing the "on" pulse width). But that starts to get quite technical....
...once the "bucket" (inductor) is full, it can't carry any more water. So in order to move more power in a given time, you have to move the bucket faster--but without overflowing it 😉.
yeah I avoid float math like a plague..even at 48mhz, it hurts..I hope u are not stuck at 20mhz pic..that used to be their 'fast' stuff..float at 20mhz would be a deal-killer..
It actually runs at 64MHz, though with a 4-cycle core...so 16MIPs. I'd be so much better off with a cheap 8051-1T core Nuvoton processor...easily running 24MIPs for a lot cheaper. All peripherals would be better EXCEPT for the ADC (analog to digital converter)--the PIC has a pretty solid essentially 13-bit ADC on it.
Once again...I'm seeing the same issue on the readings: solar array voltage within 1-2v of battery voltage (instead of being 2-3x battery voltage)
Maybe the solar panel is shaded and only 43 v. Maybe the solar panel are defective and only 43v going to the MSB . IF only 43v then the battery cannot be set for 30v but need to be set and use for 12v battery . My MSB is set for 66 volt and will not work if solat panel is shaded and is 80 volts . Something not right .
if you remove some of the abstraction layers (including the above "MAP"), it is actually much easier to control PWM with varied frequency--and keeping the pulse width from exceeding the saturation value (i.e. Value = pulse width--code limited to the maximum allowable for the voltage range, while Top can be separately adjusted to change the frequency without changing the "on" pulse width). But that starts to get quite technical....
I like the 'bucket' explanation!!..I think thats what appears to be going on..very small bucket-- (torroid)..so when I boost pwm freq, things work better and heating also goes up..
I rewrote a scan routine that simply goes thru every pwm freq with an appropriate bucket size and display results.. then in mppt() I just dither the bucket size a little to try and maximize power..
will see if this goes anywhere today..
so when I boost pwm freq, things work better and heating also goes up..
Yeah, the more power you're dealing with, the hotter things will get.
It's only if you were getting into severe switching losses that heat would start to really get out of hand. But with a TLP350 on a pair of FETs...I think you'll be quite fine.
The main thing to watch is the panel voltage...if it's getting pulled down to battery voltage, you're losing a heap of available power.
more testing...
//content.invisioncic.com/g308908/monthly_2022_07/image.thumb.jpeg.6e5466ff5d8fa50d7f702f8e7da85429.jpeg
real world scan of pwm values shows multiple peaks.. this is probably because I have paralleled two different stacks of panels, each with different characteristics..
I reduced the number of scan freqs down to a minimal 3.. like low,medium, high.. it was more of a waste of scan time to go thru 7 or 8 different frequencies..the toroids dont seem to really care all that much what freq they switch at.. so I have gone with L-M-H of 25khz, 40khz(this was the original msb freq used only), and 60khz..under some conditions, it will select 60khz over 25khz, etc..
I did what Sid said to do-- got rid of the map() on the pwm and just go between 2 at the low end and TOP-2 at the high end.. freq ends up being just the TOP value..hence pwm= 2 to freq-2
I also got real and implemented an interrupt routine--one that doesnt conflict with the standard delay()..so now its clean and simple to watch for keys hit while running any routine..
I added multiple procedures to generate different frequency ranges for the toroids, but this appears to be basically a non-issue..
//content.invisioncic.com/g308908/monthly_2022_07/IMG_20220717_112846.thumb.jpg.77d5c38477cb649f5a0599e84e52a3dc.jpg
//content.invisioncic.com/g308908/monthly_2022_07/IMG_20220717_112831_9CS.thumb.jpg.04e7e52fd60afd4e65aae7e6c1543c79.jpg
<a class="ipsAttachLink" data-fileext="zip" data-fileid="1471" href="/applications/core/interface/file/attachment.php?id=1471&key=1442450aefc248d208cd7f93ba6bb300" rel="">makeSkyBluer.zip
That looks a lot better......PVv is about double from BATTv, so it's actually doing some power conversion.
Still I would expect to see a lot more of a "power curve" from the panels...but I haven't done any real-world testing to see for sure.
That looks a lot better......PVv is about double from BATTv, so it's actually doing some power conversion.
page 2 of msb's installation flyer says 'Voc of PV should be in the correct range and when its 1.5 or 2 times than the battery voltage then its best efficiency'
max PVv is limited to 160v..
so their statement makes not alot of sense on systems where battery voltage is a low 12v or 24v or even like me 33v.. 33 x 4 = 132v..
my next testing will be to add another solar panel in each branch of PV supply..that will add anoth 40v give-or-take due to temperature..
msb is the most basic asynch buck circuit.. the reverse clamp rectifiers from switching to gnd are basic hi-voltage dual to-247 rectifiers..NOT Schottky.. diode Vf likely 1.2v or MORE..(2 silicon junction to get to 160v tolerance)
so efficiency can never be as good as more advanced designs-- like synchronous and beyond..
I have to hoist up several large solar panels to my solar shed roof.. will take a few days..but I am looking forward to better performance by the increased PVv headroom..
page 2 of msb's installation flyer says 'Voc of PV should be in the correct range and when its 1.5 or 2 times than the battery voltage then its best efficiency'
I have 2 64v solar panels in parallel and the battery is set at 32v . My reading is 62v solar panel out and 30 v in to battery and 236 watts is about the best I get for my 24v system . Your reading is probably the best you get for 24 v system as your battery is 29v and 222 watts . Adding 40v solar panel will allow you to go to 36v system I think . Your MSB is working perfect for 24v system . I get sunlight 12 hours a day now .
I have 2 64v solar panels in parallel and the battery is set at 32v . My reading is 62v solar panel out and 30 v in to battery and 236 watts is about the best I get for my 24v system .
if u look at ur solar panels on the label it should tell u the Voc and the maximum power out voltage..
and current at that voltage.. lets do the math and see how much u are actually harvesting..
and current at that voltage.. lets do the math and see how much u are actually harvesting..
64 VOC Sunpower panel and I think the max is 7 amps for this MSB 40 A MPPT . My MSB 60 A MPPT get twice the wattage and over 500 watts most days at 66v battery setting but the solar panel is in series at 120 volts .
msb is the most basic asynch buck circuit.. the reverse clamp rectifiers from switching to gnd are basic hi-voltage dual to-247 rectifiers..NOT Schottky.. diode Vf likely 1.2v or MORE..(2 silicon junction to get to 160v tolerance) . I have to hoist up several large solar panels to my solar shed roof.. will take a few days..but I am looking forward to better performance by the increased PVv headroom..
My 2 solar panels output 100 volts and 9 amps to the MSB 60A which is 900 watts and the MSB 60A send 66v and 550 watts to the battery . The best is 60 percent from my MSB 60A but I have 12 hours of sunlight . That is not enough to run a 12kw inverter for more than 8 hours during the day but I have battery to run at night . The MSB 30A is useless and my MSB 40A is only use for 24v system charging the battey to 29vdc max .
and current at that voltage.. lets do the math and see how much u are actually harvesting..
the wattage spec is a clever way to provide bigger numbers simply because of the higher battery voltages..looks impressive.. but the truth is that the real amount of 'work' that the msb can do is the same regardless of the battery voltage..7 to 9 amps is about all they can do..for the 30A and 40A models..the 60A model i suppose can do twice that..
the 60A model i suppose can do twice that..
Yes the MSB 60A is the one that is useful since my inverter run at 60v most of the time . I only connect 2 solar panel is series so less heat and maybe last longer . The instruction say 4 panels in series and parallel is possible . My MSB get 5 to 7 amps from the solar panel most of the times and is not very much but I have shading from 7 trees .
the wattage spec is a clever way to provide bigger numbers simply because of the higher battery voltages..looks impressive.. but the truth is that the real amount of 'work' that the msb can do is the same regardless of the battery voltage..7 to 9 amps is about all they can do..for the 30A and 40A models..the 60A model i suppose can do twice that..
The max PV limit rising with battery voltage is an interesting one. I can see it being possible, nothing special there at all really, but the way it scales is a bit fishy. VOC = 80V for 12V, 105V for 24V, 160 for 36V, 160 for 48V. I haven't looked at the circuit in detail for any MSB charger TBH, but a schematic would be interesting to look at to see how exactly things are arranged and how that range of voltages would be possible or required at all for that matter.
the wattage spec is a clever way to provide bigger numbers simply because of the higher battery voltages..looks impressive.. but the truth is that the real amount of 'work' that the msb can do is the same regardless of the battery voltage..7 to 9 amps is about all they can do..for the 30A and 40A models..the 60A model i suppose can do twice that..
Well, the reason there's bigger wattage numbers at higher battery voltages...is because that's just how it works.
Max 60A?
If you do 12v * 60A = 720W max
If you do 24v * 60A = 1,440W max
If you do 48v * 60A = 2,880W max.
If generalized to mechanical levels: "Work" = watts. "Volts" = speed. "Amps" = torque.
The issue is not "how much work can be done." It's a matter of amperage, which is what generates heat. Doesn't matter if it's 60A at 1v (60W of work) or 60A at 100v (6,000W of work), the same amount of heat will be generated (as long as switching losses are ignored for the purposes of the point here).
The amperage limitation is most likely mostly a limitation of the inductors.
My personal 2 cents is that the "7-9A" number from the modded MSBs has to do with not quite running the inductor just right.