 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Electronics : Nanoverter and Backfeeding with a GTI
  Page 3 of 7   |
|||||
| Author | Message | ||||
| rogerdw Guru  Joined: 22/10/2019 Location: AustraliaPosts: 815 |
Thanks Mad, I had a feeling you were using HF GTI's. I know when I've looked into it before, I got the impression that most believed that only LF ones would work. Clearly your experience proves otherwise. Sorry for the same question here too mab1 ... are yours HF or LF versions. Just trying to understand the concept for future experimenting myself. Thanks Mad and mab. Cheers, Roger |
||||
Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
Any GTI should work. The only thing I have heard about like this was that there was talk that the GTI need galvanic isolation such as the Aerosharps with there toroid. I used an aerosharp for a short time. I know others have used them a lot, most of my experience with GTIs off-grid has been with Zeversolar and ABB Aorura. The type of off-grid inverter that will work with a GTI though is a different matter. Maybe there are HF ones that will do it but I have never heard of it. Edited 2023-04-27 16:42 by Madness There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| nickskethisniks Guru Joined: 17/10/2017 Location: BelgiumPosts: 428 |
Indeed! That's why I find this problem so interesting. I'm still using the original ozz inverter, and is tested and working with Diehl ako platinum lf, GE hf, aten hf and other chinese hf gti's. The Diehl had to see ground and the right polarity of the AC before it wanted to start. |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
The HF types I have used measure resistance to ground on the DC side and will not work with less than 1 megaohm. I have never tried connecting line and neutral around the wrong way, I just connect them as they are marked. Neutral must be bonded to ground in one and only one place in your house wiring. There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| mab1 Senior Member  Joined: 10/02/2015 Location: United KingdomPosts: 173 |
TBH i don't think I've ever really thought about how backfeeding works through the dc-ac modulation stage of an inverter- i was under the impression that all lf type sine inverters would work and the hf type double conversion wouldn't cos the voltage boost stage was unidirectional. Roger: my sunny boys are: SB1200 LF with 1300W pv SB1700 LF with 2000w pv TL4000 HF with 2 x 480w pv The tl4000 does work but at low power output uses burst firing which makes the a.c. voltage fluctuate a bit. Also, when using the ASHP, if it's drawing 2300w or >3000w the tl4000 seems to interact with it and the a.c. voltage goes a bit into oscillation, so i tend to turn the tl4000 off for the duration. My guess is that it's just an unlucky combination of response times of the inverter/gti/ASHPs inverter drive, but possibly the big trannys in the smaller inverters provide a stabilising effect? |
||||
| rogerdw Guru Joined: 22/10/2019 Location: AustraliaPosts: 815 |
Thanks for the additional info Mad and mab1. It's a confusing concept, but just good to know it can be made to work with the right combination of equipment. mab1 you mentioned clockmanfr ... and if I remember correctly he uses older LF SMA's as well. Cheers, Roger |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
Doesn't need to be confusing, if you have a suitable off-grid inverter just connect AC power to the GTI. The power from the GTI is added to your OGI's output, if there is excess power the OGI feeds back into the battery. However it needs to be controlled to prevent overcharging, I have the regulator shown above that takes care of that by applying PWM to the DC input from the solar panels to reduce the power as required. The regulator can also turn your hot water heater and other loads to use excess power. Once DC battery current reaches a limit I have set the hot water is turned and as soon as current drops from a cloud it turns off. There are several advantages, it is a very efficient way to convert sunlight directly to AC power. You need just 4mm square cable for the panels as you normally have 10 or so in one string operating at around 400VDC. GTIs in Australia and other places that use lots of them can be had for little to no money. All the the Aorura GTI's I have were either free or under $100, they have a fault caused by a relay that is easily repaired. They should then work for another 6-7 years before needing the relay replaced again. Solar panels can be bought for little money also from people upgrading there solar, I recently got 12 250W panels and a Latroncs LF GTI with all the mounting rails etc, wiring and isolators for free. Just keep an eye out on market place. It was getting these free extra panels that prompted me to upgrade my off-grid inverter. The thing about LF and HF GTIs is a furphy, it doesn't matter which, if it can connect to the grid it will connect to a suitable off-grid inverter. Edited 2023-04-28 07:43 by Madness There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| rogerdw Guru Joined: 22/10/2019 Location: AustraliaPosts: 815 |
Thanks, very helpful. Makes it sound so much simpler ... and also easier to swallow knowing it's not just theory. I went back and had a closer look at your controllers and can see the concept now. I certainly have some loads I could switch in and out when I get to that stage too ... and knowing you can also throttle back the dc going into them adds even more control. I regularly see stuff on marketplace, though rarely free ... but still happy to pay for good gear, and have a friendly solar repairer who'll sell as many used inverters as we like for $25 each. In another thread you show your nano daughter board on the back of your display ... what size display or part number is that? It looks bigger than the usual ones. I'm trying to work out which ones I'm going to use for my MPPT builds. Thanks. Cheers, Roger |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
The LCD I use for everything and Poida has also is 20 X 4 line. If you can buy GTI's for $25 each then ask the solar guy to keep a couple of ones that he thinks are good. No doubt he would be able to help with panels at the right price also. Edited 2023-04-28 10:45 by Madness There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
I have hit a brick wall and my head hurts! Yesterday I removed one choke from this new big inverter thinking that perhaps I have too much choke. After that, it ran fine with the 2XSINE code (as was when I had 2 chokes) so I connected the inverter to the house very late in the afternoon and ran it overnight without an issue. But as the sun came up and the 3 GTIs started producing power glitches started appearing in the sine-wave and they would disconnect and reconnect and as their output increased it become worse. So I switched the off-grid inverter off and thought about what to do next. Next step I put one of my EG8010 controller boards in the inverter and tried that. The bloody thing was playing up just the same as with the Nanverter controller! Should have tried that earlier, anyway I now now there is something not quite right in either the power board, choke or the toroid. Without connecting a GTI it runs fine so I am scratching my head ATM. The 2 controller boards, Nanoverter and EG8010 ran fine on the small test inverter and I get the same issue with either in the big inverter so I am fairly confident there is not a problem with them. Fortunately, I have plenty of everything to swap bits around to isolate what is wrong. I know the small test inverter ran really well with the Nanoverter 2XSINE so I also am fairly confident that it will be fine in the big inverter once I get it sorted out. It has me mystified as to why it runs so well until the GTIs start working a bit. There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
InPhase Senior Member Joined: 15/12/2020 Location: United StatesPosts: 178 |
Are these GTIs known to work otherwise? Could it be that one or more are doing something wonky and causing the problem? |
||||
| pd-- Senior Member Joined: 11/12/2020 Location: AustraliaPosts: 122 |
Is there sumthing funky with the diods in the fets on the power boards. Im running offgrid with AC coupling & a genset back feeding but with very different hardware. The only thing i see that mucks things up is jitter in frequency or voltage. 1 we know the 2xsine code is good because it works on the baby inverter. 2 if the gti's wont go online then there is a problem with either PF or frequency/voltage stability 3 if the gti's go online then unusual things happen then its time to swap out one bit at a time. |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
I have been using GTI's in this way for around 7 years now with the EG8010 controller. The GTIs have been working flawlessly with the EG8010 off-grid inverter. Most days I am generating in excess of 25KWH and at least 10KWHs is being used to charge my battery via the off-grid inverter. What this thread is about is getting the same thing to work with the Nanoverter controller. That I have been able to do with a small test inverter, what has been the issue is getting this to work in my newly built large inverter. I have established that there is a fault in the large inverter. Edited 2023-04-28 16:11 by Madness There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
There is something wrong, I have taken a break from it today to do other things I need to get done. It could be the power board, could be the choke(s), could be the new 3-core torroid, I will get it worked out over the weekend. Just need to swap components till I isolate where the fault is, if I did not use GTI's I would not know it is faulty. Although something is wrong and one day it would have caused something to go wrong. While it was running overnight the heatsink was running warmer than I would expect but I thought that may have been due to the 2XSine and additional switching losses that Poida mentioned. With 300 - 500 watts overnight the heatsink was 32 degrees and the torroid 29 with an ambient temperature of less than 20 overnight. This makes me think that something on the powerboard is probably the culprit. Yes, the 2XSine worked so well on the small test inverter, this is a big positive for me, otherwise, I might have given up on it. In fact, when I put the EG8010 controller in the big inverter to get it going with it I may have left it there, but it was a step I should have taken earlier to know that the new inverter was good. The GTIs would connect but disconnect when their power output rose above a few hundred watts or if something like our house water pump started. Now the house is being powered by the EG8010 which has been powering the house for the last 7 years. The GTI's connect to it happily and run all day without a single disconnect or complaint of any kind. I have another smaller EG8010 inverter with a single-core torroid that has been wired in as a backup. It works happily with the GTIs as well, when I get this new inverter working the one that has been powering the house for so long will become my backup if something goes wrong. I have a manual transfer switch wired in so I (or anyone else in the family) can change between inverters in a few seconds. There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
I have corrected the sketch here.  I believe this is produced by this Bipolar modulation. From this video  There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| analog8484 Regular Member  Joined: 11/11/2021 Location: United StatesPosts: 92 |
I am just catching up on this interesting thread. Is this the OGI system that has been AC coupled to 3x5kW GTI's for 7 years? If so, can you share the general specs of the EG8010 OGI (# fets or kW) including the battery system (type and kWH) ? Also, what is actual max backfeed/excess power from the GTI's that you have seen? I am curious because the general rule for reliable AC coupling is 1:1 ratio between the OGI and GTI rated power capacity and the continuous backfeed charging current must not exceed 0.5C of the battery system. |
||||
| analog8484 Regular Member Joined: 11/11/2021 Location: United StatesPosts: 92 |
It's actually unipolar modulation. Different PWM modulation can certainly affect the inverter behavior but it can't stop backeed charging from AC coupled GTI's. GTI's are current sources that will push the island grid voltage as high as it needs to pump as much power as possible (limited by PV/sunlight). So, the backfeed current will charge the batteries through the body diodes in the inverter MOSFET's when the voltage gets high enough even if the MOSFET's are turned off. I took a quick look at the nanoverter code (2xsine) and it appears to have a pretty slow control loop with pretty high PID gain. It appears to assume fairly stable and linear loads which is not really valid for AC coupled systems. In AC coupled systems, island grid voltage can jump 10's volts mid cycle (e.g. large load turn off's) and some GTI's can continuously inject periodic and non-linear power disturbances due to anti-islanding mechanisms. Your GTI problems could be due to transient voltage instability caused by the interaction between the voltage rise driven by the GTI's and the inverter control loop response. Since you are off grid, if your GTI's are IEEE1547/UL1741 compliant and has configurable grid settings then I suggest you disable anti-islanding and set the range for voltage and frequency to as wide as possible, and set the trip time for out of range voltage and frequency to the max allowed (default is typically 160ms). You may also need to tune the PID gain parameters. Edited 2023-04-29 03:00 by analog8484 |
||||
Clockmanfr Guru Joined: 23/10/2015 Location: FrancePosts: 427 |
Mab1, "Hmm. Well the three gtis I'm using with my powerjack are all SMA sunny boys - neither cheap nor Chinese i don't think, and they have to meet the uk grid spec. They can be programmed for off grid ( i think clockmanfr does that with his), to make them less sensitive to the kind of glitches you get when running off an inverter instead of the grid, but mine are still set for uk grid tolerances, and only disconnect if I'm using the MIG welder on max power or starting the lathe. I still don't understand why the modulation method would make a difference TBH, as long as the 230v out is approximately sinusoidal, so I'll be very interested to see how this pans out. It's good to know the eg8010 based inverters back feed ok though, as that's what I've got as a backup if the PJ ever fails." ................................................ I used to reset them so i could sequentially shut down when the AC voltage gets to high at 252vac, i.e., with SMA old toroid types. Now i don’t bother anymore and they stay on the old settings that i get them at. Note, I have found that any GTI that is over 3000w, does tend to trip in and out as the GTI surges the AC lines, and if the batteries are insufficient to clamp down any back feeding to the batteries, ie not enough storage i find the GTI over 3000w just trips out. I understand, although i haven’t tried any yet, that HF GTI's will be okay, again keeping below 3000w. All my GTI's are TT earthed with a 1.5m copper pole in the ground, as each of my GTI's tend to be in our 9 separate buildings connected into our mini Grid with about 500 meters of AC cabling and volatge Sag can be usefull. Earthing is difficult and controversial, and i spent some time discussing this with 'Warpspeed'. Only the neutral (I designate the neutral after my EMI AC output filters) at the Main Power inverter is connected/bonded to a separate TT Earth rod. The neutral cable is insulated all the way to the TT Earth rod including the actual top of the rod that is above ground. For other TT Earth at that particular installation building i have another separate TT Earth that is 3m away from the Power Inverter neutral/bonded TT Earth rod. I trust the above is helpful Edited 2023-04-29 04:34 by Clockmanfr Everything is possible, just give me time. 3 HughP's 3.7m Wind T's (14 years). 5kW PV on 3 Trackers, (10 yrs). 21kW PV AC coupled SH GTI's. OzInverter created Grid. 1300ah 48v. |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
The video in the link also shows unipolar modulation, but the screenshot I posted is what they call Bipolar. I am just going off what they say, if anyone has more information on the subject I would be interested to see it. Edited 2023-04-29 06:42 by Madness There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498 |
The OGI is my modified version of the Ozinverter which includes the 24 fet power board that I designed and is mentioned many times on this site. It has a toroid that I rewound from 2 3KW Aerosharp GTI toroids, battery is a 48V 37.5KWH FLA forklift battery. I have seen 120A current to the battery from the OGI, it is now regulated to 80A maximum. The back feeding ratio has never been an issue, I can be using 500W while 80 plus amps are going to the battery and it is 100% stable. There have never been any disconnections etc except with this new inverter which has now been found to be faulty. This is the OGI I have been using for the last 7 years. The photo was taken Wednesday, 23 November 2016 so not quite 7 years.  This is the new OGI that I am working on. Same power board, 3 core torroid and Nanoverter control card.  Edited 2023-04-29 07:05 by Madness There are only 10 types of people in the world: those who understand binary, and those who don't. |
||||
| Page 3 of 7 |
|||||
 Print this page |
