New Feature Request on Hypertherm Cutters

[beefy]

Hi Jim,

I can use my 1250 to mark hole centres very effectively BUT my method requires the post cut air flow (about 10 seconds) to stop before my system can send the very short duration signal to turn the torch on momentarily. This video shows a cutter using a similar method but the cutter in use does not have the post cut airflow issue http://www.youtube.com/watch?v=ZLHkV8C0DsQ

Because of this problem I've abandoned a hole marking job. I will take over an hour to mark just the holes themselves, and most of that hour is the post cut air flow (and compressor) running. My only option now is to fit my z-azis with a mechanical centre punch / hole marking system. It would be a great feature to have a method of disabling or reducing this air flow duration to maybe just a second. With the hole marking the torch is on for no more than a quick "crack" so it shouldn't really heat up at all.

I realise this could cause a risk of users forgetting to flip the switch back to normal use, and the nozzle overheating so even if this feature was controlled by a signal from the cnc controller (e.g. a terminal connection inside the plasma cutter).

I'm guessing management won't even consider this for future cutters, because I've never heard of anyone else using my method of hole marking, but I can only try.

Cheers,

Keith

[jimcolt]

My home cnc machine (PlasmaCam) marks hole centers perfectly... almost immediately even with the postflow. Their software uses a short duration plasma firing as well.....but waits to start the timing cycle after it sees arc transfer to the plate. They monitor the plasma arc voltage to do this, but you can also monitor the arc transfer output from your powermax1250.....which is pins 12 and 14 on the cpc connector on the rear panel. Pins 12 and 14 will indicate contact closure instantly when current is sensed on the work ground cable...from there you can issue your off signal to accurately mark hole locations.

You can also stop the preflow at any time by issuing a very short start signal blip during postlow, in other words hit the start and instantly release it during postflow....and the postflow will stop. We do not recommend doing this often, as it will affect consumable parts life...as it will in any plasma torch.

The cutters in the video (Thermal Dynamics) does use postflow, and TrueCut (the machine builder in the video) uses he method I describe above using the arc transfer output to determine when to start the timing for making the marks in the material. Over 80% of the plasma systems they sell are Hypertherm....and this feature works the same.

Our engineers will not consider making the postflow shorter as it is a necessary funtion for consumable life as well as torch cooling.

Wish I had a better answer for you!

Best regards, Jim Colt

Hi Jim,

I can use my 1250 to mark hole centres very effectively BUT my method requires the post cut air flow (about 10 seconds) to stop before my system can send the very short duration signal to turn the torch on momentarily. This video shows a cutter using a similar method but the cutter in use does not have the post cut airflow issue http://www.youtube.com/watch?v=ZLHkV8C0DsQ

Because of this problem I've abandoned a hole marking job. I will take over an hour to mark just the holes themselves, and most of that hour is the post cut air flow (and compressor) running. My only option now is to fit my z-azis with a mechanical centre punch / hole marking system. It would be a great feature to have a method of disabling or reducing this air flow duration to maybe just a second. With the hole marking the torch is on for no more than a quick "crack" so it shouldn't really heat up at all.

I realise this could cause a risk of users forgetting to flip the switch back to normal use, and the nozzle overheating so even if this feature was controlled by a signal from the cnc controller (e.g. a terminal connection inside the plasma cutter).

I'm guessing management won't even consider this for future cutters, because I've never heard of anyone else using my method of hole marking, but I can only try.

Cheers,

Keith

[beefy]

Thanks very much Jim.

I did half expect Hypertherm would not want to make it possible for users to accidentally damage their torches. As often happens when a customer screws up they will blame the machine / manufacturer.

You've gave me a few more ideas to play with and I will pursue them. Hopefully Mach has the speed to do what you mentioned fast enough, but if it doesn't I will make a little hardware circuit to do it directly.

Thanks again,

Keith.

[jimcolt]

Sounds good, and keep me posted, I'll help you any way that I can.

Jim

Thanks very much Jim.

I did half expect Hypertherm would not want to make it possible for users to accidentally damage their torches. As often happens when a customer screws up they will blame the machine / manufacturer.

You've gave me a few more ideas to play with and I will pursue them. Hopefully Mach has the speed to do what you mentioned fast enough, but if it doesn't I will make a little hardware circuit to do it directly.

Thanks again,

Keith.

[Largemouthlou]

Happened on this accidentally on Friday and thought of this post.

Did normal drill operation in SC, turn odd THC in Mach cut profiles.. It pierced the first mark but all others (20) were very little blips.. Seems that is where the extra delay is in mach, with no THC marked in mach it follows the SC set pierce delay.. I was cutting .125 aluminum at the time.

Worked on my older versions of mach anyways.. Let me know if this works for ya..

[jimcolt]

If the "plasma center puching" process is properly set up, starting the arc on timer only after the cnc sees the arc transferred signal (current on the work lead indicating the arc has transferred to the plate), then the pierce dimple will be exactly the same every time. If you are starting a timer for this type of process from the initiation of the start signal....then you will see variations in depth of the center punch, here's why:

-The first pierce of a blowback style plasma starts very quickly because there is no air flowing at the torch.
-Subsequent starts of a blowback torch (if the torch is in postflow) take longer as the air has to shut off and then reactivate in order to move the internal components to refire the torch.
-Torch start / transfere times will also vary slightly based on different starting heights, consumable wear, etc.

Machines that are designed to succesfully do plasma center marks use the arc transfer signal to start the timing process.....all of the above listed timing issues are solved and the center marks come out uniform every time.

Jim Colt Hypertherm

Happened on this accidentally on Friday and thought of this post.

Did normal drill operation in SC, turn odd THC in Mach cut profiles.. It pierced the first mark but all others (20) were very little blips.. Seems that is where the extra delay is in mach, with no THC marked in mach it follows the SC set pierce delay.. I was cutting .125 aluminum at the time.

Worked on my older versions of mach anyways.. Let me know if this works for ya..

[beefy]

I'm looking into building an independent electronic circuit to do my "blipping". I could see how Mach itself would handle it but there's two things I have no control of:

One is the response time. I want real fast response and I want it repeatable. Will Mach respond fast one time but maybe not so fast the next, don't know.

Second is the current level at which I get the "arc OK" signal. I would like to be able to set the "arc OK" current level, so an independent current sensing circuit can give me this. I'm hoping I can make a non-contact sensor using a hall effect device.

If I'm successful, then in addition to getting rid of the post flow waiting time, I'll have the fastest possible response time, independent of Mach. The circuit will effectively give the "torch on" signal to the plasma cutter, then as soon as current is sensed (at a level I set) in the ground cable, the "torch on" signal is removed again. Hopefully this will give me the cutest little crater in the steel surface, that will be just about as accurate as a centre punch mark.

My experience with hole centre marking is that anything more than the shortest crack doesn't leave a nice round accurate crater, the crater becomes more of a shallow hole and loses its roundness. As for piercing hole centres I tried that once and immediately abandoned it.

Jim, can you see anything I'm missing with my idea.

Keith.

[jimcolt]

Sounds like it will work to me.

I don't use Mach in any of my machines....so am not familiar with the process delays associated with it. The PlasmaCam has a function for marking holes.....and it seems to work flawlessly even with the postflow active. PlasmaCam does not use the arc transferred output from the plasma, rather, they monitor the DC Raw voltage and watch for the drop in voltage when the arc transfers to the plate. From this they have instant response for timing purposes, and the amount of time it takes to fire the arc...even if it varies...makes no difference.

Jim Colt

[beefy]

Thanks very much Jim, I was so busy focussing on current, it never crossed my mind to use voltage. That seems like a much more "solid" and accurate method of achieving perfect timing. I certainly like the minds of the Plasmacam designers, they seem to strive for perfection in their control system.

So that seems to mean there's a voltage present before the arc fires, and when the arc fires, that voltage drops. It's times like this I wish I had a digital storage oscilloscope so I could see the timing of the voltage becoming present after the torch on signal. My system monitors the arc voltage but when it's not cutting it's at zero.

I wonder if sensing the voltage rising from zero would work or if that's not possible because the voltage must be present before the arc is struck.

I don't suppose you'd have any idea of the rough timing of the presence of the high voltage. Just off the top of my head I'm guessing it goes something like this:

1. The cutter receives the "torch ON" signal
2. The high voltage is established ready for the arc "striking"
3. The air solenoid turns on, moves the parts in the torch, and the arc is struck.
4. A drop in voltage will now happen

Sorry I can't be a more normal customer Jim

Cheers,

Keith.

[jimcolt]

A bit more complicated.

Power supply activates..DC output is dead shorted from Nozzle to electrode. No voltage.

Air solenoid activates....about 200 or so miliseconds later air pressure building at the torch separates the nozzle from the electrode, air is ionized and current flows from nozzle to electrode.....pilot arc is created. DC voltage can be measured (I don't know what that voltage is), likely between 150 and 300 vdc

Pilot arc attaches to material (work ground cable), current ramps up to cut current, voltage drops to cutting voltage (between 70 and about 160 volts DC).

Measuring current on the work ground clamp should be just as accurate as measuring the voltage change between pilot arc and transferred arc. Contact closure on output of the arc transferred signal is exactly the same time as the voltage transition from pilot arc to transferred arc.

Jim Colt

[beefy]

Fantastic information Jim, thanks very much.

Even if I never make use of it, I love having some understanding of how it works (there must be a nerd in me ).

Seeing as I don't know what the exact voltage levels would be I'll go for monitoring the current at the lowest level I can, and if necessary add some delay too.

Thanks again Jim, great service and great information. Hypertherm all the way.

Keith.

[jimcolt]

If your plasma has the rear panel cpc connector....pins 12 and 14 close when the arc transfers to the plate....everytime, and the timing is perfect. No need to use any other means for monitoring current! The output is a dy contact mechanical relay.

Jim Colt

Fantastic information Jim, thanks very much.

Even if I never make use of it, I love having some understanding of how it works (there must be a nerd in me ).

Seeing as I don't know what the exact voltage levels would be I'll go for monitoring the current at the lowest level I can, and if necessary add some delay too.

Thanks again Jim, great service and great information. Hypertherm all the way.

Keith.

[BTA Plasma]

You could try just dropping your pierce delay. A seperate tool for drill operation in sheetcam should be used when marking holes.

[beefy]

If your plasma has the rear panel cpc connector....pins 12 and 14 close when the arc transfers to the plate....everytime, and the timing is perfect. No need to use any other means for monitoring current! The output is a dy contact mechanical relay.

Jim Colt

Thanks again Jim.

I don't have that connector Jim but I can just as easily access the terminal block inside my 1250. I'll give that a try before I go to the bother of making / buying a current sensor. The reason I was thinking of having a dedicated current sensor is I could tune it to detect low levels of current like 5 amps, and terminate the "torch on" signal immediately. It would also be faster than the "arc OK" mechanical relay output.

Saying all that it is logical to try what is immediately available and then improve the system if I don't get the results I need. I just have a habit of swatting a fly with a sledgehammer.

Keith.

[beefy]

You could try just dropping your pierce delay. A seperate tool for drill operation in sheetcam should be used when marking holes.

Thanks BTA.

I don't actually use any pierce delay. My code used the "ActiveSignalOuput1" and a 100mS sleep command within a macro. So ASSUMING there's no delays from Mach/PC my "torch ON" relay is only turned of for 100mS. However, my hole mark is still too much, so I also cheat and disconnect the ground lead, which greatly improves the "centre punch mark" to a point I personally consider satisfactory. If it wasn't for the 10 sec post cut air flow I'd stay with this method.

However, now I have to sense the current (or use the "arc OK" signal) which means leaving the ground wire in circuit. With my present system, even with nothing but a 100mS signal for the "torch ON" my mark is still too "blown out" when the ground lead is connected (that's why I disconnect it, which helps greatly). Hence my thoughts on a dedicated electronic circuit to control the "torch ON" signal (speed - bypassing Mach/PC) and sensing the ground cable amps at a very low level. The slowest part of the system will be the mechanical relay I use for the "torch ON" signal so I'll have to find the fastest one available. I can't use anything electronic there because my 1250 required a dry relay contact for the "torch ON" signal.

At present I have a dedicated tool in Sheetcam, a dedicated post processor, and a dedicated macro in Mach for the hole marking. Once I get all the above sorted I'll play with the post processor to use conditional statements so I don't have to generate 2 separate gcode files (hole marking & normal cuts) and then manually merge them in Notepad.

Keith