CNC Interface and Torch Voltage Capture

[Rambozo]

Any time you don't see electronics smoke is success...

Yeah, sometimes it's a real challenge keeping the magic smoke, inside the chips.

[EmptyNester]

As time has gone on, I've found multiple posts about building an Arduino THC. They always have a lot of discussion and disagreement on voltage isolation and positive versus negative arc voltage.

Then, they always just end. I've yet to see a resolution.

I did some more testing with a scope. It reinforced the belief that the divided voltage coming out the CNC port is isolated. If you measure it, you get a reasonable positive voltage.

If you hook the ground of the scope to the ground clamp and the positive to the CNC port, you get a very dirty negative signal with voltages well out of the digital logic range. So, it looks like its bad to have earth ground connected to any logic circuit in the overall system (torch, Arduino, motor control). The PC is not an issue for the normal interface as I'm using an isolated BOB with seperate power supplies.

Since the Arduino runs on a wall wart with no ground, there shouldn't be an issue with ground loops between the Arduino and the CNC port.

I looked at the 5 volt supplies I have for my motor controllers. The chasis-A/C ground on that is indepenent of the negative signal. So, I think that it is isolated enough that it wouldn't cause the any problems. (I need to run that by my EE friend.) But, it probably wouldn't be a bad idea to put some opto isolators on the logic signals that go to and from the break out board.

I now have all the parts for an op-amp filter circuit. Once I figure out if optoisolators are necessary, I can get any final parts and do the electronics build.

[sportbike]

I've seen a few posts across the internet regarding the output of the voltage circuits on the CNC connection being sketchy and sometimes dropping off midway through a cut. Not sure if it is really an issue, just something to keep an eye on.

I had kind of forgotten about Arduino. I have another potential use for it for a speed controller on a small clutch Dyno I hope to someday get completed.

I have seen a few THC applications that just end as well. Seems weird.

[EmptyNester]

I spent the day at an Analog Devices seminar. They presented isolators for a couple hours. They have some really cool stuff that could be used for this project - if I could solder fine pitch packages/surface mount - which I can't.

I run two power supplies on my Break Out Board so that it is truly isolated (it doesn't share a common power supply).

I realized that I could take one potential ground loop out of the picture by powering the Arduino from the 5 volt power supply that runs the motor side of the BOB and motor controllers (rather than its own supply).

The updated system diagram is below:



I haven't build the op amp circuit yet, but tonight I wired the Arduino to run off the motor's 5 volt supply and had it reading the torch voltage.

It was pretty noisy, but I didn't have the op-amp circuit yet and the filtering was pretty crude. But, it was reading the voltage, scaling it and then displaying it.

At this point I feel pretty sure that this hardware configuration will work from an isolation viewpoint, which was my biggest concern. I just need to start working on the filtering to get a decent reading.

[Rambozo]

Sounds great. I really like the approach you are taking with this.

[sportbike]

I spent the day at an Analog Devices seminar. They presented isolators for a couple hours.

I bet that was a blast.
Probably all kinds of hot chicks there....just like any other engineering related conference....

[everlastsupport]

I bet that was a blast.
Probably all kinds of hot chicks there....just like any other engineering related conference....

Sometime they get hot girls behind the counters to move the goods.

Looks like the project is coming along. Oleg is sending me a PP50 soon and I might have to build a version of this as well.

[SeanMurphy265]

I think you are going in the right direction! I like that you are using the full arc voltage instead of the divided arc voltage. I'm sure there is a very good reason why one would use divided arc voltage for THC. Using the full arc voltage will allow anyone with any plasma cutter to use your THC. I also like that you are developing the OK to move as well. My setup does not use the OK to move, and I know other people have had to jump the OK to move pins to work. This could potentially open up the cnc plasma market for all plasma cutters to have a THC ability.

[EmptyNester]

Looks like the project is coming along. Oleg is sending me a PP50 soon and I might have to build a version of this as well.

Let me know when you get it. I'm planning on open sourcing everything when it works, but I'd like to get somebody with electronics/software experience to be an early tester and debugger.

[EmptyNester]

I think you are going in the right direction! I like that you are using the full arc voltage instead of the divided arc voltage. I'm sure there is a very good reason why one would use divided arc voltage for THC. Using the full arc voltage will allow anyone with any plasma cutter to use your THC. I also like that you are developing the OK to move as well. My setup does not use the OK to move, and I know other people have had to jump the OK to move pins to work. This could potentially open up the cnc plasma market for all plasma cutters to have a THC ability.

I haven't looked at the okay to move pins yet. Have you heard of any problems with them on the PP50?

I didn't have any special plans for that yet, I just figured it would be cleaner to put all the signals through the Arduino. But the side benefit is that they are there if you need to do anything.

I had thought about putting the E Stop through it, but decided not to for now.

[everlastsupport]

Let me know when you get it. I'm planning on open sourcing everything when it works, but I'd like to get somebody with electronics/software experience to be an early tester and debugger.

Sept 24 the plasma are in, Oleg told me today.. Not sure how long to me, maybe 7 days. Should have a PP50 and PP70. And I have a PP60 already. So I can test them all. Glad to test things out. Maybe have a PIC version too

[SeanMurphy265]

I haven't looked at the okay to move pins yet. Have you heard of any problems with them on the PP50?

I didn't have any special plans for that yet, I just figured it would be cleaner to put all the signals through the Arduino. But the side benefit is that they are there if you need to do anything.

I had thought about putting the E Stop through it, but decided not to for now.

This guy has several post, but he had some troubles with the OK to move function.
http://www.cnczone.com/forums/plasma...bot_4_0_a.html

[Rambozo]

This guy has several post, but he had some troubles with the OK to move function.
http://www.cnczone.com/forums/plasma...bot_4_0_a.html

Thanks for the link. I like the idea of a current clamp to monitor arc instead of just the arc ok signal, anyway. Another variable to measure and tune, and can be read out in Mach, too.

[everlastsupport]

Thanks for the link. I like the idea of a current clamp to monitor arc instead of just the arc ok signal, anyway. Another variable to measure and tune, and can be read out in Mach, too.

That was my plan years ago as well, but I never had the time to get the CNC dream going. Reason being, back then we didn't have the CNC port.

I still think it will work that way as well.

[sportbike]

Thanks for the link. I like the idea of a current clamp to monitor arc instead of just the arc ok signal, anyway. Another variable to measure and tune, and can be read out in Mach, too.

I think the external CandCNC method is probably more robust than however the internal is being done.

If I am not mistaken, John from CNCzone has since switched over to a Hypertherm unit. He is the one who was having unstable voltage and similar issues causing the torch to quit halfway through the part, etc.
Not sure it if was just wiring or settings or not. His solution was to swap to a Hypertherm. Cost about $1500 more than the Everlast I believe he noted.

[SeanMurphy265]

I think the external CandCNC method is probably more robust than however the internal is being done.

If I am not mistaken, John from CNCzone has since switched over to a Hypertherm unit. He is the one who was having unstable voltage and similar issues causing the torch to quit halfway through the part, etc.
Not sure it if was just wiring or settings or not. His solution was to swap to a Hypertherm. Cost about $1500 more than the Everlast I believe he noted.

That would be correct! Xalky's thread over there also talks about some noise troubles he had, but he was using a Longevity multi process unit with CandCNC.com thc and electronics.

[EmptyNester]

Xalky's thread over there also talks about some noise troubles he had ....

I didn't read the entire thread, but I did see where he said that he used the divided voltage. Depending on where the noise is being introduced, you could have the case where the noise is a much greater percentage of the overall signal with using the divided voltage instead of the full voltage.

[EmptyNester]

I've been working on the software over time and spent this weekend getting the hardware prototype done.

My goal was to have enough to allow me to test my table without THC functionality. I believe I'm there. It works on the bench, I just need to try it on the CNC table. (Currently focusing on getting all SheetCam and Mach configured properly so I can try to cut.)

I figured that I'd build the prototype cheap and easy. If/when it works well, I I'll do a clean build. I didn't want to spend a lot of time in case I had to rework anything.

So I could test the software, I wired the buttons before I had an enclosure. I make the wires long, so now its a bit of a rats nest.

To prevent any issues with ground loops, I wanted to put the electronics in an insulated box. Project boxes are expensive and I didn't feel like making a wood enclosure for it. So I went to Walmart and bought a cheap plastic food container. They didn't have much to choose from and I ended up having to get one with a vent in the middle. (The hole in the lid.)



The single push button on the left is to set the mode (disabled, run, manual torch moves). I haven't wired it yet because its a late addition.

The two red buttons are for voltage up and voltage down. They will also allow moving up and down through the menuing system.

The black buttons on the right are for the menuing system and are "Menu/Select" and "Cancel".

I looked at a number of different prototype boards. I ended up using one that is made for the Mega board. I put the voltage divider on the board (it includes a filter cap and zener diodes for overvoltage protection). You can see the resistors left of center on the board.

You can see the corner of the relay board on the right behind the Mega board. This is for the plasma trigger.

I haven't finished the voltage sensing circuit. I need to add the amp/filter. I have an op amp circuit that filters off anything over 100 hz. Hopefully that will help with noise.



The box has a DB-9 at each side. One goes to the CNC port and the other goes to the motor control. Since I knew I'd have to work on it, all the wiring connects to the board with plugs that go on header pins.

I had to change the LCD display mounting. It usually mounts on top of the Mega board, but I wanted the proto board there. So, I made a couple "extension cords" that all them to be separated.

[sportbike]

Looks pretty good to me. I use containers line that for all sorts of things. No sense in trying to make everything neat and tidy when you are 99% sure you are going to want to add in another feature or change something.

All it needs is for you to break out the sharpie and label it :-)

[EmptyNester]

So, I have the THC running enough to receive "Torch On" commands and then trip the relay and do feedback of the Arc Good signal. (Don't have Mach looking at the Arc Good yet.) And, I've successfully done my first cut.

So now I'm going to start working on the analog voltage filter. But I was doing some general web surfing and it looks like a number of people have just done a serial link to Mach and implemented the THC in "Brains". But, these are like the other posts talking about homebuilt THC's - they are from around 2010, and then just ended after a couple posts.