Hi

I am trying to source a Grid Tie Inverter (about 20-40KW) that regulates the DC input Voltage within defined limits . The inverter's DC input is connected to a DC/DC converter that has an adjustable constant current output so the inverter needs to regulate the DC bus voltage . Note, I do not want an MPPT regulated inverter input , it must be a constant voltage regulated input . For example , the Inverter should regulate the DC input voltage at say 400V DC if it is a single phase inverter connected to a 230V single phase AC supply or about 800V DC for 3 phase (400V L-L) AC supply .

Do you have any suggestions/comments ?

 

In my tears of engineering I never heard of a device regulating it's INPUT voltage. how could that be done?? It is the power source device that determines the input voltage, and if that is an adjustable DC/DC converter then that device must set the voltage.
But you state that the device supplies a constant current
It is not possible to achieve what is requested,so I suggest instead describing what the overall function you desire , rather than requesting an impossible creation.
AND, one more consideration is that if this is a "for profit" project, commercial consulting fees will be applicable.

 

Hello,

I dont see why this would be impossible. That's because a lot of input sources are 'soft' not as hard as we would like. It does depend a lot on the source though and what you connect between the source and the inverter.
For example, if you were to connect a 10 Ohm resistor between a 12v battery and an inverter, the inverter load would determine the voltage on the inverter input. Feedback would be the voltage, and the control law would simply be a function like Vin(t)=K*Pout(t) or something like that or more complex.
We could do a short study on this mathematically to see what would happen. This would be necessary because as we allow the output power to change, the input voltage changes, and that could also affect the output power level, so we'd have to take that into account. It depends on how much the input affects the output and how much the output affects the input.

Strange feedback circuits sometimes come up. Another example is an LED that regulates it's own forward current using an op amp. Normally you would need a reference voltage to set the forward current at some set level that doesn't change much, but since the LED is an approximate voltage reference, you can use that to regulate the current through the very same LED. Sounds crazy, but it works because the forward voltage dv/di changes much less than the forward current di/dv, or another way of looking at it is dv/dt is much less than di/dt. That means the voltage, being a nonlinear function, sort of levels off while the current changes almost linearly. It reaches a state of equilibrium where the DC voltage is constant and so is the DC current. The breakeven point would be something like A*dv/dt=B*di/dt or A*dv/di=B*di/dv, roughly speaking.

Of course, the system in question has to be modeled and tested using math or maybe even a simulator.
There is one catch, the input cannot go too low on its own because then lowering the output power would not be able to be used to raise the input voltage back up again.

 

The main problem with schemes like that is the possibility of active negative resistance slopes. It's very easy to start power oscillations in control loops with non-linear loads.

You can get some very odd transfer functions depending were you are on the slope.
https://en.wikipedia.org/wiki/Negative_resistance

https://forum.allaboutcircuits.com/threads/dealing-with-negative-resistance.107553/

 

The inverter's DC input is connected to a DC/DC converter that has an adjustable constant current output

Why? Why not use one with constant voltage output in the first place? Most of the DC to DC converters I have seen will do either.

 

My response is to what was requested, not to what I think is needed to do whatever it is. And the TS has asked for a rather non-obtainable creation.
"Regulating an input voltage from a constant current source" ?? For those who claim it can be done, an explanation of the method will educate the rest of us. So please post the explanation of how it would be done.

Now I am heading out to the Field Day Exercise and so will be gone until Monday. So there is time to create a solution. Good Luck!!!

 

My response is to what was requested, not to what I think is needed to do whatever it is. And the TS has asked for a rather non-obtainable creation.
"Regulating an input voltage from a constant current source" ?? For those who claim it can be done, an explanation of the method will educate the rest of us. So please post the explanation of how it would be done.

Now I am heading out to the Field Day Exercise and so will be gone until Monday. So there is time to create a solution. Good Luck!!!

Hi again,

It should be obvious that you cannot regulate anything with a constant anything else, it has to be variable and change with feedback.
Maybe I had presumed wrong, but it sounded like he wanted to vary the output power to regulate the input voltage, which may be doable.
Will there be stability problems? That would have to be investigated through math or simulation or hopefully both.

Hey, have a good trip there when you go out.

 

I am trying to source a Grid Tie Inverter (about 20-40KW) that regulates the DC input Voltage within defined limits .

I'd be surprised if you could find one. Surely it would be the responsibility of the DC source to provide the regulation?

The inverter's DC input is connected to a DC/DC converter that has an adjustable constant current output

In that case, providing a variable load to the converter output could change its output voltage to suit what you want. But it won't be easy if you're handling 20-40kW.

 

You can regulate anything as long as you can develop a feedback system that is stable. As I was saying before, there are some unusual feedback systems that do not seem like they would work, but yet they do, and that is because some mechanism works in a way that allows that to happen. Of course it could take some extra ingenuity to get it to work right.

Now that i think about it though, i would have to ask why anyone would want to do this. A guess would be that they use the voltage source for something else in addition to the converter input, and they would like to have that voltage be somewhat constant at the same time that they are operating the converter.
There would be other things that have to be looked at though before this system can be worked out, if in fact it is possible, and that would be determined in the process.

Somewhat of a comparison could be made between this and a max power tracker I think. The max power tracker has to keep the solar array at a given point where the solar array puts out the maximum power available from the available solar insolation at the time. That is somewhat the same thing, because the inverter is used to regulate the solar array operating point, which is again also the input to the inverter/converter. The load power is varied in order to get the array to that point. This involves measuring several quantities and using that to determine the proper feedback over time.

 

My question remains, if what you want is constant voltage, why are you using a constant current source?