Low cost test setup for 3 phase pv inverter

[butech]

Hi,

does anybody know a low cost (<1000$) solution for testing a 3 phase grid connected pv inverter (power rating <1.2 kW 3 phase 210Vrms). Usually a 4 quadrant ac power source (chroma et. al.) would be needed, but commercial ones tend to be very expensive (even at this moderate power level 600W would be also ok). Basic operation could be tested using three resistors, but it is not possible to test synchronization and or maybe reactive power control. My first idea was to build a circuit with which acts like an ac electronic load (in "constant" voltage mode with three sinusoidal voltages) maybe using something similar to three apex high power opamps?

Or has anybody a simple idea?

[f4eru]

what modes, faults, etc.. do you want to test ?
What frequency ?

[NiHaoMike]

I have never seen a 3 phase grid tie inverter with a power rating less than a few kW. I suppose a sufficiently large rotary phase converter (to keep the impedance low enough for stable operation) could work.

[butech]

what modes, faults, etc.. do you want to test ?
What frequency ?

This is for an academic project, we want to test the basic operation of an self designed inverter. Therefore, basic efficiency at different power levels, anti-islanding mode, synchronisation of control to grid frequency. As modes we want to test active power delivery to the grid, as well as reactive power compensation depending on the grid voltage. Frequency is nominal 60 Hz.

[f4eru]

60 Hz? according to your profile, you're in Germany, why 60Hz ?
if frequency is fixed at the line frequency, I can recommend a 3-phase transformer (with some capacitors to compensate for inductance)
https://www.ebay.de/itm/Drei-Phasen-Regeltrafo-380V-Stelltrafo-Spartrafo-Ringkerntrafo-Trafo-regelbar/352293036598?hash=item52064d2236:g:tLUAAOSw-2JfEL5D

Another possibility is to use a motor inverter, followed by a 3-phase LC filter.
more complex solution, but can vary the frequency. You also need to reinject the resulting DC power somewhere. (brake chopper, or reuse it at the input of your D.U.T.
For the motor inverter, I can recommend an omron MX2. You ccan vary the voltafe and frequency separately by writing a few lines of code.
For the filter inductor I can recommend using a big asynchronous motor with the rotor blocked, followed by some caps.

[David Hess]

What about testing it with an externally excited synchronous motor, if you can find one?

[jbb]

If money’s really tight, you can do a lot with a variac as follows:
- make sure you’ve got good protection: undervoltage, overvoltage, overcurrent etc. If possible I like to do it in hardware for my development units
- think about safety! You have dangerous voltages, large capacitors, risk of debris flying into your eyes, risk of small fires etc
- internally disconnect output inductors from AC line and short term together (models 3 phase grid with zero voltage)
- connect line voltage sensors to 3 phase line so PLL can lock
- set line voltage feedforward term to zero
- you can now test current loop stability against a low impedance and verify that the phase current aligns with the voltage as you expect.  Note that some PLLs can lock to the negative sequence component of the 3 phase line, so check phase rotation too (ie measure at 2 line voltages and 2 line currents)
- once control loops stable you can return the inductor connection to normal
- use three phase variac to begin testing at low voltage
- use variac to test over varying line voltage (efficiency, VAR injection )

[butech]

60 Hz? according to your profile, you're in Germany, why 60Hz ?
if frequency is fixed at the line frequency, I can recommend a 3-phase transformer (with some capacitors to compensate for inductance)
https://www.ebay.de/itm/Drei-Phasen-Regeltrafo-380V-Stelltrafo-Spartrafo-Ringkerntrafo-Trafo-regelbar/352293036598?hash=item52064d2236:g:tLUAAOSw-2JfEL5D

Another possibility is to use a motor inverter, followed by a 3-phase LC filter.
more complex solution, but can vary the frequency. You also need to reinject the resulting DC power somewhere. (brake chopper, or reuse it at the input of your D.U.T.
For the motor inverter, I can recommend an omron MX2. You ccan vary the voltafe and frequency separately by writing a few lines of code.
For the filter inductor I can recommend using a big asynchronous motor with the rotor blocked, followed by some caps.

Yes you are right I am in Germany, but actually the unit should be tested for US three phase voltages, therefore 60Hz nominal.But a test with 50Hz should not be a problem, for testing the basic operation.

[butech]

If money’s really tight, you can do a lot with a variac as follows:
- make sure you’ve got good protection: undervoltage, overvoltage, overcurrent etc. If possible I like to do it in hardware for my development units
- think about safety! You have dangerous voltages, large capacitors, risk of debris flying into your eyes, risk of small fires etc
- internally disconnect output inductors from AC line and short term together (models 3 phase grid with zero voltage)
- connect line voltage sensors to 3 phase line so PLL can lock
- set line voltage feedforward term to zero
- you can now test current loop stability against a low impedance and verify that the phase current aligns with the voltage as you expect.  Note that some PLLs can lock to the negative sequence component of the 3 phase line, so check phase rotation too (ie measure at 2 line voltages and 2 line currents)
- once control loops stable you can return the inductor connection to normal
- use three phase variac to begin testing at low voltage
- use variac to test over varying line voltage (efficiency, VAR injection )

Thank you for your post. You have noted quite a lot of things to consider. Essentially I did much design of commercial power electronics in a range of 10W - 10kW (mainly SMPS) therefore I am aware of the safety issues. The main problem is that I did not do much about power feeding into the grid. Therefore the used equipment is limited to this kind of applications, where you just can use some kind of (mainly) DC load to test the electronics. Money is not really that tight, but as I do not know if there are any following projects, it will be difficult to spend more than a 4 digit sum for equipment.