diy MPPT data

I figured I'd quit hijacking that poor guy's thread, and post data here of my diy MPPT tracker. Here's the latest picture:
I waited until like 4:45PM or so before turning on all the panels to keep the amps down. I'm still testing it out, and don't want to blow up my battery bank. There are 20 280W panels in 4 strings of 5 panels per string. This one started at around 53 amps output to 48v, and the sun was gradually setting, so the power (and amps) kept falling. Then my stupid computer fell asleep after me not touching anything for some time, and it woke up with much lower amps because the sun had gone down some. The short spikes of lower current were when a couple clouds drifted past the sun, but they weren't enough to really block it much.
ampsvstimeAllPanels4-30PM.jpg

Comments

  • drraptor
    drraptor Solar Expert Posts: 218 ✭✭
    Re: diy MPPT data

    I was actually learning a lot from your post. As I'm going to buy or DIY a MPPT charger in near future InsALLAH.


    Just change "put computer to sleep" to never.
  • bmet
    bmet Solar Expert Posts: 630 ✭✭
    Re: diy MPPT data

    In the service I once ran the calculations for a navigational 'Schuler Loop' (F-4s used analog computers). That was enuff. Now, I happily buy the tools to do that work :D
  • MPaulHolmes
    MPaulHolmes Solar Expert Posts: 35 ✭✭
    Re: diy MPPT data

    What's this?? More data? haha. I finally got the transition from MPPT to constant voltage mode working pretty well I think. According to my fluke multimeter, there's a fluctuation of the "constant voltage" of about +/- 0.01v. Good enough for government work. I dumped one of the 6v batteries, since my ... long story. I'd really rather not talk about it. HAHA.

    MPPTConstantVoltageMode7-11-2013.jpg

    What I have it do is, if the voltage drops below the "constant voltage point", a timer starts. If it stays below the constant voltage point for a certain specified time, it switches back to MPPT mode.
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: diy MPPT data
    I dumped one of the 6v batteries, since my ... long story. I'd really rather not talk about it. HAHA.

    So, this graph is done with only 7 batteries, thatr is 42V battery bank?

    That's the advantage of custom controller. If you remove one battery, you can re-program it for any voltage of the bank. Must have a custom inverter too :D
  • MPaulHolmes
    MPaulHolmes Solar Expert Posts: 35 ✭✭
    Re: diy MPPT data

    Yep, 42v battery bank. Sort of odd ballish, but you are right. Just change the voltage stopping point to whatever you want, and hit compile! haha. Or just make it a programmable setting... I should do that. Ya, it's also a custom inverter too.
  • drraptor
    drraptor Solar Expert Posts: 218 ✭✭
    Re: diy MPPT data

    so what I will need for a 30A MPPT ?

    A pic controller, PWM kit etc ?
  • MPaulHolmes
    MPaulHolmes Solar Expert Posts: 35 ✭✭
    Re: diy MPPT data

    I don't think you could do a kit that's cheaper than what's available for 30A. The savings happen when you really crank the current and voltage up to hundreds of amps and hundreds of volts. Actually, everything gets easier. You use a sledgehammer for every problem that a clever engineer could work around.
  • drraptor
    drraptor Solar Expert Posts: 218 ✭✭
    Re: diy MPPT data
    I don't think you could do a kit that's cheaper than what's available for 30A. The savings happen when you really crank the current and voltage up to hundreds of amps and hundreds of volts. Actually, everything gets easier. You use a sledgehammer for every problem that a clever engineer could work around.

    some of options ava. are

    http://www.solar-electric.com/sbchco2512vm.html $182.
    http://www.solar-electric.com/blue-sky-energy-solar-charge-controller-sb2512i-hv.html $187
    http://www.solar-electric.com/blue-sky-energy-solar-charge-controller-sb2512ix-hv.html $222

    So IMHO, a MPPT can DIYed for alot less than this.
  • MPaulHolmes
    MPaulHolmes Solar Expert Posts: 35 ✭✭
    Re: diy MPPT data

    You could get a functional thing maybe for less than that, but with my approach, you already passed $182 on the input capacitor. Or one of the 2 contactors. Or the IGBT.
  • drraptor
    drraptor Solar Expert Posts: 218 ✭✭
    Re: diy MPPT data
    You could get a functional thing maybe for less than that, but with my approach, you already passed $182 on the input capacitor. Or one of the 2 contactors. Or the IGBT.

    Contractors are easily ava. second hand for like 2.5K to 6K ~ 25 to 50$. Yes I can go beyond 30A limit.
  • MPaulHolmes
    MPaulHolmes Solar Expert Posts: 35 ✭✭
    Re: diy MPPT data

    I did the first high power test, all panels, with no clouds. I made sure two 12200BTU air conditioners were running off of the inverter before turning the charger on so the charging amps wouldn't be ridiculous.
    This was the transition to constant voltage mode at 1:05 pm or so. The current oscillation is much more calm now while holding at the constant voltage.

    At this point, the current out of the charger had fallen to around 85 or 86 amps. It did over 85 amps continuous with an IGBT hotspot temperature of about 12 degrees C over ambient. Ambient was around 115 F inside the shop. Thermal cutback starts at 75 degC, so I still had quite a bit of room.
    FirstHighPowerTestTransitionToCV.jpg

    Here's what the power in and power out looked like at around 12:35. At that time the current out of the charger had made it to a little over 88 amps. Not very interesting when there are no clouds!:
    FirstHighPowerTest.jpg
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: diy MPPT data

    Where do you buy these monstrous components - IGBTs, Capacitors, Inductors?
  • MPaulHolmes
    MPaulHolmes Solar Expert Posts: 35 ✭✭
    Re: diy MPPT data

    600v 600amp igbt module from Digikey:
    http://www.digikey.com/product-detail/en/CM600DY-12NF/835-1023-ND/1989675 (but I got a new one on Ebay for like $40 I think. It was from someone I trusted, and still had all the paperwork from the company proving they were new)

    600v 500uF capacitor. These are amazing. Developed for the military. Probably 5 times better than the competitor:
    http://www.futureelectronics.com/en/technologies/passives/capacitors/film-capacitors/Pages/5211230-700D50796-349.aspx?IM=0

    300uH 200amp continuous inductor:
    http://www.digikey.com/product-detail/en/CH-200/CH-200-ND/1984836

    The inductor runs pretty cool at 90 amps (the peak so far). Well, the shop is 120 deg, and the inductor is 10-15 degrees warmer than that, so maybe not "cool", but well below danger territory.
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: diy MPPT data
    600v 500uF capacitor. These are amazing. Developed for the military.

    Never seen anything like this one.

    500uF looks a little too small. At 50A and your 10kHz switching, the ripple will be around 10V - may go too far away from the MPPT point at the extremes.

    Or do you do have high voltage strings, so you get like 300V and 20A?
    300uH 200amp continuous inductor.

    $400 for an inductor looks lika a lot to me. Have you thought about winding your own?
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: diy MPPT data
    NorthGuy wrote: »
    $400 for an inductor looks lika a lot to me. Have you thought about winding your own?

    Looks like a lot to me too: http://www.solar-electric.com/xaxwmp60amps.html

    DIY is obviously not the cheap option for charge controllers. :roll:
  • MPaulHolmes
    MPaulHolmes Solar Expert Posts: 35 ✭✭
    Re: diy MPPT data

    They have higher uF versions, but I just went with that, since it's fairly small. I am doing around 200v at 25A. It was a lot for the inductor. I tried winding my own toroid inductor and it worked for lower current. I haven't tried a high current version though. I just had that one laying around from a few years ago because I was using it (along with some heavy nichrome wire for added resistance) to test a motor controller when I didn't have a motor.

    I'm not using the input current and input voltage at all for computing the maximum point point. Just maximizing the output current (since it's connected to a 48v battery pack. Well, actually maximizing the average of the last 512 samples of the output current. The output capacitance is a 4700uF electrolytic rated for 450v. I didn't post the link to that one since it was just a random capacitor I had laying around, and I have no idea what part it was, and it's stinking hot outside! I'm not checking. hahaha.

    I could run this one up to around 350v output though at 200 amps. It's been running at 50degC ambient shop temperature with an output of 90 amps peak (around noon) 12 hours per day. Of course, it's not working quite as hard in the morning and evening. haha. Darn that sun. I'm going to test it out at 350v in the next few weeks, so we'll see how it goes. Of course, it won't be at 200 amps, since my panels won't quite do 200 amps at 350v. Maybe I can sneak on to the Frito Lay solar field that's just south of where I live. They have a huge array. I'd have to also sneak 350v worth of lead acid batteries with me.
  • mtdoc
    mtdoc Solar Expert Posts: 600 ✭✭
    Re: diy MPPT data
    Looks like a lot to me too: http://www.solar-electric.com/xaxwmp60amps.html

    DIY is obviously not the cheap option for charge controllers. :roll:

    Not at that level of voltage and amperage.

    On the other hand Linear Technology and Texas Instrument now both offer small scale MPPT controllers on a chip that can be had for a few bucks each! . For small scale charging someone could probably incorporate a few of these into a fairly low cost DIY box.

    ADD- For a larger inexpensive DIY mppt controller - big enough to actually charge 12V battery there's Tim Nolan's Arduino CC
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: diy MPPT data
    I'm not using the input current and input voltage at all for computing the maximum point point. Just maximizing the output current (since it's connected to a 48v battery pack. Well, actually maximizing the average of the last 512 samples of the output current.

    Whether you use the input measurements or not, they do affect the performance.

    MPPT point is a point. There is some voltage and current associated with it.

    When you turn on your transistor, the capacitor start discharging, so its voltage drops. The current from panels stay the same (may increase a bit). As a result, the total power taken from panels decreases. As seen by panels, you're now below the MPPT point.

    Now, when you turn it off, panels start to charge the capacitor. Voltage goes up. It passes the MPPT point and goes beyond it. The current decreases, and total power decreases too. As seen by panels, you're now above MPPT point.

    Therefore, because of the ripple, you do not sit exactly at the point, but oscillate around it with your switching frequency. The bigger the capacitor, the lower the amplitude of the oscillation. Therefore, with bigger capacitor you do not move very far from MPPT and, as a result, produce more energy. 500 uF capacitor is probably good for your high voltage setup - the voltage should vary only 1-2%, but bigger capacitor would still be better.

    If there was 10% ripple (as you should expect from your controller if you run input at about 70V) or more, the oscillation would be wide around the MPPT point and panels would spend only a fraction of the time in the vicinity of the MPPT point. This would lead to smaller harvest. At the same time, the efficiency measured as a ratio of outgoing to incoming power would be roughly the same regardless of the size of the ripple.

    With 150V drop you should have problems with heat from switching losses in IGBT. How do you deal with that?
  • MPaulHolmes
    MPaulHolmes Solar Expert Posts: 35 ✭✭
    Re: diy MPPT data

    Oh I see what you are saying. That makes a lot of sense.

    I'm using a pretty small gate resistor (3 Ohms) and the VLA500-01 igbt driver turns it on and off hard, so it's not in the linear region very much. The laminated bus with the capacitor with 100uOhm makes for a small turnoff spike even though it's happening so fast. It does get warm, but nothing to write home about.
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: diy MPPT data

    You switch at 10KHz. This is an audible frequency. Could be making terrible noises. Can you hear it? Is it loud?
  • MPaulHolmes
    MPaulHolmes Solar Expert Posts: 35 ✭✭
    Re: diy MPPT data

    I hear it barely, but my inverter is right next to it, making a buzzing noise. Then I usually have a fan blowing if I'm working out there as it's 120 degrees. haha. Then I usually have my geared head milling machine running at the same time. So, overall the 10kHz isn't noticeable to me as EVERYTHING is stinking noisy out there.
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: diy MPPT data
    I hear it barely, but my inverter is right next to it, making a buzzing noise. Then I usually have a fan blowing if I'm working out there as it's 120 degrees. haha. Then I usually have my geared head milling machine running at the same time. So, overall the 10kHz isn't noticeable to me as EVERYTHING is stinking noisy out there.

    Milling machine is a really good way to silence noisy electronics!