Close

Turbos and Filament controls

A project log for Vacuum System for Deposition

I have been putting a system together for the past few years to do different kinds of deposition.

jerry-biehlerJerry Biehler 04/29/2014 at 07:110 Comments

Well, the turbo I had hoped would work seems to have a short circuit somewhere in the pump. With no documentation available I dont even know where to start. The plan is to sell this pump and controller as well as another controller for a Osaka Vacuum turbo on eBay and hopefully that will cover the price of the controller for the big pump I have. Then I cross my fingers again that that pump is OK.

I got the little Leybold Turbovac 50 installed on the RGA, this will allow me to sample partial pressures up to 150mtorr in the chamber. I learned some stuff about these little leybold turbo drives in the process. First, and most important, is the cabling. The manual lists 5 wires, three for the phases of the pump, one shield, and one ground. No where in any of the manuals do they mention that this shield is at +160v DC! I had the shield and ground swapped and blew just about every logic IC in the drive. I bought an older NT-50 drive off ebay to replace it and, of course, it didn't work right. I lucked out and Leybold gave me the schematics and waveforms for the drive since they dont repair them anymore and managed to get it working. One open resistor and a bad comparator later and it works.

Next thing is the evaporator control, the big transformer has got to go, it just pulls too much on 120v and the variac that is used to control it cant take that kind of power for long. And then there is the issue with the ground being used as a neutral. Pretty much everything is wrong.

So my solution is using a big old 5v, 600A switching power supply to supply current to the filaments. I am using some big mosfets, some LEM current sensors, and a Teensy 3.1 to create a PID loop controller that should allow me to control the power to all the filaments simultaneously. The boards are made and stuffed, now I just need to program and test. 

Discussions