High Frequency TIG welder causing UNO issues

``When I was in the overhaul yards aboard a nuclear submarine, TIG welders would drive the instrumentation crazy. If it will do that to very well shielded systems, I doubt you will be able to shield an Arduino well enough.

``When I was in the overhaul yards aboard a nuclear submarine, TIG welders would drive the instrumentation crazy. If it will do that to very well shielded systems, I doubt you will be able to shield an Arduino well enough.

OMG there's another missile launched. Will someone tell that bl**dy welder to stop !

Hey there all
It's late down here but I really appreciate the conversation and suggestions.
In two days I will be having a crack at the modifications and some kind of report. "News of the weld?"
Down here we like building noisy submarines and we just decided to build 12 more. It dawned on me that maybe if you have a noisy submarine and you make lots of noise about it then the other noise doesn't matter so much?
What about 1 inch thick plate for a box?!
I can go to 1 and 3/4?


30 thou of copper is all you'll need for a box. Steel is a very poor conductor of current . Failing copper, go for aluminium.

You would think after all these years impossible deadlines would be easier?
To more important things and to close this loop.
Rather than read the ramblings of a hack if you would like an understanding of TIG process i suggest you refer to the suppository of all wisdom in Wikipedia. I guess the important point is that the HF arc initiation is a spark similar to a Tesla coil. The point being to allow a non contact initiation of the ionized flame without contaminating the Tungsten electrode.
A clean shaped tungsten electrode allows very precise focus of the flame. One dip in the puddle or touch with the filler wire contaminates the Tungsten and the accuracy is lost and the ark starts dancing all over the place.
Some would say “Man why are you doing this and why don’t you just use a MIG?”
I would say " Man i own 5 Mig’s and one of them is a Pulsed Mig $$$$ and the sparks, smoke and splatter while exciting is a bit boring". Great for blasting welds all over the place for production welding but for me it is a bit like decorating a cake with a sledge hammer (really big hammer). My personal preference of process is TIG and it is expensive but I don’t care!
During my downtime since my last post i have spent a lot of time experimenting with the suggestions acquired here and I thank you all for the guidance.
Implemented the INPUT_PULLUP code.
Capacitor across the switch.
Battery to the board trial.
Top of the range shielded switch cable which I then isolated with a relay.
Shielded box.
The PULLUP code makes so much sense and has taught me some trickery that I would not have come across on my own but does not seem to fix the problem.
Capacitor across the switch made a big difference but there is an “Over run” after releasing the switch which in practice is something i could get used to but not ideal. I would prefer the 2 second delayed start than the 2 second delayed stop. I tried every suitable available CAP!
Battery to the board made no difference.
At this point i decided it must be the cable for the switch so i ran the switch through its own relay powered from the wire drive power supply with no resulting change in behavior.
Now I know there will be those of you who will say “Dude give it up. Arduino no good in this environment!”. You are probably right but the quality of weld i am getting and the simplicity of action coupled with the ability to update code as I learn is exciting.
It was my four year old who first suggested that “Maybe the robot brain is scared and doesn’t want to help you weld? Its hiding!” and that was when I first thought switch to power the board.
This works and as I wrote previously it is the shame i will have to live with. I find that solutions come eventually. Until then I thank you again for your assistance.
Trust that my next project will be more complex but I will choose one without the noisy environment.
My belief is that products don’t come with any consideration for drink holders anymore!! (see photo)
Please excuse my silliness.
Until the next time.
All the best.
int_hayes out.

I reckon you'll get there in the end! You got shielded box, and RF chokes on cables etc.

You probably tried big electrolytic capacitor (eg 500 microFarad) between arduino 5V supply rail and arduino 0V ground? That would be for smoothing and decoupling.

Actually, if any other chips are used that don't have decoupling capacitors, should put decoupling capacitors as close to the power pin as possible, eg..... 100 nanoFarad decoupling capacitors.

Hello all. I am new to this forum. I am also a beginner in Arduino & electronics.
I have developed a wire feeder to be used for DC pulsed welding of Stainless steel using Arduino Uno.
After reading this post I have understood that there are some noise issues during HF starts.
Before I actually use my feeder, I would like to prevent such problems.
So coming to my question,
Int_Hayes, how did you actually solve the problem? When you say "switch to power the board" it means something like a spike guard?

Ithis is my last project of linear TIG welding

Here's another topic on the cold wire feed on the drive stepper motor and arduino.

I work as a welding/machine tool engineer & have made a tig seam welding plc controller for airbus which works fine, now. I had to deal with hf problems so I reverse engineered Hitachi suppression units, as the arduino only seemed to crash when plugged into the welding set or positioner. I then went on to cross wire a small inductor transformer so that it cancelled out the interference waveform in its coil. This even works directly connected to the torch so I can measure voltage.. It was a complete pain but I hope that helps as I was ripping hair out myself.. I am having problems with building a cold wire feeder programming wise as using millis is driving me mad but can't use delay due to the reaction time when adjusting the retract & speed etc.. If you have Any pointers would be great, also any problems you have with suppression etc I can help no peoblem.

Im not sure if it was you who asked for the solution to this.
I have built units for airbus and I'm surprised you don't get a blank screen every now and then even dc welding.. Then again my unit is incorporated with the welding turntable which tig welds aerospace pipes.
As ive invested a lot of time and money, 4 odd years! Solving this problem and it's patented I can make you a unit which will cancel out hf to the point that you can connect a cheap multimeter direct to a tig set with hf on for a few minutes and still get a voltage reading.
Im only offering this as your not a company as i don't want my work duplicated.
Let me know.

Is this like .... a false trigger issue?

See what happens if you put a big electrolytic capacitor eg...few hundred microfarad (or more) between the supply pin of the Arduino and the ground (0V) pin of the Arduino.

High power RF might well sail straight past that - for real protection you need feed-through capacitors and
a totally shielded enclosure preferably with RF gaskets.

In fact an electrolytic cap's leads might act as a loop antenna and make things worse, depending on many
variables of course.

In an environment with severe RF flying around nothing beats a decent Faraday cage.