Please recomend secondary side synchronous rectifier drivers for use with HiperLCS and HiperTFS?
Hi,
Please recommend secondary side synchronous rectifier driver chips for use with HiperLCS and HiperTFS?
Does Power integrations plan on making any of its own?
コメント
The best situation for a synch rect driver is to actually be able to switch off before the pri fet switches, instead of simply reacting to it switching, which is the modus operandi of the SRK2000 sec side sync. rect. driver.... (so SRK2000 is a "reactive" type sync rect driver, instead of being "predictive"...the Inno switch is superb and is a "predictive" synch rect driver)
Page 15 of the SRK2000 (synch rect driver used in DER385) datasheet warns that some mosfets will have too much bond wire inductance and will instigate faulty behaviour of the synchronous rectification....and the mosfet may need swapping out......such sensitivity seems very poor, and in fact, indicates that these kind of "reactive" type sec side sync rect drivers are dangerous.
Power integrations employ a Superb group of SMPS engineers to make demo boards and eval boards, but it says it all that they have only used a “reactive” type sec side sync rect driver chip on one occasion (DER385). That was the SRK2000 discussed above. (whose datasheet tells of the perils of using it)
And in DER428, power integrations use “natural drive” (no chip) sync rectification for a 2TFC with 14A output….but even then only do sync rect for the power diode…not across the freewheel diode.
This says it all to me about “reactive” type sync rect drivers. They are plain dangerous. They promise vast reductions in sec side rectification losses, but are too risky. Do you agree?
Power Integrations has never tried to make a synch rect driver IC like the SRK2000, and would you agree, that this is because they are very poor and not worth making?
The innoswitch of course, coordinates the pri and sec side fets superbly, and so is a brilliant synch rect driver.
In a perfect world, that would be a better solution, but the primary side of the LCS does not talk to the secondary half - the same with the TFS-2.
The "reactive" solution used in the DER-385 and DER-428 worked quite well with a few caveats, like any other engineering solution. Symmetric layout was necessary in the DER-385 for proper balance between phases. Also, SMD mosfets would be preferable to leaded through-hole mosfets to reduce the effects of the leads on operation. It was also necessary to optimize the RDSon value for the mosfets to balance off loss against having enough signal to guarantee proper switching.
Often it is necessary to work with what you have.
Details on sync rectifier controller for LCS (as well as application considerations) can be found in the report for DER-385, available at the PI main web site. We have never tried using a synchronous controller for TFS-2, but a self-driven synchronous rectifier scheme was used in DER-428, also detailed at the PI main web site.