Linkswitch-TN High-side BuckBoost with opto
Posted by: Quintin
on
I would like to know if it is possible to use the LinkSwitch-TN in a High-Side Buck Boost configuration, using a TL431 and an opto for the feedback. The application notes only list a low-side buck boost. I am expecting a maximum voltage of 390Vdc at the high side, and I would like -3.3Vdc at the output.
If I use a TL431 as the feedback-reference, can I still expect very tight regulation tolerances on the output?
Will the circuit be safe if the LinkSwitch were to fail?
Thanks
Comments
You should be able to use opto coupler. Since output is 3.3V there is a headroom problem so you will need to use TL431V which has a 1.25 voltage reference. The output of the TL431V can go to a sufficiently low voltage to allow drive an opto LED fro 3.3V
Thanks for the input. I would like to attach my shcematic just so everyone is clear on what i would like to do. Please give me your comments on this topology.
Hello Quintin,
PI_Surak already answered your question. Don't forget RZ, check http://www.powerint.com/sites/default/files/product-docs/an37.pdf page 4.
Cheers,
PI_Crusher
Hello Quintin,
I have to assume you need a floating power supply for driving high-side MOSFETs, part of a bridge topology (or similar). What make the original “low-side” Link-Switch application to be low-side or high-side conversion circuit is not the schematic but how you connect the input and output voltages. All you have to do is to use the same original schematic with no changes at all, see http://www.powerint.com/sites/default/files/product-docs/lnk302_304-306.pdf. Just connect the negative rail to the Source of your High-Side MOSFET, the DC output to your MOSFET Driver, and you’re good to go.
Be aware, there is a catch. Your Link-Switch application becomes now a floating configuration with the “ground” connected to a high voltage, high frequency switched output. You have to provide the high-voltage input not trough traditional slow AC rectifying diodes, but trough some very fast high voltage diodes capable to charge a small input capacitor (eventually trough a small inductor or resistor). Our part can stand up to 700V; I do not see any problem with your 390V requirement.
Cheers,
PI_Crusher