Post Go back to editing

Fly-back converter using LT3757 to generate 170V from 12V

Hi,

I am trying to use the LT3757 to create a flyback converter whcich converts 12Vto 170V. I have based my circuit off the typical applications circuit on page 30 of the data sheet and a hack-a-day Nixie power supply from here: https://github.com/tonyp7/170v-nixie-power-supply - I have used a DA2034 coil (10Apk, leakage=0.25uH, interwinding capacitance=128pF) instead of a DA2032 (3Apk, leakage=0.150uH, interwinding capacitance=67pF) - my mistake when writing down the product number - both coils are 1:10.

The purpose of the circuit is to provide a bias for some custom detectors.

However, I am not getting any output at Vout... I suspect that my Vds waveform is a bit kooky (see image below, plot T scale is in us) there is a huge spike when the gate switches off - would this have caused damage to the Vsense input of the LT3757? or is this causing Vsense to shutdown the oscillator?


Also I have noticed that the pulses seem to be very short when comparing them to the switching wave forms of the data sheet - below are my waveforms. I have the RT=140kOhm = 100kHz so I was expecting a pulse to begin every ~10us but seem to get one every ~24.5us, have I misunderstood this?





History on the circuit:

I initially noticed that I had misread the pin out and connected the coil polarities the same way (i did get Vout=~50V)  but have since changed the polarities so that they are opposite - now I get no output

I also did not have the RC circuit attached to the primary as I had followed the nixie tube set up - I have now added one as per the data sheet.

This is the schematic I have used:

I would appreciate any advice or pointers.

If there is any useful information that would be good to help diagnose what is going on I can provide it.

Thanks

  • When the FB pin voltage is low, the switching frequency folds back, as described in the datasheet (page 13).

    Try to increase the value of the 10nF cap connected to the SENSE pin to 0.1u or above. If it helps, then please use the suggested DA2032. The DA2034's parasitic capacitance is very large.

  • Thanks for the advice W.G. I really appreciate it!

    Increasing the capacitance on the sense did indeed help, the gate pulse was extended improving the duty cycle (see new traces). It did not get me the entire way to my desired Vout but increasing the capacitance further improved the duty cycle and Vout more but not completely.

    I have as you suggested ordered the correct DA2032 and will solder them in place to test again (with the original 10nF cap for sense).

    As you can also see the voltage spike on Vds is still there, albeit with a smaller amplitude, so I will be adding a snubber into the circuit following the Linear Technology Application Note 19. For completeness I will post my results.