Bad noise

I tied up a few loose ends hardware wise, and decided to do some rough testing with the extruder. In order to drive the stepper, I connected it to the 4th axis output of my mill. I turned on the heater circuit, and waited until it hit temp. Then I turned on the stepper power system. This is where the trouble started.
Temperature readings began jumping around wildly, which caused the actual temperature to drop, the system would see that the temp was reading very high and turn the heater off. I was reading the temp through the serial port monitor on the Ardunio. Thinking it might have been a strange ground loop problem, or timing interference on the CNC system with the serial monitor I tried isolating the board. Separate board power supply, and I attached a serial LCD display for temp readout. No difference, still crazy noise when the CNC was on.
I thought noise might be getting through the thermocouple, so I grounded the shield with no change. Although I checked for conduction paths between everything and the thermocouple, I even tried taking the thermocouple off the extruder and moving it and the board across the room.
A little scoping around and I believe I have found the culprit, the switching power supply that drives the CNC. When it's on there is enough noise radiated back into the power line to put 50mv of noise on an oscilloscope with it's probe grounded that is plugged into the same house circuit. An EMI filter is on it's way from Digikey. A new power supply is being researched as well, as I plan on making an all in one power source/cnc controller and I need a supply that will play nice in that app.
The other big thing is that I have decided I am not satisfied with the current extruder design. Although it does function relatively well, I am not happy with it's ease of manufacture. Too many operations on too many parts, and absent a bunch of fixtures, very time consuming to make en masse. Additionally the whole thing seemed rather heavy, and reducing weight in this area would seem crucial to good performance.
I have almost completed a reworking of the heater block, nozzle, tube and heatsink parts. I am still working on a redesign of the filament drive mechanics. This where I feel the most can be gained, this part weighs the most.