I've designed two PCBs using the LT3758- one is a SEPIC, one is an inverting converter. I want to use them to generate about 100V and -100V respectively. I've never designed a switching power supply before and am aware that they can be finicky, so I've used the schematics and suggested PCB layout from the datasheet as closely as possible. I've also simulated both of the boards in LTSpice, and while I know simulations never correlate perfectly to the real world, both of the simulations worked ok.
When I first made up the SEPIC board it didn't work, so I populated a second board. I was extra careful with this second SEPIC board and once it was populated it worked nicely. I've just populated an inverting converter board and it also doesn't work.
What I've found though my troubleshooting is that the GATE output the LT3758 on both of the non-functional PCBs isn't firing properly. It is as if the LT3758 is stuck in a fault state, even though:
1) The SHDN pin voltage is almost at VIN.
2) INTvcc pin is at 7.8V. It is using the internal regulator and is bypassed with a 4u7 capacitor.
3)SS pin is at 2.5V, and is bypassed with a 0.47uF
If I pull the SHDN pin low then INTvcc and SS both drop to 0V.
What I have noticed is that when I pull the SHDN pin low then then GATE pin goes to 0V, but when I pull the SHDN pin high then the GATE pin sits at a constant 0.68V. This happens on both the SEPIC and inverting converter board and I'm at a loss as to why.
I've inspected all the soldering work under a microscope and it looks ok. I'll come back to it again tomorrow with fresh eyes but if the soldering was bad then I'd expect the LT3758 to not power up at all (no voltage on the INTvcc or SS pin)
I've attached the schematic for the SEPIC board. It is taken pretty much straight from the datasheet, although the MOSFET and diode are different.
Any help would be greatly appreciated.
[edited by: BJHenry at 12:41 AM (GMT 0) on 1 Mar 2019]