Close

Switching controller chip

A project log for Nixie tube High-Voltage Power Supply

The path building my own 170V power supply to drive nixies

esenesesenes 08/16/2018 at 13:573 Comments

Here it comes the hard point: once decided that this will be a boost-type power supply, the hard point is the selection of the PWM controller IC. In the jungle of the DC-DC controller chips, we have to consider:

Since the idea is to build something new, all the usual design are not the way. This rules out the commonly used chips for nixie power supplies:

... but makes the journey more interesting !

So what are we looking for ? To begin:

  1. A chip by Linear Technology, in order to have a model in LTCSpice to simulate it. If you are new to LTCSpice you can take a look to the numerous tutorials, like this.
  2. A package that can be soldered by hand, so no packages without leads ... and I don't even mention the ball bonded ones
  3. Internal switching transistor or external to be seen
  4. I don't really care of the efficiency, as long as it's reasonable.

<draft to be finished>

Discussions

esenes wrote 08/17/2018 at 08:50 point

Thanks a lot for your comment !

Yes, the idea of the interal switch chips was actually considering the addition of an external mosfet, as it is done on this https://threeneurons.wordpress.com/nixie-power-supply/hv-supply-kit/ using the MC34063 chip.

I'll still try without the voltage doubler and see what happens ! Worst case it can always be added in a second time ;)

  Are you sure? yes | no

K.C. Lee wrote 08/17/2018 at 10:37 point

That looks reasonable.  You won't need 5W.  I got by some were between 1-2W for 16 segments x 20 Nixie old burnt in display probably from a pinball machine machine.

  Are you sure? yes | no

K.C. Lee wrote 08/16/2018 at 14:34 point

Let me pare down your choices.

>Switching frequency solution (up to 1 MHz)

Might want to do a bottom up approach here.  The diode recovery time for one that meets your *voltage rating* likely would be limiting your switching frequency.

Also skip the 555 solution as it is a piss poor choice.  It doesn't have reference source, inductor current control.  Seen much too many fail design that have huge inductors because of the latter.

If it is a non-isolated DC-DC converter, then your switching element should have high enough voltage rating. i.e. 180V or higher.  You might be able to something designed for offline flyback, but otherwise rules out most of the low voltage ones with integrated switches.  You can possibly use external cascode MOSFET/BJT to increase the voltage rating of internal switch.

The duty cycle is a function of the output voltage to input voltages.  Keep that in mind when looking at duty cycle limits of your parts and your minimum operating voltages.

I cheated a bit in my nixie supply as I needed a 80V and a 160V rails.  So I used a voltage doubler which also helps to drop the switch voltage rating by a factor of 2 and let me use the tiny 100V/1A MOSFET in my collection.

  Are you sure? yes | no