Midinite 150 first impressions

12346»

Comments

  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: Midinite 150 first impressions
    ChrisOlson wrote: »
    What firmware do you have in your XW-MPPT60's? It should be 1.05 Build 0006.

    I don't really know. For some misterious reason, SCP doesn't show the software build for SCCs.

    I didn't try to watch the input voltage as BoB did. I only watched watts. So, I I haven't seen the effect that BoB has described. But as soon as I see a weather with clouds passing by (may be even today), I'll watch it, and post the results.
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Midinite 150 first impressions
    NorthGuy wrote: »
    I don't really know. For some misterious reason, SCP doesn't show the software build for SCCs.

    Yes, it does. Go to the System Status screen then press the Enter button. Select System Settings, then scroll down to View Device Info. Pressing the up and down arrows will show the firmware versions in each of your Xanbus components and it should say this:

    Attachment not found.

    --
    Chris
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: Midinite 150 first impressions
    NorthGuy wrote: »
    I'll watch it, and post the results.

    I watched it in the passing clouds, and I couldn't see the effect. Looks like I have newer software.

    My MPPT point is 94.2 (by docs). When it is darker and clouds change fast (400-500W on 3kW array) XW is hunting for MPPT changing in range between 80 and 100V. Two SCCs are not synchronous. Looks like, in fast changing conditions, they have hard time finding the gradient and may wonder away from MPPT point. Apparently, there's a room for improvement here, but since production is low anyway, the impact is rather small, perhaps no worse than 20W. When condition stabilized and the Sun came out almost completely (about 2kW on 3kW array in present conditions), both XWs locked at 89.9V very quickly, in a matter of seconds, and stayed at this number.

    It would be good to know how Classic behaves in such conditions.
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: Midinite 150 first impressions
    ChrisOlson wrote: »
    Yes, it does. Go to the System Status screen then press the Enter button. Select System Settings, then scroll down to View Device Info. Pressing the up and down arrows will show the firmware versions in each of your Xanbus components and it should say this:

    Yes! Thanks. It is the same as version as yours.

    I knew it was under "View Device Info", but I looked under "MPPT 01" menu and it is "View Device Info" too, but it doesn't show the version. That confused me.
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Midinite 150 first impressions
    NorthGuy wrote: »
    It would be good to know how Classic behaves in such conditions.

    I've been watching it at the same time you've been watching it. Our conditions here today are solid overcast off Lake Superior all morning. The panels are working off what I call "Cloud Shine". The sky is bright but no shadows being cast. Both the Classic and MPPT60 have been operating their respective arrays at 38-39% nameplate. My Vmp is 92.4 just like yours. The voltage has been running around 85-88 on both.

    About 15 minutes before solar noon the sun popped out thru a hole in the cloud deck. I was sitting on a stool in the utility room eating an egg salad sandwich when it happened. Fast as you can snap your fingers the MPPT60 dialed the east/west arrays up to full nameplate - 3,000 watts. By the time I could get my eyeballs from that controller to the Classic, the Classic had also dialed the south facing up to full nameplate - 3,750 watts (we put in three more panels when we installed the MPPT60 here). It all happened so fast I could not catch what the input voltages were doing during the 1-2 seconds when both controllers ramped up.

    But watching what the two controllers do with input volts otherwise reveals that the Classic is unloading the array every so many seconds to power sweep it. The MPPT60 is not. It is constantly adjusting the input voltage a tiny bit on each side of Vmp to test for changes. And if it detects a change it ramps it up or down so fast that you couldn't even catch it with a video camera. I got my Sun on the MPPT60's output for awhile and it never "glitches" or unloads the array. Putting the Sun tester on the Classic's output reveals that every so many seconds it totally unloads the array to power sweep it, and the ammeter on the Sun makes a quick snap to zero, then back up.

    I am using the Fast Timed Auto in the Classic and I have the interval set at 15 minutes. I had always assumed that it would not do this unloading of the array for 15 minutes unless it detected a change of more than 10% in the output. But it's not working that way. So I'm confused about what that setting actually does - it doesn't appear to do anything. I think the default is 5 minutes for that setting.

    So from my observation platform up here in Wisconsin I'm saying there is absolutely nothing wrong with the MPPT60's tracking. It is very, very fast, and it is accurate. It rivals the Classic any day, any time.
    --
    Chris
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Midinite 150 first impressions
    boB wrote: »
    I would really like to see that video !
    PS... I think you mean what they call shade- tolerant Fast Sweep. I've seen their papers on that.
    Nothing new I'm pretty sure. I am guessing that they probably won't say exactly how they do it though.

    boB,

    I cannot find that video I watched. I don't remember if it was on their website or on their YouTube Channel. It was really cool because it showed what happens during shading with diagrams and a guy talking in the background explaining it. For a newbie like me it was very educational. All I can find right now is in their documents on it and I'm sure this is the one you are referring to:
    http://www.schneider-electric.com/documents/support/white-papers/seshadetolerantwp.pdf

    No, they don't say how they "do it". I imagine that's a secret. But it is not the same old dynamic tracking that we all know about. Nor is it the P&O type like they used to use in the old controllers. I think it's a form of dynamic tracking because it appears to never really unload or power sweep the solar array. Watching my XW-MPPT60 work, they are doing something with constantly "testing" the Vmp by varying the input voltage a slight amount, and it looks like the software reacts very fast if it detects a small change on one side or the other that it thinks is "better".

    I wish I knew more about this stuff. Whatever they did over the old Xantrex units works. And it works pretty dang good here. In that video I watched I remember it said they have dual trackers in their GT inverters. There was a guy in the video showing it and explaining it, and that inverter had doors on it that opened up to reveal all the stuff in the "sections". It was either 480V or 600V three-phase. But the same technology used in those is now used in their little controllers too.
    boB wrote: »
    So you were operating in a high ambient environment ? That'll certainly make it run hotter. But the Classic should adjust its current limit downward when it gets too hot. If it does that, you should notice the bottom RED LED light up on the MNGP if it does go into
    current limit. I don't suppose you saw that happen ?

    Yep - it did that on July 19 - I noted it in my logs. It cut the current back to 70-75 for most of the afternoon. I was trying to explain it to the other guys here that it does that when it gets hot. I knew about that because I overheated a Classic once with a wind turbine and it got so hot that it went to current limit and burned the clipper up. The clipper was doing fine until the Classic went to current limit and sent more power to the clipper. I took some of the weight off the tail on that turbine to cut its power back a little, plus make a new clipper with elements cooled by hydraulic oil. It's been fine ever since.

    Despite trying to wreck one of these controllers, I've never managed to break one yet. But I am concerned that operating it at continuous temperatures near its limit is bound to shorten its life. That's why I installed dual controllers recently - I thought it would be better to take some of the stress off it.
    --
    Chris
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: Midinite 150 first impressions
    ChrisOlson wrote: »

    Very interesting article. Thanks for posting.
    ChrisOlson wrote: »
    No, they don't say how they "do it". I imagine that's a secret. But it is not the same old dynamic tracking that we all know about. Nor is it the P&O type like they used to use in the old controllers. I think it's a form of dynamic tracking because it appears to never really unload or power sweep the solar array. Watching my XW-MPPT60 work, they are doing something with constantly "testing" the Vmp by varying the input voltage a slight amount, and it looks like the software reacts very fast if it detects a small change on one side or the other that it thinks is "better".

    I don't see why the whole scan should take long time. You just let the input capacitor to charge up, then discharge it to to battery voltage or so. Should take 10ms (well, 50ms if you have real big input capacitor). If you have 100kHz A/D, it'll give you 1000 samples of voltage and current along the line - the whole curve. Just find the maximum and move your operations there. If you do that every 10 seconds, it's only 1% downtime, during which you loose, may be, 20% of what you would otherwise get - 0.2% efficency hit. If you do that every minute, it's less than 0.1%.

    And you don't even need to do it very often. The MPPT point depends on insoltation and temperature. If your MPPT point voltage is high, but production is low, you know right away that one of the strings is shaded. Then you do the scan more often. If not, you do not scan at all or do it infrequently.

    The task of finding MPPT point is easy and doesn't really require any complicated algorithms.
  • boB
    boB Solar Expert Posts: 1,030 ✭✭✭✭
    Re: Midinite 150 first impressions

    I'm afraid I don't know the exact sweet spot. Sorry. But if you (Chris) found that you are more comfortable using 2 and you are definitely
    getting more output power, then I can't argue with that.

    What I can argue about is that conditions change often and one measurement will not necessarily be repeated the next time around.

    On efficiency measurement, you must have separate input and output measurement gear (simultaneous readings and not moving around)
    and they must be very accurate if you trying to measure 1 to 2 or 3 % accuracy.

    You can most likely easily measure a difference of 80% and 85% but you'll play hell trying to measure the difference between
    say, 97% and 98% efficiency. You'll just have to take my word on that one.

    What I did at OB on the MX (originally) was to use calibrated shunts on input and output and calibrated Fluke meters
    ( 4 meters total) and took a digital picture of all of them because as soon as you write down one reading, the others
    will change. I'm not sure if that was even very accurate. The best way to go is to use two of those Yokagawas
    or equivelent with computer control/graphing and then slowly change the input voltage and output current and
    have the efficiency graphed that way. A sanity check on the metering also means that you would measure the
    same input and output with both metering systems and make sure they read the same. Of course, you would
    want to make that experiment on both input and output to make sure the readings and matching doesn't change
    much with input voltage and current.

    Also guys, you cannot measure efficiency by feeling how hot it makes your hands. For example, if you have
    a penny dissipating 5 watts in the air, you will burn your finger. If you have a stove top skillet dissipating
    5 watts, you will barely feel it get warm. This has to do with heat sink area which the XW has more of.

    The XW MPPT could have been designed more efficiently but they chose to use a larger heat sink rather
    than smaller heat sink and fans.

    On the XW 60 MPPT older tracking, I don't doubt that it would not find the right hump in multiple MPPT V conditions but
    it just spent a long time finding where it was. There isn't a way to measure MPP V and be off the MPP V point at
    least some times. The compromise is how long you are spending off of it vs. on the MPP V. The shorter the better.

    The best way to evaluate performance is to find which unit gives you the most kW-Hours at the end of the day.

    I would say Chris, to do what you did and use two units running cooler than one unit running hotter. Especially
    if you get more kW-Hours out of both of those rather than the one all by itself. They'll last longer too. And a
    spare controller is always good to have.

    We are going to try and update our XW MPPT 60 controllers and see what they have done better.

    Thanks,
    boB
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Midinite 150 first impressions
    boB wrote: »
    I would say Chris, to do what you did and use two units running cooler than one unit running hotter. Especially
    if you get more kW-Hours out of both of those rather than the one all by itself. They'll last longer too. And a
    spare controller is always good to have.

    We are going to try and update our XW MPPT 60 controllers and see what they have done better.

    I'm not getting a lot more harvest. But I am getting more with two controllers. I logged from both this afternoon when the sun came out, and the peak power is really not all that different from what I was getting in my logged data with just one Classic. And I don't have a way at present to measure insolation here (looking at getting a meter that does that). But I am getting slightly more kWh from the dual controllers and I can only attribute that to the fact that I'm not stressing out one controller and losing some my solar energy to heat because of pushing the controller too hard. And it is immediately obvious that the Classic is way happier - the internal fan is running this afternoon, but the Turbo fan only came on a couple times for quite sort duration - vs running constantly before.

    To my way of thinking, that just has to be way better for the longevity of the controller.

    You will be able to update your test XW MPPT60 controllers with the new Combox quite easily when it becomes available. The Build 0006 firmware is available on Schneider's website for download as a zip file (no need to call tech support to get it anymore). And no need for the old Firmware Updater Tool anymore.
    --
    Chris
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,728 ✭✭✭✭✭✭
    Re: Midinite 150 first impressions
    boB wrote:

    We are going to try and update our XW MPPT 60 controllers and see what they have done better.

    Thanks,
    boB
    Unless your 60-150V is really old I doubt you will see any difference. I think there is some not quite right info in this thread about the 60 AMP controller. Most of the tracking improvements, the white paper from Dr. Swingler, and the technical advance in the 600V controller do not pertain to the 60 AMP controller. I am not saying that it is not good example of current charge controllers but it is not any different than it was 5 years ago.
    The new Grid-tie inverters and the 80 amp 600V charge controllers do share this algorithm. I noticed with my matched 2KW tracked arrays similar data that Chris noted on with array facing different directions. The place I disagree with him is when an array facing SE and one facing SW are in moving clouds. The power harvest over the day with a controller on each array can be + 40% in clouds compared to a single controller trying to pick the mppt.
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Midinite 150 first impressions
    I am not saying that it is not good example of current charge controllers but it is not any different than it was 5 years ago.

    If the MPPT60 is older than 2010 it probably has the old firmware in it. And Schneider told me there was some MPPT60's that got sold after 2010 that sat on retailer shelves after the transition that still had the old firmware in them, and some of those never got updated. So it's worth checking to see what you got if you bought the controller anything but recently (at least that's what I was told).
    The power harvest over the day with a controller on each array can be + 40% in clouds compared to a single controller trying to pick the mppt.

    While it can, I don't think it's necessarily true that it will be by a significant amount that makes a difference in the big picture. I spent quite a bit of time when I put these additional east/west arrays in this spring testing it with my Classic by operating different portions of the arrays individually and comparing outputs with other sections of the arrays and what the controller said was ideal Vmp for each one.

    I determined that as long as you do not wire panels in series that are facing different directions, it does not make any difference. Read that paper from Schneider and it explains why. You're not dealing with shading with the arrays pointed in different directions - just how much insolation each array is getting because of its angle of incidence with the sun. And therefore a slight difference in Vmp between the different arrays. But, surprisingly, I found that the Vmp was not different enough to be a problem. A good controller should find the total global max output point of the whole thing and not track just one part of it. And that's the key as to why it works. I spent a lot of time figuring out how much capacity to face in each direction so it would match the Classic's maximum output capacity for at least 5 hours per day. And if you do it all right - you can harvest some SERIOUS power in a matter of only 5 hours with one controller, and have a big "boost" on each side of that 5 hour period without a single moving part in the system (except the screaming fans in the controller). I done it. And it works.

    My experiments on the cloudy (totally overcast) day proved to me that the conditions are even better - basically every array is receiving almost identical insolation and it "just works", albeit at lower output from the perfect sunny day.

    When you have clouds intermittently shading one section of the array, but the others are in bright sunshine - THEN two or more controllers will beat the single one hands down. But I don't know how much.
    --
    Chris
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Midinite 150 first impressions

    Good thing I'm not grading this "science class". :p:D
  • boB
    boB Solar Expert Posts: 1,030 ✭✭✭✭
    Re: Midinite 150 first impressions
    ChrisOlson wrote: »
    I'm not getting a lot more harvest. But I am getting more with two controllers. I logged from both this afternoon when the sun came out, and the peak power is really not all that different from what I was getting in my logged data with just one Classic. And I don't have a way at present to measure insolation here (looking at getting a meter that does that). But I am getting slightly more kWh from the dual controllers and I can only attribute that to the fact that I'm not stressing out one controller and losing some my solar energy to heat because of pushing the controller too hard. And it is immediately obvious that the Classic is way happier - the internal fan is running this afternoon, but the Turbo fan only came on a couple times for quite sort duration - vs running constantly before.

    To my way of thinking, that just has to be way better for the longevity of the controller.

    You will be able to update your test XW MPPT60 controllers with the new Combox quite easily when it becomes available. The Build 0006 firmware is available on Schneider's website for download as a zip file (no need to call tech support to get it anymore). And no need for the old Firmware Updater Tool anymore.
    --
    Chris

    We have been working with Schneider locally because we are building an E-Panel for their 24V Connext inverter. I thought you were using a Connext somewhere ?

    So, we may be able to have them help us to update this controller here. Can't remember now if I have that controller or if I gave it to Ryan ?
    Maybe we have two of them now ?

    I would expect your MPP voltage to be about the same with the array divided up. The internal fans being on and the turbo not coming on very often
    is just about right I'd say. Temperature wise.

    boB
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: Midinite 150 first impressions
    ChrisOlson wrote: »
    When you have clouds intermittently shading one section of the array, but the others are in bright sunshine - THEN two or more controllers will beat the single one hands down.

    You need to have these sections at leat a mile apart for this to happen.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,728 ✭✭✭✭✭✭
    Re: Midinite 150 first impressions
    Good thing I'm not grading this "science class". :p:D
    Definitely agree this time! It is very, very hard to keep logic in this with the sunlight changes. Even 2 perfectly tracked arrays can delta your data. And then you throw in being offgrid and the man behind the curtain really gets nervous.
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,728 ✭✭✭✭✭✭
    Re: Midinite 150 first impressions
    ChrisOlson wrote: »
    Yes, it does. Go to the System Status screen then press the Enter button. Select System Settings, then scroll down to View Device Info. Pressing the up and down arrows will show the firmware versions in each of your Xanbus components and it should say this:

    Attachment not found.

    --
    Chris
    I just looked and this build added an update to scroll mode, boost charging, and added float time and the time of day the cc went to float. Too bad the old Xantrex forum went away. A guy here on the forum Solar guppy did the beta development for grid-tie and we were talking once when I was doing the offgrid firmware beta. (which was not much change) He told me very little changed in the unit after it was sold. Some of the decisions that would have made it as good as the Morningstar were not adopted for cost reasons (at this early time). A lot of this is probably here in the wind/sun oldies.
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Midinite 150 first impressions
    boB wrote: »
    We have been working with Schneider locally because we are building an E-Panel for their 24V Connext inverter. I thought you were using a Connext somewhere ?

    I dunno. I guess they're changing the name of everything to Conext now for anything to do with renewable energy. But all the equipment I got still says Xantrex on it - right on the front. The invoice for our inverter says it's a Conext. On the front of it says Xantrex and don't say anything anywhere on it about being a Conext.

    So you tell me what the deal is with Conext :confused:
    --
    Chris
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Midinite 150 first impressions
    NorthGuy wrote: »
    You need to have these sections at leat a mile apart for this to happen.

    LOL! That's true. I drug out my notes from when I tested all this and here's an example of what I got on April 6

    I watched until the sun was shining at exactly 90 degrees to the east facing panels near sundown. I determined that standing on the ground sighting along the panel until the sun appeared to be just on the edge of them. 2.25 kW installed capacity on the east, .75 kW on the west - turn off the west facing combiner breaker and let the controller track it.

    I had 83.9 volts and 111 watts.

    Turn off the east and turn on the west and let the controller track it.

    I had 84.7 and 340 watts.

    Turn it all on and let the controller track it and I had 84.5 and 440 watts.

    So if you add up the two individual you get 111 + 340 = 451 watts is what it should've been, or could've been with two controllers.

    Done with a Classic 150..

    The times when you can get shading on one set of panels from a cloud, but not on the other, would only happen once in a blue moon because they're only separated by about 6 feet of roof over the peak.
    --
    Chris
  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    Re: Midinite 150 first impressions

    just a thought on this chris. if it does not fold back amps very often i question the viability of running 2 separate classics as you are now doubling the losses by running 2 ccs. these losses will be more appreciative than the occasional current fold back. only other factor may be the equipment lifespan due to the higher heat. ultimately your call though.
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Midinite 150 first impressions
    niel wrote: »
    just a thought on this chris. if it does not fold back amps very often i question the viability of running 2 separate classics as you are now doubling the losses by running 2 ccs. these losses will be more appreciative than the occasional current fold back. only other factor may be the equipment lifespan due to the higher heat. ultimately your call though.

    That's true. I haven't a chance to see how it does on a really bad day yet. But, frankly, I think these controllers are pretty darned efficient at very light loads like you'd get on a bad day (where arrays are running at ~15% nameplate). So, truthfully, I think the losses with two controllers are going to be really, really small on the bad day. And I'm seeing an improvement with two on the good day.

    I'm definitely going to stick with dual controllers just because they don't get hot. I don't like it when my equipment gets that hot.
    --
    Chris
  • fca1
    fca1 Solar Expert Posts: 124 ✭✭
    Re: Midinite 150 first impressions

    Hi there been away from one week and here are some readings for compare..
    Last24 battery voltage comparison (morningstar mppt60-RED, midinite classic-BLUE, victron shunt-GREEN)
    the more accurate is the victron shunt, but the midnite is very distant from mppt60 (im not using morningstar batt sensor)
    Attachment not found.

    AmpsIn vs Fet temperature last 24 also

    Attachment not found.
  • mtdoc
    mtdoc Solar Expert Posts: 600 ✭✭
    Re: Midinite 150 first impressions

    Nice!good to see real data!
    fca1 wrote: »
    Hi there been away from one week and here are some readings for compare..
    Last24 battery voltage comparison (morningstar mppt60-RED, midinite classic-BLUE, victron shunt-GREEN)
    the more accurate is the victron shunt, but the midnite is very distant from mppt60 (im not using morningstar batt sensor)
    Attachment not found.

    Have you tried adjusting the Midnite's battery voltage offset to better match the Victron and Morningstars?

    FWIW my Classics match my Outback FNDC monitor and my Fluke 87V DMM well. My outback inverter voltage is off by 0.2 volts but I use the offset adjustement to get it to match the other 2.
    AmpsIn vs Fet temperature last 24 also

    Attachment not found.

    Cool - literally! I assume this is from your Classic? Your Fet temp is running about where mine runs - around 50 C. What is your ambient temp?
  • fca1
    fca1 Solar Expert Posts: 124 ✭✭
    Re: Midinite 150 first impressions

    Hi, today i had a good day from the midnite as yesterday i got the batts very low ... so midnite worked at the value i'm expecting ~70-80 amps .. and the temperatures vs ampsout vs pv amps...

    Attachment not found.
  • fca1
    fca1 Solar Expert Posts: 124 ✭✭
    Re: Midinite 150 first impressions

    Hi , today my midnite reached 70.2C very hot .... also was the day that it produced more power 14.2KW ,
    don't know if that drop in amps (near 14:00) was the midnite cutting because of temperature , what is temperature cut of the midinite, i think i saw a graph with that, but dont remember where..
    thanls


    Attachment not found.
  • ChrisOlson
    ChrisOlson Banned Posts: 1,807 ✭✭
    Re: Midinite 150 first impressions
    fca1 wrote: »
    Hi , today my midnite reached 70.2C very hot .... also was the day that it produced more power 14.2KW ,
    don't know if that drop in amps (near 14:00) was the midnite cutting because of temperature , what is temperature cut of the midinite, i think i saw a graph with that, but dont remember where..

    I was pushing 20-22 kWh thru my Classic in five hours last July and I think it starts de-rating itself at around 80-85C. If you're only at 70C, that's cool running yet. Get it up to around 4,800 watts output for 2-3 hours and then you'll see what nice and hot is.
    --
    Chris