• Have you tried out dark mode?! Scroll to the bottom of any page to find a sun or moon icon to turn dark mode on or off!

diy solar

diy solar

Syncing 2 off grid inverters to achieve 2 phases of 3 phase

Ahh, yes, a neutral can be considered a phase (even though it is actually only part of a phase), but it is a rare item indeed that uses neutral, and two hots to make use of phases...

The phases would be wrong for a 3-phase motor, maybe it would fight or show poor power factor as it put current into the oddball phase.
It would be great for a starting winding, better than a start capacitor in a single-phase induction motor.

Given those two phases you can make the 3rd phase with a couple transformers. Due to transformer voltage tolerance it won't be perfect.

If you only use two out of three phases, you have two sines, 120 degrees shifted from each other. It's not single phase, and it's not good for anything other than rectifying it, or turning one additional phase off to get single phase. Running a 3 phase device on only two of the three phases is asking for trouble.

It is single phase, riding on a common-mode sine wave

We use it all the time over here in the US. 120/208Y comes to a building (e.g. apartment complex.)
Each unit gets two hots and a neutral. Two 120V (to neutral) circuits and a 208V (L1 to L2) which is just fine for single-phase appliances like stoves, air conditioners, etc.

The rest of us use 120/240V split phase, bigger motors and heaters fed 240V but no neutral. The Big motors in the apartments get 208V but no neutral, and don't care about the common mode (104V I think.)
 
OP hasn't answered my question directly, but I assume that the unit will fault and shut off if either phase 1 or 2 goes missing but will run "fine :rolleyes:" without the third phase. The pumps are most likely single phase (fixed speed). The question is whether they are 230 or 400v.

Edit: Since this is geothermal the pump might not even be in the unit. Please disregard........
Yes, exactly. With phases 1 and 2 connected it starts up and runs, with or without phase 3. Take away phase 1 and it doesn’t start at all (this phase runs the main control and monitoring board, maybe some circulation pumps too). Give the machine only phase 1 and it will start up and be contactable over network but reports “inverter error”, so the inverter which powers the compressor needs at least 1&2 for it to run, 3 being optional.
 
So my initial idea was to access the sine wave generator in two off grid inverters and synchronise them so that the high voltage stages were producing synchronised outputs, to give me my two hots and neutral.

No one really came up with a reason why that won’t work yet, just that no one has done it, understandably I suppose - it’s kind of niche. But is it easier to just derive two hots and a neutral from a single phase 230V inverter, maybe with a centre tapped transformer as in split phase?
 
Last edited:
Probably because nobody has the design documents, schematics, or information on the chip sets used, so you would have to start by reverse engineering it, and then determining if the oscillators could be synchronized, or if you could buffer one to drive the other. It’s a non-trivial exercise, And getting stackable inverters isn’t that hard.
 
Yes, exactly. With phases 1 and 2 connected it starts up and runs, with or without phase 3. Take away phase 1 and it doesn’t start at all (this phase runs the main control and monitoring board, maybe some circulation pumps too). Give the machine only phase 1 and it will start up and be contactable over network but reports “inverter error”, so the inverter which powers the compressor needs at least 1&2 for it to run, 3 being optional.
Hots connected, not phases...

A phase is the electrical wave between two hots...
The conductors are not the phases...
 
So my initial idea was to access the sine wave generator in two off grid inverters and synchronise them so that the high voltage stages were producing synchronised outputs, to give me my two hots and neutral.

No one really came up with a reason why that won’t work yet, just that no one has done it, understandably I suppose - it’s kind of niche. But is it easier to just derive two hots and a neutral from a single phase 230V inverter, maybe with a centre tapped transformer as in split phase?
Syncing could be done but it would be difficult (not worth the trouble) and you might as well go with a 90 deg phase shift to have true 2 phase. A transformer can give you a "center tap" to create a single phase neutral (split phase), but not a true 2 phase neutral. The "neutral" in the wye connected 3 phase system is not a voltage "center tap" due to the phase difference in the hot legs, even though it is a physical center tap.

If I were trying to run the heat pump from a cheap 230v inverter, I would first investigate whether the heat pump is drawing current from the neutral connection and if so, what the actual compressor operating voltage was. If the compressor is 230v, then it might be possible to modify the heat pumps inverter/vfd to accept 230v single phase without making it unreliable. This might involve upgrading the rectifiers (or just connecting all three inputs together to use all of them), modifying the capacitor precharge circuit/timer, addressing filter capacitor total value and ripple current, and defeating "lost phase" detection. If your actual operating condition will always be as stated: "It’s a 12kW heatpump and only pulls 2.7kWh in the hot water cycle I’m interested in running during the summer.", then maybe only the lost phase detection needs to be addressed. I am assuming that the different units in that statement was a typo, and that you meant kW in both places.

I share your curiosity and desire to make things work the way I need and want, but I look for the simplest solution and modifying the cheap inverters is the hardest solution, so I wouldn't start there.
 
Syncing could be done but it would be difficult (not worth the trouble) and you might as well go with a 90 deg phase shift to have true 2 phase. A transformer can give you a "center tap" to create a single phase neutral (split phase), but not a true 2 phase neutral. The "neutral" in the wye connected 3 phase system is not a voltage "center tap" due to the phase difference in the hot legs, even though it is a physical center tap.

If I were trying to run the heat pump from a cheap 230v inverter, I would first investigate whether the heat pump is drawing current from the neutral connection and if so, what the actual compressor operating voltage was. If the compressor is 230v, then it might be possible to modify the heat pumps inverter/vfd to accept 230v single phase without making it unreliable. This might involve upgrading the rectifiers (or just connecting all three inputs together to use all of them), modifying the capacitor precharge circuit/timer, addressing filter capacitor total value and ripple current, and defeating "lost phase" detection. If your actual operating condition will always be as stated: "It’s a 12kW heatpump and only pulls 2.7kWh in the hot water cycle I’m interested in running during the summer.", then maybe only the lost phase detection needs to be addressed. I am assuming that the different units in that statement was a typo, and that you meant kW in both places.

I share your curiosity and desire to make things work the way I need and want, but I look for the simplest solution and modifying the cheap inverters is the hardest solution, so I wouldn't start there.
Yes, 12kW when the scroll compressor runs flat out in top gear, probably with immersion heater added too for maximum heating effect during winter, but in summer (which is when I want to solar power it), when using the compressor at low gear to just bring the integral hot water tank up to 40C for showering, 2.7kW is the max I have observed, in fact it starts around 2kW and builds progressively over about 20 mins to peak at 2.7kW. kWh was a silly typo from autopilot fingers.

I will look into whether there is current being drawn on the N connection but I don’t have much appetite for making internal mods to my heatpump which cost around $16k installed. I guess that’s also an argument for not messing with the mains power going to it, but I’m assuming there’s enough protection built in that I can’t fry it unless I do something silly.

The installation manual makes clear connection to 230V single phase is an option with removal of some jumpers on the connection terminals, but also that the 400V 3 phase and 230V single phase versions of the appliance are different and that I don’t get off so easily as just rearranging those jumpers and connections to power the thing with 230V. In fact it warns of “material damage”.

IMG_2341.pngIMG_2342.pngIMG_2343.png
 
The installation manual makes clear connection to 230V single phase is an option with removal of some jumpers on the connection terminals, but also that the 400V 3 phase and 230V single phase versions of the appliance are different and that I don’t get off so easily as just rearranging those jumpers and connections to power the thing with 230V. In fact it warns of “material damage”.

View attachment 277831View attachment 277832View attachment 277833
OK, so you can run it on 230V single phase, the 3-phase is just a red herring for when you have a 3-phase feed. That's good!
or in fact just the 2 phases I need to run my ground source heat pump.
Should read: "Or in fact the single phase supply I need to run my ground source heat pump."

But the answer to your original question is still "Conceptually yes, but it's a lot of work and not worth the trouble, especially since you can run everything off a single-phase supply."
 
OK, so you can run it on 230V single phase, the 3-phase is just a red herring for when you have a 3-phase feed. That's good!

Should read: "Or in fact the single phase supply I need to run my ground source heat pump."

But the answer to your original question is still "Conceptually yes, but it's a lot of work and not worth the trouble, especially since you can run everything off a single-phase supply."
No, as stated the version of the heat pump I have is 400V 3 phase and the manufacturer literature states that the 230V single phase version they sell is internally different.
 
No, as stated the version of the heat pump I have is 400V 3 phase and the manufacturer literature states that the 230V single phase version they sell is internally different.
Oops, I missed that. So your original question is how to make 400V 3-phase (or maybe only single-phase?) from a couple of random inverters? Do you have inverters in mind?
 
No, as stated the version of the heat pump I have is 400V 3 phase and the manufacturer literature states that the 230V single phase version they sell is internally different.
I searched the site you linked but didn't find the wiring instructions. Now I see why it has the potential to draw so much power. Internal heaters. That makes everything more difficult. I can't possibly offer any more ideas without having complete technical information or the actual unit in front of me. It might still be possible, with extensive modifications, but too much risk. I just use my excess solar to heat water with a conventional 230v water heater!
 
I searched the site you linked but didn't find the wiring instructions. Now I see why it has the potential to draw so much power. Internal heaters. That makes everything more difficult. I can't possibly offer any more ideas without having complete technical information or the actual unit in front of me. It might still be possible, with extensive modifications, but too much risk. I just use my excess solar to heat water with a conventional 230v water heater!
The internal heaters won’t be used, I just want to power the main board (control, monitoring), water circulation pumps and scroll compressor up to 2.7kW max. It just needs 2 x 230V phases connected (400V between them) plus neutral
 
Oops, I missed that. So your original question is how to make 400V 3-phase (or maybe only single-phase?) from a couple of random inverters? Do you have inverters in mind?
I think I described them as cheap generic off grid inverters because I have a few under different brand names, Changhi, Mensela, Excellway, all which seem to come off the same production line with exactly the same basic design (different sizes 3kW - 7kW) and all with this same control board with a 3 pin (Vin, Vout, gnd) connection from control board to main board.

IMG_2345.jpegIMG_2346.jpegIMG_2347.jpeg
 
I think I described them as cheap generic off grid inverters because I have a few under different brand names, Changhi, Mensela, Excellway, all which seem to come off the same production line with exactly the same basic design (different sizes 3kW - 7kW) and all with this same control board with a 3 pin (Vin, Vout, gnd) connection from control board to main board.

View attachment 277858View attachment 277859View attachment 277860
OK, so find the numbers on the chipset and poke around with a scope to look for where the sine wave is being generated (probably PWM to the gate drivers on the MOSFETs) and see if you can disconnect from one and connect to the other, with optional additional buffers.
3 pin (Vin, Vout, gnd) connection from control board to main board.
Do you mean the one called "Multi-Function Display"? That might just be a display/controls board,, the main board is probably generating the main oscillator
 
Some inverters of that type do not have the output isolated from the battery and some do. I don't know why they do that, but it could cause problems when connecting them together.
 
I want to say that victron inverters can support two legs of a three phase system in limited circumstances, paging @sunshine_eggo

They can be configured 120, 180 and 240° out of phase and full floating L2 as well as floating L2 that will revert to original when de-synced from a source.

When full floating, the minimum phase angle results in about 180V between L1 and L2 when L1 is synched to a source and L2 is hunting for something to attach to.
 
Some inverters of that type do not have the output isolated from the battery and some do. I don't know why they do that, but it could cause problems when connecting them together.
Yeah, we've already been over the difficulties of syncing two inverters that aren't made for it, but different vendors, different designs definitely add more complexity and pitfalls. At least the OP has inverters that are 'similar', if not even equal sizes, bit that's an implementation optimization if he can even find the main drive signals.
 

diy solar

diy solar
Back
Top