Hello ; We have a three phase soli 1

Hello Jorge, this is Cesar

I read your description of the Delta(3-phase input) and (Star or Y output) voltage stabilizer.

Here is something to think about.
Load current and voltage regulation are measured at the LOAD. That means that (even though this application is AC), (on a phase) you have to treat one wire as hot and the other wire as return, and you are delivering power to the LOAD. Yes, I see you have a 3-phase application. So if you have identical loads on each phase, then it is ok to measure the load current and voltage regulation on one pair of wires. And then you can claim that the same regulation must be occurring on the other phases. And if the loads are matched, then it does not matter that you elected not to use the neutral wire on the secondary side (where the neutral is where the 3 phases (of the secondary side) connect at a common node in the middle). But if the loads are not identical on each of the phases, then you can use the neutral, and monitor the load current and voltage regulation on each of the phases individually.

Here are some useful notes :
to be able to characterize an assembled voltage stabilizer, you need to document component-level data FIRST :

1) you have 3 transformers ( each single phase type)
so, for Transformer 1, apply vin, and measure vout at no load, and then with a 10Amp load. Transformer 2, apply vin, and measure vout at no load, and with 10Amp load. Transformer 3, apply vin, and measure vout at no load, and with 10Amp load.

2) connect the 3 transformers in Delta (input config, primary side). Connect the secondary sides in delta config.
Now you have to apply 3 phase input voltage. So measure and take scope pictures of your 3 phase input voltage. Measure and take scope pictures of your 3 phase secondary voltage. Change the secondary side config to Y (and don't use the neutral wire) - measure and take scope pictures of your 3 phase output voltage. One more change - this time, use the neutral wire and measure and take scope pictures of your 3 phase output voltage. This is one way to characterize the components you are using in this application.



That's about all I can mention for now. Cesar

Hello cesarcesar;
Thanks for your reply, yes we do exactly the load test for each individual transformer as you suggest, that is applying 220 VAC at the primary and checking for 127 volts output, with a load equivalent to 0.578 times the total load of the stabilizer we are designing, then we have good regulation, about 5 % or better.
The problem comes with a three phase balanced load with the neutral disonnected.
It is possible that impedance is lower (better) if we change to delta config also in the secondary?

Jorge.

Hello again, let me offer a link for comparison
it shows that typical regulation is 10 % to 25 %


On a previous email, I indicated that it is important to measure / characterize each transformer when you apply vin, and measure vout (and regulation) at no load, and then at 10Amp load. This tells you "how much" the secondary voltage dips with (light and heavy) load. This dip on the secondary, is also going to show up when you are trying to stabilize the voltage (while drawing the intended load current). But if you do not characterize the "dip", you may think there is something wrong with your stabilizer circuit.

And, it is important to define your rules for measuring regulation. You can choose using voltage-to-NEUTRAL readings, and determine a value. Or you can use voltage-to-voltage values, and determine a value. And you can use power transfer ratio [a] ( power in ) vs (power out with no neutral) and power in vs power out using neutral.

The other item that occurs to me is : voltage stability
Just for a moment (forget the use of neutral)
Without neutral, the voltage should have a certain stability and that is a certain number. And this number should apply to each phase, (and the load should not be toooo light). Then, with neutral, the voltage should have a certain stability and that is a certain number. But, we can't be saying things like : using the neutral, the voltage is stable, and not using the neutral the voltage is not stable. The voltage value (in each scenario) should not jump around. For the case of not using neutral, maybe 3 fluke meters are needed : one on each set of secondary windings of the star connection (not using neutral) ?

Let's keep crunching on this one.