http://www.ce-mag.com/archive/02/03/goksu.html
or look at chapter 6 of:
http://eepia.kaist.ac.kr/~gwmoon/lec_pe.htm
The WEB is full of PFC and charge pumps all
variations,.....!
or look at chapter 6 of:
http://eepia.kaist.ac.kr/~gwmoon/lec_pe.htm
The WEB is full of PFC and charge pumps all
variations,.....!
ChocoHolic said:
Some years back the word was, power factor correction would be a requirement for PC SMPSs and other consumer electronics in the near future. That time has passed and I'm wondering what happened, as well.
I used the passive and active boost method of PFC in 1986/87. Don't like the passive method much. The current averaging inductor for single phase input is large for a medium power converter, but not so bad for 3 phase input. Active boost method was an R&D effort only, but seemed like a good method at the time.
Amen to that! Just thinking about a 1KW flyback gives me nightmares. I was given the duty of re-designing (make it work) a 200 watt flyback with many secondary voltages. Tried our best to talk the director of engineering into changing to a buck derived topology such as a forward converter. We were told again, to just make it work. We did the best we could with what we had to work with, then a few years later a decision was made to re-design using something other than a flyback.
The good thing about a flyback is there is no output inductor and the bad thing about a flyback is there is no output inductor. Of course, it does not have a transformer, but a coupled inductor. For high input voltage low current they can be ok. For low input voltage and high current, I say no to flyback. Just depends on the application, I guess.
Thanks for finding the great links to PFC!
hi everyone..interesting topic..though i am not an expert (by any means), you are right about the whole PFC issue, but i believe that current trends are set for SMPS > 300W, and i think due to the cost of PC supplies, a lot of these do not have this feature..basic need for this is to make sure that the phase of I and V are nearly the same (cos (angle) = 1)
SMPS isnt the easiest topic...magnetic component selection being the most frustrating...but i think that gapped transformers are used to extend the BH curve so that the core is safe from saturation (correct me if im wrong please)...as far as i understand it, MOSFET designs are much simpler, as you only need a current limit resistor at the gate and limiting zener and capacitor from gate to ground. design is a lot simpler than for bipolar where the transistor has to be biased using atleast 2 resistors..the only problem with FETs is the as the current rating gets higher, jucntion capacitances tend to increase (on/off time problem) and also Rds on MAY increase (this has been improved over time)
as for the output filter, this is a hard topic for me...all i understand is that the configuration of the diodes depends on the topology...BUT in all cases, the resonant frequency of the LC filter MUST BE atleast 10 - 20 times greater than the switching frequency or the ESR in the capacitor will cause problems...
im here to learn..so any suggestion would be greatly appreciated
SMPS isnt the easiest topic...magnetic component selection being the most frustrating...but i think that gapped transformers are used to extend the BH curve so that the core is safe from saturation (correct me if im wrong please)...as far as i understand it, MOSFET designs are much simpler, as you only need a current limit resistor at the gate and limiting zener and capacitor from gate to ground. design is a lot simpler than for bipolar where the transistor has to be biased using atleast 2 resistors..the only problem with FETs is the as the current rating gets higher, jucntion capacitances tend to increase (on/off time problem) and also Rds on MAY increase (this has been improved over time)
as for the output filter, this is a hard topic for me...all i understand is that the configuration of the diodes depends on the topology...BUT in all cases, the resonant frequency of the LC filter MUST BE atleast 10 - 20 times greater than the switching frequency or the ESR in the capacitor will cause problems...
im here to learn..so any suggestion would be greatly appreciated
I find MOSFETs actually harder to use in applications requiring galvanic isolation and a control circuit connected to the output side
To drive a full bridge of MOSFETS from an isolated source you'll need a transformer and four mosfet driving circuits made with discrete components or expensive ICs. Driving MOSFET gates directly from a transformer doesn't allow for fast rise and fall times so you end with higher than reasolable switching losses and big dead times
Having the control circuit tied to the secondary side also requires an auxiliary transformer-coupled power supply to power the control circuit at startup
In the other hand, a bipolar half/full bridge is easily driven from a transformer, a capacitor and some diodes and resistors, requires no drive current from the control circuit due to proportional drive schemes and is able to self oscillate in a controlled way and provide auxiliary power to the control circuit for start-up
Bipolars doesn't provide lower losses but help reducing component count and costs
In the other hand, IGBTs have I/O/R capacitaces an order of magnitude lower than MOSFETs so fast rise and fall times are easily obtained with very simple driving circuits [even with direct transformer-gate drive]
Try to design an AT or ATX 250W PC PSU using only MOSFET switching devices and you will find the crude reality : Higher component count, higher complexity, higher costs [maybe not so high in ATX units because a standby supply is already required]
My next step is to try self-oscillating startup principles applied to IGBT full bridges, since fast and cheap bipolar switching is not suitable for more than 10A/350V and there are some nice, cheap and fast 20A/600V TO-220 IGBTs in the market waiting for somebody to make good use of them
To drive a full bridge of MOSFETS from an isolated source you'll need a transformer and four mosfet driving circuits made with discrete components or expensive ICs. Driving MOSFET gates directly from a transformer doesn't allow for fast rise and fall times so you end with higher than reasolable switching losses and big dead times
Having the control circuit tied to the secondary side also requires an auxiliary transformer-coupled power supply to power the control circuit at startup
In the other hand, a bipolar half/full bridge is easily driven from a transformer, a capacitor and some diodes and resistors, requires no drive current from the control circuit due to proportional drive schemes and is able to self oscillate in a controlled way and provide auxiliary power to the control circuit for start-up
Bipolars doesn't provide lower losses but help reducing component count and costs
In the other hand, IGBTs have I/O/R capacitaces an order of magnitude lower than MOSFETs so fast rise and fall times are easily obtained with very simple driving circuits [even with direct transformer-gate drive]
Try to design an AT or ATX 250W PC PSU using only MOSFET switching devices and you will find the crude reality : Higher component count, higher complexity, higher costs [maybe not so high in ATX units because a standby supply is already required]
My next step is to try self-oscillating startup principles applied to IGBT full bridges, since fast and cheap bipolar switching is not suitable for more than 10A/350V and there are some nice, cheap and fast 20A/600V TO-220 IGBTs in the market waiting for somebody to make good use of them
hi eva
i do agree with you about what you said regarding MOSFETS...i didnt want to go too much into it since my knowledge is shallow...i would have to say though that for direct chip driven topologies such as the forward, flyback, buck or boost, the implementation of the MOSFET is much simpler than BJT's...as for half bridge/push pull or even full bridge, i would have to say that isolation would be an issue...as to how big an issue i am not sure..BUT i have seen it done in many UPS systems utilising IR driver chips for isolation !!...let me know what you think...i do not have much design experience such as yourself...so i may be wrong...may i ask how you go about designing your PSU's ?...would you be kind enough to share a design and explain how you go about designing?
i do agree with you about what you said regarding MOSFETS...i didnt want to go too much into it since my knowledge is shallow...i would have to say though that for direct chip driven topologies such as the forward, flyback, buck or boost, the implementation of the MOSFET is much simpler than BJT's...as for half bridge/push pull or even full bridge, i would have to say that isolation would be an issue...as to how big an issue i am not sure..BUT i have seen it done in many UPS systems utilising IR driver chips for isolation !!...let me know what you think...i do not have much design experience such as yourself...so i may be wrong...may i ask how you go about designing your PSU's ?...would you be kind enough to share a design and explain how you go about designing?
Eva said:
Hi Eva (eslamejor indeed
)Sorry I missed this interesting thread when it occurred, just in case you're still around...
Your design seems to make a lot of sense, do you have any schematic or PCB layout available for the 1kW smps you described in this thread by any chance?
Best regards
SMPS AT & ATX
hi Eva, I have some experience in high power, but now i´m working with low power and i´m totally lost it´s a new world. I could see that you have a good experience in PC power supplies so i would like to know if you can help .... i´m having some problems to design a proportional base drive in a half bridge aplication....Do you have any experience on this kind of drive?
Where can i find more information about that?
hi Eva, I have some experience in high power, but now i´m working with low power and i´m totally lost it´s a new world. I could see that you have a good experience in PC power supplies so i would like to know if you can help .... i´m having some problems to design a proportional base drive in a half bridge aplication....Do you have any experience on this kind of drive?
Where can i find more information about that?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.