1) With forward design, multiple output, does it really matter which winding is used for regulation? I think that as long as the PWM regulates on one output, then all the windings will have the correct voltage (not like flyback, for example)
2) I need a bias winding for the PWM. I would prefer to make it flyback flavor for simplicity. With the rest of the transformer in forward mode, plus the active reset, I'm afraid there won't be enough energy in core to supply the bias supply. If so, I'll just go ahead and make the winding forward mode as well.
regards
gene
2) I need a bias winding for the PWM. I would prefer to make it flyback flavor for simplicity. With the rest of the transformer in forward mode, plus the active reset, I'm afraid there won't be enough energy in core to supply the bias supply. If so, I'll just go ahead and make the winding forward mode as well.
regards
gene
I would need to think about this a little more to be sure, but I don't think the forward converter regulation issue is much different from a flyback. So I would regulate the highest power output, but be sure none of the others go out of spec. Yes, I would make the bias winding a forward winding, but it will cost you another inductor.
I will design with population option so that I can choose any of the outputs for regulation. At least the flexibility will let me recover if it turns out wrong.
Seems to me that flyback and forward are a lot different. On forward, no matter which output is chosen for reference all the other outputs are set by turns ratio. Since no power is stored in core, no problems of one winding starving another, right?
Seems to me that flyback and forward are a lot different. On forward, no matter which output is chosen for reference all the other outputs are set by turns ratio. Since no power is stored in core, no problems of one winding starving another, right?
A forward converter and a flyback converter are a lot different, but the secondary voltages of the forward converter are related by the turns ratios. Neglecting the diode drops and assuming the output inductor currents never go to zero the average output voltages of the forward converter are Vave=D*Vsec.
The output voltages are related by the turns ratios, but there is the so-called issue of "cross-regulation", meaning they will not all regulate as well as the one output voltage that is regulated, reason being the different magnetic couplings, winding resistance, different rectifiers etc etc.
As mentioned by sawreyrw, regulating the highest power output is a good idea, but its still necessary to check that the other voltages do not go too high. you may use post-regulation (e.g. a linear reg) or take of a turn or so to keep these voltages in spec.
As mentioned by sawreyrw, regulating the highest power output is a good idea, but its still necessary to check that the other voltages do not go too high. you may use post-regulation (e.g. a linear reg) or take of a turn or so to keep these voltages in spec.
One reason I bring this up, is the high-power winding may not always be active. This supply is going to +/-40 (amp), +/-7 (pre-regulated analog supply) and +5.0 (logic). It is conceivable that the 40V portion may be inactive at times. Maybe the answer is that the 40V rails need a minimum load in order to keep the regulation - so place some active or passive load to keep things tidy.
back at it with this supply
The active reset smps is *much* harder to tame than I ever thought - good learning experience though and it's actually coming along well. Here's the problem: I used a FAN7390 for the high-side active reset mosfet driver. It is working correctly at 85-90 Vin, with duty cycle limited to around 15%. But - set the duty cycle to 50%, and FAN7390 works until the supply is turned off. Then, and only then, it develops a power to ground short. Always at power down. I did try to drop the Vin slowly - the pwm UVLO is set to 50V, and as soon as it hi 50V, it shutdown as expected, but also the FAN7390 developed the short again.
As near as I can see, the voltage on its pins are within spec. Even at 90Vin, and 50% duty cycle, the main mosfet Vds peaks at around 250V. The mosfet is fine with this, and the FAN7390 is also supposed to good to 600V.
Fairchild-semi suggested a different version of the part, that has some additional protection. I don't like the option because the package is larger. But, at this point, think that a better solution is gate-drive transformer.
Has anyone got some experience with these active-reset forward designs?
The active reset smps is *much* harder to tame than I ever thought - good learning experience though and it's actually coming along well. Here's the problem: I used a FAN7390 for the high-side active reset mosfet driver. It is working correctly at 85-90 Vin, with duty cycle limited to around 15%. But - set the duty cycle to 50%, and FAN7390 works until the supply is turned off. Then, and only then, it develops a power to ground short. Always at power down. I did try to drop the Vin slowly - the pwm UVLO is set to 50V, and as soon as it hi 50V, it shutdown as expected, but also the FAN7390 developed the short again.
As near as I can see, the voltage on its pins are within spec. Even at 90Vin, and 50% duty cycle, the main mosfet Vds peaks at around 250V. The mosfet is fine with this, and the FAN7390 is also supposed to good to 600V.
Fairchild-semi suggested a different version of the part, that has some additional protection. I don't like the option because the package is larger. But, at this point, think that a better solution is gate-drive transformer.
Has anyone got some experience with these active-reset forward designs?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.