CNC Interface and Torch Voltage Capture

[EmptyNester]

Since you are working with Arduino for the THC, here is something interesting I stumbled across that might interest you.
As an alternative to Mach 3, a complete open source Arduino based CNC, GRBL. I can really see using this for some smaller projects that have been on the back burner for a long time.

http://dank.bengler.no/-/page/show/5470_grbl

http://www.shapeoko.com/wiki/index.php/Grbl

And a slick little driver board to go with it.
http://blog.makezine.com/arduino/grbl/

Ram,

Thanks for the links - those are pretty cool and I hadn't seen the shapeoko or grbl before (and I've done a lot of searching for Arduino/CNC info).

Its a shame the Grbl board can't handle bigger motors or 4 axis. I wouldn't mind replacing the individual motor controllers with an Arduino for the CNC.

I had looked at the Linux based open source CNC software but decided to go the traditional path. I've had projects before where I've taken on too much - then couldn't get to the point that it was functional. Since the THC is optional, that seemed like a good starting point. After it's working and dialed in, I may start looking at other software and control systems.

On the topic of Arduino projects, I think a 3D printer would be cool. I'll be curious to see if I can make one from steel with the plasma table when its done.

Rege

[SeanMurphy265]

The guy that makes the carriage kits that I purchased has been working on a 3D printer.

http://www.needfulthings.net/shopdro...y;threadid=108

[EmptyNester]

I have a 16 character by 2 line display that came with the Arduino Mega 2560 I bought on eBay.

I've been thinking about what type of capabilities I'd want on a THC and how a menu system would work on the small 16x2 dispaly. Assuming the use of four push buttons (<Up>, <Down>, <Menu/Select>, <Cancel>), I came up with the following:



When operating, you can adjust your target voltage.

The menu system would be disabled when the torch is on. To configure it when the torch is off, you'll want to be able to:

• adjust target voltage
• calibrate the voltage reading so the display is accurate
• adjust the anti-dive setting (which I expect would be to damping very fast voltage jumps when going over existing cuts)
• test the different interface lines
• display software version info

I expect the anti-dive would detemine when you're crossing an earlier cut by a faster than normal voltage change. It would then "dampen" the response if it thought that was the case.

Does that seem like a complete list?

[SeanMurphy265]

I do have a question about voltage. For THC to work it reads the voltage and makes the adjustment up or down to get the correct voltage.
1. How do you know the correct voltage for the optimal cut height?
2. How would the voltage read if the torch is too high or too low? Will the voltage read more or less than the correct voltage if the torch is too high or low?

[Rambozo]

During operation I would display the target voltage and the current measured voltage with a display updating time that allows you to read it, say somewhere around 2 to 4 updates per second, and the typical up down indicators for torch motion signals. Other than that, I like your menu structure.

1. There is a direct correlation between torch height and voltage, so you can calibrate one to the other. By measuring the voltage at the desired cut height, you will calibrate your system for all the variables at that point in time. However, electrode wear will cause this calibration to drift a little. I'm not sure if that would be a something that needs compensation or not.
2. The voltage goes up as the torch height increases. (longer arc = higher voltage)

[EmptyNester]

During operation I would display the target voltage and the current measured voltage with a display updating time that allows you to read it, say somewhere around 2 to 4 updates per second, and the typical up down indicators for torch motion signals. Other than that, I like your menu structure.

Thanks Ram.

I menu I have now gives:

• Torch on Status (on/off)
• Arc good status (on/off)
• target voltage
• current voltage

The display is too small to show the up/down arrows, but that makes sense. I'll have to abbreviate more.

Sean's question made me realize that I should have an option to set the cutting voltage based on what's current. That would allow you to start cutting manually and hit the button to get the initial voltage setting.

You comment on tip deterioriation over age makes a "cut time" timer make sense. Once that's in place and its understood, it could be compensated for automatically. Would you only track the time on the tip or would you also track electrode or swirl ring time separately?

[Rambozo]

You comment on tip deterioriation over age makes a "cut time" timer make sense. Once that's in place and its understood, it could be compensated for automatically. Would you only track the time on the tip or would you also track electrode or swirl ring time separately?

From what I have read, the only thing you are compensating for is electrode wear, as this makes the effective arc longer, hence a higher voltage, even though the tip to metal height hasn't changed. I do not know how much wear takes place, so I'm not sure if this is anything to worry about.
I like the one touch calibration while the torch is running. I guess if you could stop the THC, then jog the Z axis while cutting you could find the best height then tap a button and have the THC take over. Something like the set button on a cruise control. Might take a little of the trial and error out of building a chart of best torch heights for your particular combinations.

[EmptyNester]

I like the one touch calibration while the torch is running. I guess if you could stop the THC, then jog the Z axis while cutting you could find the best height then tap a button and have the THC take over. Something like the set button on a cruise control.

I was thinking about how I'd go about setting this up the first time. So, I figured I'd just do a manual cut with a stand-off and use the set voltage function.

Adding the ability to save pre-set voltages for materials and thicknesses would probably be good. Probably the quickest way to implement that is to use the serial interface on the Arduino and write a small PC .NET application to save and recall those settings.

I've updated the menu chart with suggestions.



I took a vacation day today. My goal is to move from the Arduino development tools to Eclipse. For smaller projects like this, it usually takes longer to get the tools set up than to write the first pass of the code.

[Rambozo]

From what I understand the number of piercings has a lot to do with tip life. So I would say that the tip life feature should count number of starts and integrate that into the number of inches cut. I'm sure a simple formula can combine the two. For example, each start equals 10" of cut time. or something like that. I've been looking at how long consumables last in CNC and it seems like if the machine is setup right, they last much longer than cutting by hand. There is a lot to be said for machine consistency.

[EmptyNester]

As I've been working on the menu software I was thinking about calibration and voltage control.

With a 10 bit A/D the range of values read is 0 to 1023. With a max 225 volt input (allowing for a little overshoot) - that would give about .22 volts per A/D count.

Because of the greater resolution in counts, would it make more sense to use the "count" value instead of the voltage to control torch height? The THC would then display set point and actual values for the "count" instead of voltage. This would remove the need to calibrate the voltage reading it and give more accurate control. (Though, you'd have to have some hysteresis to keep from bouncing the torch up and down).

With the "one touch calibration" Ram mentioned - this could work well. You'd go to a menu to set the voltage. While in that mode, the THC is not working. While cutting you adjust the Z-axis to the desired height then hit the "set" button and the THC takes off with the current value as the setting.

[EmptyNester]

Spent some time working on the menus. Rearranged and simplified them as I worked on implementing them. The new structure is below.



The "Cruise Control" name was Rambozo's idea. It will allow you to run without height control until you adjust the height to where you want it. When you hit <SELECT>, it will then take that voltage readnig as a setting and start running (at least in theory).

I'm still not sure if I need to keep the "Calibrate". It's intent is to allow adjusting the actual voltage display.

"Tip Life" is a simple time counter that can be reset. (Potentially could used to compensate for electrode wear in the future.)

"Anti-Dive" would attempt to detect when you're crossing a previously cut line and not send the torch crashing down. This would probably work by detecting a more rapid change than with curved metal. I'm wondering if the setting might be able to be caclulated by using the cutting speed (in IPS) to calculate how fast the drop would be and how long it would last when crossing an existing cut.

Here is a diagram of the hardware configuration I currently have.

[Rambozo]

I would keep the Calibrate option. Since there are inline resistors in the full arc voltage output lines and several possibilities of voltage divider circuits (external and internal) It would be nice to be able to measure at the torch and calibrate the system to read correctly. That way you can use published data tables as a starting point for materials you have never cut before, and be pretty close right off the bat.

[sportbike]

I'm using a PP50.

The pp50 has a CNC connector with arc voltage, divided volts, ok to move, etc? I wasn't aware of that, I thought only the 60C had that?

[Rambozo]

From what I understand even older versions of the PowerPro 256 used to have that. I'm sorry they dropped it, but I understand the decision.

[EmptyNester]

The pp50 has a CNC connector with arc voltage, divided volts, ok to move, etc? I wasn't aware of that, I thought only the 60C had that?

I got my PP50 a year ago, and it has one. I'm pretty sure they still come with the connector.

[EmptyNester]

I would keep the Calibrate option. Since there are inline resistors in the full arc voltage output lines and several possibilities of voltage divider circuits (external and internal) It would be nice to be able to measure at the torch and calibrate the system to read correctly. That way you can use published data tables as a starting point for materials you have never cut before, and be pretty close right off the bat.

Okay, I'll keep it.

If I can find a reasonably priced TTL to RS-232 level converter with isolater, I could add a serial interface and PC app. That would let you save settings based on materials and then download them. But, that would be after I get the basic functionality working.

[everlastsupport]

Okay, I'll keep it.

If I can find a reasonably priced TTL to RS-232 level converter with isolater, I could add a serial interface and PC app. That would let you save settings based on materials and then download them. But, that would be after I get the basic functionality working.

Have you thought about a discrete level shifter with two transistors, cap, diode and some R's, if all you need is TX/RX/GND? That will keep the price down but take a little space (maybe not with the caps for the MAX chips).

Are you looking at optically coupled? You can drop two in on the TX/RX lines for pretty cheap. Is this a 5VDC or 3.3VDC system (I see TTL)?

Sorry for all the questions, I just dropped in and posted. I'm up on ATMEL and MicroChip, just not the Arduino stuff. I have a Rasberry Oi on order . If I see a little any more info on the Grbl I will be getting an Arduino I am sure.

[EmptyNester]

The Arduino boards use ATMEL processors. The board I'm using (Mega 2560) uses the ATMega2560.

The interface between the Arduino and the PP50 is the Arc Good (PP50 dry contacts out) and the Torch On (Arduino dry contacts out) and the arc voltage (PP50 analog out). But, it seems that the arc voltage might be an isolated signal. So, with the Ardunio sitting inbetween the BOB and the PP50, I don't think the I/O needs any additional isolation (though, I'll be curious to see if I have noise issues). I just haven't tested the theory yet because I wasn't ready to blow up an Arduino if I'm wrong. (My table is the priority now and this is just a time filler.)

Longer term, I thought that adding a serial interface to the PC would provide the ability to add a lot more functionality. But, I'm trying to keep the hardware as simple as possible. I did find a complete one chip solution (isolated power supply, TTL to RS-232 and isolation). While it was $19, which wasn't bad for everything in one package, it turned out to be a BGA package.

So, for now I'm not going to worry about the serial interface until I get the standalone THC working. (My table construction is complete, I have to start wiring everything and setting up Mach).

I got my Raspberry PI in August (April order), but by the time I got it I had too many other projects going on. So, it's just been sitting. One of the drawbacks of the PI is the lack of interfaces. There's a German interface kit in the works, but I think it's around $70. Sorta kills the idea of a $35 computer. But, I have seen some stuff on people using Arduino's as I/O processors for Pi's.

[everlastsupport]

I was looking at the Arduino only because on the Grbl and shield Rambozo posted, but like you got busy and have not done much. Doing a couple PIC project for someone right now and keeping me busy, so I can relate.

Keep us posted on your progress, I plan on a plasma table one day. I can borrow 2 anytime, but now the same as having my own (one day).

[EmptyNester]

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

The touch-up paint on my gantry height adjustment wasn't totally dry - so I couldn't start final table assembly tonight. So, a little work on the THC.

I have a better two channel scope now (rather than the pocket scope) - so I got that out and did some voltage captures. It was consistent with what I found before, but I was able to get both the raw and divided voltage at the same time.

I got bold and hooked it up to the Arduino (running on a wall wart) and had a simple loop displaying the average analog value every 500 ms (total of 10 samples taken every 50 ms). I successfully tested it with pilot arc (no cutting) and the value the Arduino reported worked out to what I would have expected. The value was a little bouncy (noisy), so I'll have to do better filtering, but I'm pleased that it fundamentally worked. So, the simple voltage divider should be sufficient.

I'm going to add zener's for overvoltage protection. I figured I'd see what level of over voltage they can take and might change the divider based on cutting voltage rather than max voltage. While it would stress the zeners, it would give better resolution on the reading.