 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 : Parallel operation of inverters
Page 1 of 2   |
|||||
| Author | Message | ||||
| Godoh Guru  Joined: 26/09/2020 Location: AustraliaPosts: 378 |
I have a few inverters in my shed, and am wondering if anyone has run theirs in parallel. I have two Latronics inverters one is a 3kw one is a 2.3kw inverter I also have two 8010 based inverters. I remember many years ago seeing generators in a small power station being synchronised by using three lights. So I am wondering if there is a safe way to run inverters in parallel. I have plenty of solar power, but only run 24 volt batteries. I am not interested in changing my gear for 48 volt as our needs are very modest. So on sunny days I would like to use more of our panels output on larger loads if possible Thanks Pete |
||||
wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 995 |
Hi Pete, because the inverters all have asynchronous clocks running they will never stay in Phase even if you somehow manage to start them all in phase. I believe the only way to parallel them is to slave one or more output stages to a master ie have the 1 control section running two+ output stages in parallel. But I think they should ideally be identical output stages driving a common transformer and choke or 2+ identical toroids/chokes with their outputs in parallel as the output voltages need to match exactly too. Not good news I feel, but I will watch with interest to see if any other ideas surface - I'm always willing to learn too. Matching rotary machines in power applications is also a bit of a black art but a bit more forgiving I think. If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| Godoh Guru Joined: 26/09/2020 Location: AustraliaPosts: 378 |
Thanks Mike, We have bought an EV and at the moment it is charging fine on our 2300 watt Latronics inverter. I will be adding more panels to our system in the coming months as well. I figured that it would be hard, so will just keep using a dedicated inverter for the car, one for the house and another for the shed. We don't drive much so will only need to charge the car to top up after we go shopping. We just wanted to continue on the stand alone lifestyle. Thanks Pete |
||||
phil99 Guru Joined: 11/02/2018 Location: AustraliaPosts: 1783 |
Some inverter generators can be paralleled with a synchronizing cable that makes one inverter a slave to the other, achieving what Wiseguy said. My experience with paralleled generators is that in manual mode they need constant attention to maintain sharing of load and power factor. Power sharing requires tweaking the governors and power factor sharing requires tweaking the voltage regulators. Rated output can't be achieved unless they all have the same power factor. This would apply to inverters also, adjusting the voltage feedback to reduce the output of the one with a leading PF. Smaller generators and inverters may be more forgiving but for automatic operation these larger ones (2 x 1350kVA and a 1600kVA) use a Summation Unit to measure and control everything. |
||||
Murphy's friend Guru Joined: 04/10/2019 Location: AustraliaPosts: 583 |
Pete, I have been tinkering with that idea few years ago, ended up in tears  .You can parallel GTI inverters, these are designed for that job. But Our home built inverters, nah too complicated, if at all possible. It's quite a different task with rotating generating machinery. Once they are synched their rotation speed and phase stays locked together electrically so both can supply power to the load. You could wire them up with one of those TOMZN contactors which lets you connect one inverter OR the other to your power requirements. This gives you a ready spare. I use that set up to be able to grid connect my house if I need to switch off the inverter, happens very rarely now that I taught my inverter how to behave  . |
||||
| Godoh Guru Joined: 26/09/2020 Location: AustraliaPosts: 378 |
Thanks Phil and Murphy, I definitely don't want to blow my inverters up. So I have just split my loads over different inverters. I have one running the house ( really a granny flat) one running the shed, and another now to charge the car. So I will stick with simple, I can understand simple, and don't want life to get any more complicated than it is. Thanks for your replies and help Pete |
||||
| Godoh Guru Joined: 26/09/2020 Location: AustraliaPosts: 378 |
Hi Mike, I have been thinking about what you said above. I do have two 3kw 8010 aliexpress boards in the shed. From what you said I am thinking that as they both have double row headers where the 8010 boards plug in . That I could just remove one of the 8010 boards, jumper from that control board over to the 8010 socket on the other board. Maybe cut some tracks so the front end before the 8010 board is inactive and try the two boards with a big transformer. The boards are spares that I have had for a while so if I kill them then it is not too big a problem. The aim would be to build a 24 volt inverter that could run 3600 watts continuously during sunny days to charge our car up. Does this sound feasible to you? Pete |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 995 |
Hi Pete, that is essentially what I am saying and I think it should work fine. Personally I am not a fan of the ribbon cables but I consider a 24V rail supply is a bit more forgiving than 48V. I am also hoping that the Power boards you have are the ones with a totem pole buffer drivers ? If it goes bang we can talk again about a controller card (8010) with the inverse opto drive and power board like KeepIS is using. If you buy any more FETs I am very happy with the HY5608 they are almost double the rating of HY4008 and less than half the on resistance. Both desired attributes for an inverter powered from a 24V supply. Edited 2024-01-23 20:19 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
poida Guru Joined: 02/02/2017 Location: AustraliaPosts: 1388 |
I think it possible and not too hard either to have a setup that has a master oscillator, making a 50Hz squarewave. This is then used by any number of picoverters to produce their PWM outputs. I already have played with this concept and it will work but there will be imperfections of the order of one PWM pulse (i.e. 1 part in 400 of a 50Hz AC waveform) If all slaves have the same error then it will be a case of them all making a not quite perfect AC output, with the imperfections all the same size and phase. The picoverter could be used to drive any of the 8010 based Aliexpress boards. I have done this in the past. It still brings the problem of large negative voltages coming out of the FET Gates and killing the gate drive ICs. (hence my preference for totem pole drives on the power board) Easy. one small board to make 50Hz modified picoverter code to look for and use the rising edge of the 50Hz any number of picoverter boards to drive any number of power boards. Voltage feedback will need to be carefully looked at. I suspect I also gang the PWM outputs together since the AC output will be paralleled anyway. Maybe a single Vfb signal sent to all picoveter boards. that's just my initial sketch anyway. wronger than a phone book full of wrong phone numbers |
||||
| Godoh Guru Joined: 26/09/2020 Location: AustraliaPosts: 378 |
Poida, what you say sounds like a fair bit of design, blow up, stuff. It is a out of my league, I am just a retired electrician who used to fix some electronic stuff. Designing it was way above me. But if something used to work, I could mostly get it to go again. I will see how it goes with the two Aliexpress 8010 boards I have, if they blow up then I know that the idea is a dud. Thanks for your thoughts, Pete |
||||
| phil99 Guru Joined: 11/02/2018 Location: AustraliaPosts: 1783 |
That is ideal if the transformer ratios are all the same. Output current sharing will be more complex if they are not. At low to medium load current sharing is unimportant, but at full load to prevent one transformer being overloaded they must each reach 100% of their individual ratings at the same time. The least complicated solution may be to test them by temporarily connecting all the primaries to the paralleled MOSFET output stages and measure the separate secondary voltages with near 100% load on each. Turns can be added to the low one(s) to bring them all to exactly the same voltage. If getting them all to 100% isn't practical at least get each to the same percentage of rated current. With that done the primaries can be separated and the secondaries paralleled. If adding turns isn't practical the pulse widths going to each drive stage will need to be tweaked (multiplying by a constant - eg Unit A * 1, Unit B * 1.02, Unit C * 0.99 or whatever) to get exactly the same voltages at full load. |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 995 |
I think I need to clarify something, when I said : If it goes bang we can talk again about a controller card (8010) with the inverse opto drive If Pete wants to take up my offer of help, I also have a working Nano controller card with the inverse opto drive that is plug compatible. Pete can also talk to poida about a nano/pico solution and power board with my blessing, but I can't volunteer Poidas help, I can volunteer mine. I only mentioned 8010 as that is what Pete said he had on hand in a control card and I went with that, of course if you go with the nano solution you get Poidas excellent support and a configurable tweakable solution instead of it only comes in this flavour so suck it up! Unless you really want an 8010 solution (and KeepIS is happy with his) I think the nano has a lot going for it but if you dont want to program nanos etc there is an alternative 8010 solution too. Edited 2024-01-24 14:24 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| Murphy's friend Guru Joined: 04/10/2019 Location: AustraliaPosts: 583 |
When KeepIS described his 8010 control board (with wiseguy's mods) I thought I'll try that idea too. So I now have a nano based and an 8010 based control board that are plug in swappable. They both work well in my inverters  , have not managed to blow any mosfets. |
||||
| less_is_more Newbie  Joined: 03/04/2023 Location: SpainPosts: 12 |
Hi, could perhaps give some info of interest here. I was given a couple of LF inverters 700W 230V, spanish brand Atersa, who where working in parallel mode (master/slave) and were replaced when upgrading an offline farm. This means parallel operation of LF inverters is possible. You will find here the spanish manual for a bit more recent model (sorry I could'nt find it in english). Instructions to connect them in parallel is on pages 9-10. The master is still working, although load detection does'nt work, and the slave's driver and control board seem to have been destroyed after a bad repair. Atersa was of no help, didn't have ¿didn't want to send me? the circuit scheme, nor this older model manual. They say it's too old. So I reverse engineered part of it, and one of my projects (from too many!) would be to bring them back to life with a modified picoverter, reusing the driverboard which uses IR2110 Edited 2024-01-24 23:54 by less_is_more |
||||
| analog8484 Regular Member  Joined: 11/11/2021 Location: United StatesPosts: 89 |
It's doubtful that using the same PWM drive alone would be adequate for paralleling in general. Some kind of droop control on at least the secondary inverters is likely necessary for practical paralleling. Commercial inverters also use real-time comms for sharing of inverter output power/current across the inverters. |
||||
| Godoh Guru Joined: 26/09/2020 Location: AustraliaPosts: 378 |
Thanks for the input folks, seems I have opened a can of worms. I am planning on using one transformer now, ( that i need to rewind) and two 3kw powerboards. Mike does the nano board just plug in where the 8010 boards go? Is that what you mean by plug compatible? I will have a fair bit on for a while yet, I need to dig a trench ( 25 metres) for the conduit and cable, build the rack and mount the panels. Then rewind the transformer. I will have a look at the threads on the nano inverter, So far I am pretty happy with the 8010 boards, the ones I am using have all been modified as Oztules used to do. Disconnecting some of the chips that apparently made the inverters go bang. Seems I have a bit of reading to do on nano boards thanks Pete |
||||
| Murphy's friend Guru Joined: 04/10/2019 Location: AustraliaPosts: 583 |
I can answer the first, its NO, a nano is quite different from the plug in 8010 carrier. Plug in compatible usually means the control boards inputs & outputs are compatible. |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 995 |
Well Pete my news is not good. First, I did not realise that you had Ali boards I assumed incorrectly that you had the Mad style boards with a double row connector/ribbon cable to an EG8010 controller card oz/mad style. My opinion of the Ali boards is that they are expecting a lot from the IR2110/IR2113 FET driver IC but they seem to work reasonably well. But to drive 2 lots of FET bridges from the one controller is probably too much to ask. So I think that the only way for it to work is to have 2 more IR2110/IR2113 (and ancilliary parts) to drive the second boards output stage. I sense that this is nowhere the simple add a second ribbon cable to a second output stage and you're good to go that I first envisaged. So sorry for getting your hopes up. When I said if it goes bang I could help, I was looking at one of my power stage PCBs which has a ribbon connector that can connect to either an EG8010 controllers or a Nano controller that are plug compatible. Something tells me that you might just want to continue to live with separate inverters for the various tasks. If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| noneyabussiness Guru Joined: 31/07/2017 Location: AustraliaPosts: 506 |
Not sure what board pete has, but they usually have a npn/pnp booster on the output of the ir2113... so its not doing as much work as you think it is... I think it works !! |
||||
| Godoh Guru Joined: 26/09/2020 Location: AustraliaPosts: 378 |
Thanks Mike, yep the boards I have are Sunyima or similar. Yesterday I ran the home made 3kw inverter all day charging the car. It did that easily, I want to be able to charge the car with a 15 amp charger rather than an 8 amp charger, so I need at least a 3.6kw continuous inverter. That would stretch the 3kw sunyima board and rewound powerjack transformer too far, that is unless I dumped the whole inverter into a tank of transformer oil. I have seen commercial 4 and 5 kw 24 volt inverters so it is possible. My thinking is that I have only ever managed to blow up powerjack inverters in the past. ( pretty unreliable they were) so in order to join the smoky fet club I will need to experiment more. If Noneya is right then I may not get to join the dead fet club, but only testing will tell. It is a way off yet, I still need to buy some magnet wire to rewind the big transformer core that I have, make the panel rack, dig a trench, fit the panels, wire it all up to charge the batteries etc. So blowing up fets may be a way off. Thanks for all the ideas, I really appreciate the expertise here, even if some of it goes way over my head Pete |
||||
| Page 1 of 2 |
|||||
 Print this page |
