I am designing a device that combines multiple different functions into one device. It will be primarilly a combination battery charger and regenerative motor controller, but I will add a pure sine wave inverter on if I can.
At the heart of the device will be a multi-directional DC-DC converter. I am planning to use a push-pull Flyback Converter to keep the different areas electrically isolated. I need for power to be able to flow in just about any direction to/from 5 different DC stages.
I wish to check my reasoning on how this will work:
Each DC stage will have a fixed nominal DC voltage. The ratios of the windings in the transformer will match the desired DC stage voltage. The size of the wiring and transformer core will be sized for the max desired average energy transfer.
When a DC stage is transferring energy out it will turn on the transistors and send a current pulse through the transformer. Flux will rise in the core and when the pulse turns off that flux will rapidly ramp up the voltage across the other cores until the diodes in one of the other stages become forward biased and current will flow in that coil only, charging the capacitor in that DC stage only utill the flux drops back to zero.
My theory is that whichever DC stage is transferring power out will be lowering the voltage on its capacitor only and when another pulse goes through the transformer it will be the only stage to charge. I am hoping that I do not need any other mechanisim to direct the flow of energy: One stage pulls energy out and one stage pushes energy in. The energy will flow properly controlled solely by the relative DC voltages of each stage. Am I right?
I have full bridges on everything so I can alternate pulses and not saturate the core. I also hope to be able to use the transistors as ideal diodes to increase efficiency. I am also wondering if I can add a small low power, low voltage winding around the core that is tied straight to a resister. What can I learn if I monitor the voltage on that resister? Can I measure energy transferred? Core saturation?
I plan to control each stage with its own microcontroller with one master controller taking care of all of the main switching through optical isolation.
I think ATX power supplies do something similar, using multiple secondary windings to feed different voltages. Am I correct?
Thanks in advance for your feedback.
-- Paul
At the heart of the device will be a multi-directional DC-DC converter. I am planning to use a push-pull Flyback Converter to keep the different areas electrically isolated. I need for power to be able to flow in just about any direction to/from 5 different DC stages.
I wish to check my reasoning on how this will work:
Each DC stage will have a fixed nominal DC voltage. The ratios of the windings in the transformer will match the desired DC stage voltage. The size of the wiring and transformer core will be sized for the max desired average energy transfer.
When a DC stage is transferring energy out it will turn on the transistors and send a current pulse through the transformer. Flux will rise in the core and when the pulse turns off that flux will rapidly ramp up the voltage across the other cores until the diodes in one of the other stages become forward biased and current will flow in that coil only, charging the capacitor in that DC stage only utill the flux drops back to zero.
My theory is that whichever DC stage is transferring power out will be lowering the voltage on its capacitor only and when another pulse goes through the transformer it will be the only stage to charge. I am hoping that I do not need any other mechanisim to direct the flow of energy: One stage pulls energy out and one stage pushes energy in. The energy will flow properly controlled solely by the relative DC voltages of each stage. Am I right?
I have full bridges on everything so I can alternate pulses and not saturate the core. I also hope to be able to use the transistors as ideal diodes to increase efficiency. I am also wondering if I can add a small low power, low voltage winding around the core that is tied straight to a resister. What can I learn if I monitor the voltage on that resister? Can I measure energy transferred? Core saturation?
I plan to control each stage with its own microcontroller with one master controller taking care of all of the main switching through optical isolation.
I think ATX power supplies do something similar, using multiple secondary windings to feed different voltages. Am I correct?
Thanks in advance for your feedback.
-- Paul