Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Synchronous rectifiers sometimes unworkable?

Status
Not open for further replies.

cupoftea

Advanced Member level 5
Joined
Jun 13, 2021
Messages
2,507
Helped
53
Reputation
106
Reaction score
108
Trophy points
63
Activity points
13,087
Hi,
In the following LTspice sim of a 300W Full Bridge (12vin 32vout 300wout 250khz), the synch rect fets CDS rings with the Llk of the txformer, and the result is way too much loss in the snubber.
This ringing cannot be stopped since lossy damping in this loop is too dissipative.
Would you agree, that in this case, the snych rects FETs are just not workable?
 

Attachments

  • fullBridge300w_CD SR.zip
    3.5 KB · Views: 68
  • Synchronous rectifiers not good.pdf
    191.9 KB · Views: 85

Have you exhausted parts-selection as an avenue? For example maybe a much smaller FET gives you less Cdg/Cds and also is the resistor of the snubber (or part of that).

At 12V in you are in the middle ground where dumb rect Vf@If inefficiency vs sync rect complexity and switching loss are a negotiation.

Have you begun to look at low voltage GaN as candidates for this role?
 
Getting synch rec working well is a combination of > 10 year experience along with knowing what chips work well ( trial and error ) and which don't and how to conquer the limitations of some of the better ones. Also getting the total layout right and having a power circuit that doesn't contribute to spiking up the Vds when a fet diode finally has to turn off ( this can be very tricky ). Real parallel diodes are sometimes a good idea if they are near schottky equivalents. Keeping the Freq down allows real snubbers which can be useful - often active catch ckts are used here.
Good luck and don't believe too much in a sim.
 
Thanks, and now in the attached, this LTspice simulation with synch FETs is now working...to make it work needed a swap to a FET with much lower junction capacitances. The idea about the Schottkys sounds good, ....in the attached case, the voltage is too high for normal Schottkys so i reckon its worth doing this with a SiC FET.....it means not so much of the diode current will be transferred to the synch FET, but some will, so it seems like an OK compromise.
At least......do it with a SiC Synch FET, then see about getting a lower rdson FET in there at some future date.....but for the short term, put in enough heatsinking to allow it to work without synch FETs....as is known, few design schedules permit time for tuning synch rects to get working really well. If the synch FETs dont work like a dream on the first power-up of the first prototype, then the project either gets scrapped, or the synch FETs get sidelined for the "time being".
 

Attachments

  • fullBridge300w_CD SR_NOW WORKS.zip
    3.5 KB · Views: 64

Aah sorry, my sincere apologies, in the top post, i appear to have used a general Si diode instead of ultra fast....(now connected as attached)....so i have not demo'd the problems here...But there are some , and if its OK i will return with them
 

Attachments

  • fullBridge300w_CD SR1.zip
    3.5 KB · Views: 64

the short answer here is that: if you can't get a design going well ( sim or otherwise ) with good fast diodes - then the chances of turning that into synch rect are pretty slim ....
 
Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top