Another plasma component...

Andy

Thank you for the reply. I think the idea will require a fair bit of plumbing. The air supply to the plasma cutter is away from the machine and would need to be plumbed through the all the drag chains to the floating head. The air connection from the plasma power supply to the torch head is in a composite cable and is not access-able. I will have to consider whether using this air supply will be a problem. The blow back torch uses the air pressure to move parts to strike the arc and then the cutting air pressure at the torch is controlled by the power supply box for the correct cut. I actually can do this as my torch is actually a ?Franken Torch" made from parts of two manufactures, most people will be loath to start altering their expensive and carefully calibrated torch assembly.

John

On further thought this will not work. The post cut airflow is usually still happening when you want to be probing for the next cut. So yes it will work for single cuts but not for the much more common many serial cuts. I could hack the plasma power supply air solenoid circuit but the results of interrupting post cut cooling flow aren't known to me.

John

OK, so a separate valve would still be needed. The idea might still work, but not as easily as I thought.

But those guided cylinders really do look like a floating mount waiting to happen

Well, there is another reason to plumb air up to the Z axis and that is to operate a pneumatic engraver. Normally, this requires 2 feeds, one to operate the air ram and the other passing through an oiler to operate the engraver. The separate feed to operate the air ram is required to control the down pressure applied when engraving. I did buy an engraver for this purpose and kinda a have it back on the agenda as I have to replace the drag chain with a larger one and add Ohmic sensing now I have a larger plasma cutter that supports Ohmic sensing. I must get some 6mm air line and plumb it in while I am at it!

John,

There is a problem with the current setting of thc-enable and cut-height from gcode in that if you connect an analog or digital out to one of the new hal pins then they cannot be set from the GUI as they are outputs.

I have just pushed a version (Gmoccapy only at this stage) that does it by remapping M10 (currently unused) which seems to work ok.
P and Q arguments are used to achieve this
P0 = thc-enable, Q0 disables and Q>0 enables
P1 = cut-height, Qnnn. is the height

A code other than M10 could be used, I only picked it because it was unused and I didn't want to remap the digital/analog codes in case a user required them.
Edit: actually after re-reading the docs they cannot be remapped anyway...

Cheers, Phill.

Phill

Thank you for all the effort on the cutting control from Gcode. I will look at the code in a few days. The analog output was chosen because others had used them in past. One of the original ways cutting parameters were passed was using the S spindle speed word in Mach3, no analog codes there. At one time there was a short lived discussion here about creating a loose "standard" of codes used for cutting parameter control from Gcode, I just ended up using the same codes as CandCNC does. I really will not be able to work on this until Tuesday or Wednesday when the forecast is for rain. Too much outside related stuff at this time of year to work inside during dry weather. I will start porting over my hal files so I can drive the machine. Is the M10/M11 pair of codes usually used as fast on and fast off for lasers?

Guys

The guided cylinders do look like a nice starting point for a floating head, unfortunately I already have one that works. What is really needed is a good reliable non contact method to find the top of the material. I am guessing that since none of the big commercial machine builders are doing this for plasma there must be a good reason. I certainly will not be installing a crazy noisy pneumatic plate marker any time soon. My compressor lives outside to try and control the noise in the shop and I still need to wear hearing protection when cutting because of the exhaust fans.

John

islander261 wrote:
Is the M10/M11 pair of codes usually used as fast on and fast off for lasers?

That is probably correct, I also see that Hypertherm and Thermal Dynamics use M10 to turn Marker off. It would be good if there were some sort of consensus and what codes to use.

I have done some preparatory work to allow other overrides from gcode, I have done all the coding for this in Gmoccapy but have not mapped them to any M codes yet, looking at this table leads me to think I should stay away from codes in the 00-99 range as these may be used internally for LinuxCNC at a later date. I may just use some code/codes in the high 800's. I am also unsure if I should use a single M code with P and Q words, P being the item and Q being the value or whether to allocate one M code to each item and use the P word for the value...

The items I have overrides for are:
material, thc-enable, use-auto-volts, pierce-height, pierce-delay, cut-height, cut-feed-rate, cut-amps, cut-volts, puddle-jump-height and puddle-jump-delay
There are no real overheads to doing this so more items could easily be added if required (or deleted if you think I have gone crazy)

The only items mapped to M codes at the moment are thc-enable and cut-height as mentioned in this previous post.

Cheers, Phill.

Probably the wrong place to ask a question:
I'm not in plasma and think I'll never be but curious.

As I know plasma emits a lot of uv light. And I'm convinced there must be a relation between distance and light emissions. Did anybody tried to measure and control via that aspect ?

Phill

Those that have newer HT machines will be able to use your work with Pedro's RS-485 work to have full control over their plasma power supply from Gcode. I believe that the lower numbered M codes (99-199) are reserved for some sort of internal LinuxCNC use so a set of high numbers is good. There is one thing that I forgot to mention. I think one of the reasons the analog output codes are used is that you can synchronize them with machine motion so you can output changes that don't stop machine motion. I know that my use of them is done so that the THC enable command is synchronized with the start of X&Y motion to not break the motion buffer. THC disable is just a normal asynchronous code because it usually occurs after a M5. The exact format of the codes doesn't matter much from where I sit, after all the PP does all the work inserting them after you have written it.

Mike

Thanks for adding to the discussion. I am sure that someone who isn't into plasma cutting will be able to see the forest for the trees. Right now most use the arc voltage to sense the torch height for a given cutting speed and current. The present discussion was about finding the top of the work piece so that accurate pierce and initial cut height can be established, typically the torch isn't on when this happens.

John

Mike_Eitel wrote: Probably the wrong place to ask a question:
I'm not in plasma and think I'll never be but curious.

As I know plasma emits a lot of uv light. And I'm convinced there must be a relation between distance and light emissions. Did anybody tried to measure and control via that aspect ?

I think voltage is probably the best way as its much easier to measure. In my tests where I logged 16,000 readings with halsampler at different heights, the linear relationship between height and voltage was accurate with 99.4% confidence. I measured 7.53 volts per mm.