I'm trying to implement the PMP7760 reference design by TI (http://www.ti.com/tool/pmp7760)

It uses a flyback transformer namely, the G094215LF which I cannot seem to find anywhere. I found several other threads about this where nobody seems to know what's going on with it and TI hasn't seem to provide an alternative since the design is quite old. I was wondering if anyone would have any idea of what alternative I could use to it in that PMP7760 design

PMP7760: http://www.ti.com/tool/pmp7760

BOM: http://www.ti.com/lit/pdf/snvr020

Schematic: http://www.ti.com/lit/pdf/snvr019

  • \$\begingroup\$ Unfortunately the primary turn could be the usual 30, but if the source voltage is over 700 then the primary is about 48 turns of 24 gauge magnet wire. You figure out the turns ratio from there. It is an old design. Not sure why you must use this design. AC ripple on the primary is going to be high. Finding long outdated parts is more by luck than chance. Old existing stock may have dried up long ago, so it is not even in databases as an EOL part. \$\endgroup\$
    – user105652
    Dec 29, 2018 at 6:22
  • \$\begingroup\$ @Sparky256 I have a very unique setup in which I need to step down 500V to 24/12V. I have it explained here: electronics.stackexchange.com/questions/414097/… \$\endgroup\$ Dec 29, 2018 at 16:39

1 Answer 1


The best "Alternative" is to use magnetics and IC's currently supported by TI.

WebBench software can create this for you. I tried your parameters and found 68 designs including using the UCC2801. Yet not without issues , such as overtemp at Vin min, Iout max.

  • \$\begingroup\$ I just tried that software out. Really cool stuff. Thanks for letting me know. I've messaged some TI people about just finding a replacement for that transformer which seems the easiest. The PMP7760 is still an active design, it's just the transformer that is no longer made by Gci. \$\endgroup\$ Jan 10, 2019 at 2:47
  • \$\begingroup\$ That design belonged to NSC which is obsolete since the Gci no longer support TI who bought NSC. I guess the demand was too small. You must create a TI login then use their design tools to create a better parts list from your specs. webench.ti.com/power-designer/switching-regulator \$\endgroup\$ Jan 10, 2019 at 3:16
  • \$\begingroup\$ Yeah, just went through like 30 of the possible designs with my parameters set as Vin = 420V-588V (3V to 4.2V range for our 18650 cells in a 140s config for an eV racing car) and a isolated 12V output at 1A (our LEDs will probably use around 6W I think but we may need to add 1 more so it could go up to 9W). Can't find any design that doesn't have temperature issues hence why I wanted to stick with the PMP7760. I'll do some more research. Thanks for the help though. I appreciate it. \$\endgroup\$ Jan 10, 2019 at 3:23
  • \$\begingroup\$ Redo your design requirements pls so it is simpler. You need a flashing Light. How bright [? Candella] whereas I have RED 10Cd at 1m that are blindingly painful using 2.1V @20mA or 42mW and you need LEDs with 5.5W total???? why!! Stick to the real acceptance criteria \$\endgroup\$ Jan 10, 2019 at 4:46
  • \$\begingroup\$ It's for an Formula Student competition. The rules specify you need a red/green led (can be seperate or together) one facing forward and one facing backward installed on the main roll hoop. The requierments say that it must be visible in direct sunlight and clearly visible from a 3m horizontal distance with a visible angle of 360 degrees (excpet that which is blocked by the main roll hoop). From that, I've gathered that other teams seem to be using 3W RGB led in the front and a 3W RGB led in the back. (only 1 color on a time - so either red or green). That's where my 6W comes from. \$\endgroup\$ Jan 10, 2019 at 17:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.