If we have two parallel DC-boost converters connected to the common load, and If the droop control is properly adjusted to insure voltage regulation and equal current sharing end keeping the circulating current within the limits, then what are other possible causes making this system unstable?
Truly Yours
If two voltage regulators are operating in parallel, one should be set-up as a slave to the other. It isn't really practical tp expect them to operate identically. There will always be tiny differences. So, the same signal that the "master" unit is using for control, must also control the slave. If we call them "A" and "B", then as "A" boosts, "B" will see the increase and decrease. Then they will reverse. Oscillations will grow. The time delays in their control loops will not be identical. This is why you must have one operate as the slave to the other.
Dear Charles Stigers
Thanks for your answer, maybe master and slave criterion it has drawback when we lose the master one. I wish to make them operate in normal droop control not using master and slave method. And really I want to know are there any addition reasons making this parallel combinations work with unstable made.
Truly Yours
Hi Mustafa,
I totally agree with you. Master-Slave mode has single point of failure. Moreover, the two converters may be located at different physical locations, making communication between them impractical. In my experience, I have not seen any issues with droop-based implementation or a very recent decentralized approach that I'm working on. I have implemented them in experiments with four such converters in parallel, without any issues.
Cheers
Connect both the outputs through 1 ohm resistors and put a capacitor large enough so that neither of the DC-DC converters had to regulate for load current variations. If possible choose the DC-DC converters that have closest output voltage in no load and full load else the one with higher output voltage will tend to supply more load current and eventually fail. All the best
Thanks for all...
Dear Mayank Baranwal
I am trying to search the issues related to destabilizing parallel connected converters behind circulating current and output voltage mismatch. When you did your experiments did you observed any unstable behavior, for example if you shared the resistive load with CPL .
Hi Mustafa,
Certainly not with the controllers that I'm working on. I went all the way from 62.5W to 600W, both voltage stabilization and sharing were pretty good. I haven't myself tried droop-controllers on more than two converters at large powers, so I do not have the insight that you're probably looking for.
Though you should refer to some papers along these lines: A quick google search suggests - "Circulating current minimization and current sharing control of parallel boost converters based on Droop Index".
Mayank
- Badges
- Science topic