Electric Automation Forum
Forum » Power Supply » 90V to 265Vac Input with PFC & Output of 60V, 3000W
Topics: 90V to 265Vac Input with PFC & Output of 60V, 3000W on Power Supply
#1
Start by
Ram
09-08-2013 10:46 PM

90V to 265Vac Input with PFC & Output of 60V, 3000W

Currently my solution is based on IR1150 for the PFC & UC2825 for the Full Bridge Hard switched converter. Its a old design & works well though its bulky & lossy. Now i'm planning to make it more efficient & less bulky. I'm planning to use multi phase PFC based on Texas Instruments. For the full bridge, i'm wondering if i could use LLC or Phase Shift. Pls share your ideas & experience.
09-08-2013 10:46 PM
Top #2
Kok Fei
09-08-2013 10:46 PM
Hi Ram, what is your application for this supply?

Multi-phase for PFC should be a good choice as per your aim, as per present component populairty and performance, how many phase you will use ? do you have hold-up time requirement ?
09-08-2013 10:47 PM
Top #3
Ram
09-08-2013 10:47 PM
Its a Telecom rectifier. I plan to use 2phase for 160V to 265V & 4phase for Universal Input. But i cant decide between phase shift & LLC for the DC-DC section.
09-08-2013 10:47 PM
Top #4
Kok Fei
09-08-2013 10:47 PM
Hi Ram, I would suggest to use Full Bridge Phase Shift as you already have hard-switching experience. TI, Linear,.. have good controller can do the jobs well. Their references is easy to understand also, if by vote many expertise will pick FB phase shift.

LLC is a good solutions also if you able to prevent its "bad behavior", if interested you shall consider Full Bridge LLC, not Half-bridge. Reason is Full bridge will help you to handle smaller Primary circulating current, probably less hassle.

LLC, there are numbers of papers out there for references, resonant behavior is where you have to pay attention. It's power components are rather none straight forward to determine. Some IC like ST L6699, NXP TEA1716 already had some good features in it to prevent transient failure, but they are for Half-bridge, I have not try how to trick around it to work for FB LLC, anyhow, by understand them you may know better want to consider LLC full-bridge or not.
09-08-2013 10:48 PM
Top #5
Ram
09-08-2013 10:48 PM
Thanks. I completely agree with you. LLC bad behavior has caused me lots of damage when it some times fails in No Load Tests.I plan to use ISL6753 this time for the ZVS Full Bridge. What is your opinion on ISL6753?
09-08-2013 10:48 PM
Top #6
Ray
09-08-2013 10:48 PM
No load failures can happen in the phase-shifted bridge too. Most likely the same event - forced commutation of the the antiparallel diodes at light load.

No bridge is "easy". Expect prolonged development and testing times before you can claim the full benefits of bridge topologies. Pain and experience is a necessary part of the process to get to high performance.
09-08-2013 10:49 PM
Top #7
Girish
09-08-2013 10:49 PM
One option would be to retrofit the existing full bridge with passive lossless turnoff snubbers of the type shown in Carroll - . US patent 4,403,269. The leakage inductance of the transformer already ensures minimal turn-on losses. Thus, the efficiency goal can be met with minimal changes to the existing circuitry
09-08-2013 10:49 PM
Top #8
Kok Fei
09-08-2013 10:49 PM
Agree with you Ray, is true "no Bridge is easy" and looks like so far no perfect topology in place, however, the Full Bridge phase bridge weakness have better manageable switching mechanism, as its delta with conventional full-bridge is not much, not like LLC.

Minimum load is always a not to neglected consideration when come to design, minimum load is Buck family "branded" characteristic, some member you just can ignore it but for bridge better have some precaution, anyhow, phase shift still better than LLC.

Cheers !
09-08-2013 10:50 PM
Top #9
Ilya
09-08-2013 10:50 PM
1. It is against the regulations to consume 3000W/η from 90VAC line due to 16Arms (in some cases 20Arms) line allowance.
2. CCM PFC (if you know how to create a lossless snubber) is much less expensive than two or more phased PFC running in DCM.
3. The optimized full bridge topology depends on your specific needs since there is no best for everything topology (otherwise everybody would use it)
09-08-2013 10:50 PM
Top #10
Kim
09-08-2013 10:50 PM
Have you ever considered 3 phase LLC topology? I believe it is very promising. The only problem is that there isn't any analog controller for it, so you must use a DSP for it. Please refer to this, Multi-phase resonant converter and method of controlling it
US 20100328968 A1. 3 resonant inductors and 3 transformers can be one of 3 phase transformer. You may dislike 3 phase as 6 switches are a lot. The advantages of this topology are that you can use lower voltage secondary rectifier diodes and much smaller
Reply to Thread