Drill dimple issue

I have an issue when trying to make a drill mark instead of a full pierce. What I have done is that in Sheetcam I have made so that I use a drill operation for just making a pierce wherever I have a small hole, to later drill it to proper size. I tried removing the ARC OK (in qplasmac) for this operation and use a short pierce delay (~0.3s) and it does a really beautiful dimple easy to drill later instead of a full pierce. However it only works every second "pierce". First works fine, second not, third is OK, fourth not and so on. What can this be because of?

If I turn on the ARC OK it does a pierce every time flawlessly but of course it will pierce through the whole sheet in these cases.

Does anybody have an idea what the reason for this could be? I use a Mesa 7i76E with a Powermax 65 plasma running QtPlasmaC.

Spotting is a bit of a mixed bag because we are doing stuff the plasma cutter was never designed to do. We are turning the arc off before an arcok is received.

I seem to remember that the Hypertherms are fussier than other machines when a few of us were testing it out after I thought of the technique.. I get good results on my Thermal Dynamics with 220ms, but that is too quick for the HT's. So some HT testers could not do it until after the air flow stopped.

If I use 200 ms, nothing happens except a bit of air..

Also the relay dwell time can be an issue although we now start counting when a voltage threshold is reached to try and avoid this. I use 10 volts for this setting.

Its normally the first dimple that does not go well with hypertherm machines (the plasma spider forum is full of posts on the subject)

Most people just put the first pierce in some scrap / offcut... and then go from there (as you can change the cut order in sheetcam).

There have been a lot of ideas why the first pierce is not the best....

The ideas are that
a) that is the first pierce for air to begin flowing (not quite conviced of this, the reason is that the hypertherm torches use blockback start.... so every start is a break in the airflow!.... thats how the arc becomes established).

b) the consumables are cold.... and not up to temperature (probably.... the first pierce / arc will add a bit of heat into the electrode so items probably expand.... may not be very true given the duration is extremely short... how much expansion is going to take place, considering the post flow (after the arc / pierce) will cool it back down.

c) probably a bit of water in the end of the line..... first pierce will clear out anything in the line and also get the flow moving.... so that the regulator will begin moving (diaphragm changing position) .... probably who knows.

Open to any other suggestions.... don't believe anyone has given a definitive "this is it" reason.... but it's common that the first pierce (or dimple) in a sequence is always poor.....

every second one may be the duration between peirces (or dimples....) everything cooling back down and post flow stopping (properly) before the torch refires from scratch and the regulator needs to begin flowing air (move upwards / away from the seat)

Who knows!

Easier to just work around.... pierce in some scrap and then just move around touch off and refire as quick as possible... any delays.... put another pierce in some scrap (say you do some dimples.... then cut something out, then some more dimples etc).

www.plasmaspider.com/viewtopic.php?f=114...first+dimple#p174918


www.plasmaspider.com/viewtopic.php?f=58&...first+dimple#p171110

www.plasmaspider.com/viewtopic.php?f=3&t...first+dimple#p122971

etc.... you get the idea (I just searched "first dimple")

Interesting, I have however never any issues with the first pierce nor the first dimple. When cutting normal I rarely ever have a bad pierce, so this is strange.

I might need to do more research on this and do some different tests to see if I can figure it out. Because the "dimple mode" to create marks where to drill would be just a fantastic thing to have working properly.

Thanks for the suggestions, will read through the threads.

If you or anyone figures this out please let me know, as I've never had the exact cause of the issue. {note, I don't use a refrigerated dryer on my compressed air system, I have used no dryer in the past, and a membrane dryer, and now a dessicant dryer arrangement}

I do have 3-stage air filtration, but when I first started out I just had a single stage course filter (and water trap).

There are so many things that could cause the issue and variances between peoples setups

Some people use a two stage air pressure regulator setup too.... first stage on the reciever, second stage at the compressor. I use a 2 stage... because my dessicant dryer vessels won't like 10bar (150psi) which the receiver will charge up to. maybe the fluctuation in the 2 stage regulator setup can cause a intermittent blip at the initial pierce while both regulators respond to the sudden pressure drop. My first stage regulator is very large.... less pressure drop at higher flow... but it may have an effect on when it leaves the seat

No idea.... too many potentials.... easier just to be aware of it and work around it.

My first thoughts on this are:

* What is being used for the Arc OK signal - voltage or the Arc OK signal from the plasma cutter.
* What are the spotting settings in Qtplasmac. You have 2 separate settings - one is the voltage level at which timing starts, and the 2nd is the torch on time.

Personally I wouldn't want to use voltage as a sensing method to start the torch on time interval. I would just set the voltage at zero so the timing starts from when the system gets the Arc OK signal. However, that is assuming the Arc OK signal comes from the plasma cutter itself and not from voltage sensing. Remember, in Qtplamac you have 2 ways of getting the "Arc On" signal - Mode 0 which uses voltage sensing, and Modes 1 & 2 that use a digital input Arc OK signal.

Jim Colt who was the big Hypertherm rep claimed that the Arc OK signal from Hypertherm machines is very accurate / repeatable. He uses a Plasmacam system which he said does the timing from when the system gets the Arc OK signal and he gets perfect repeatable hole marks, whether post cut air flow is on or not.

Some years ago I designed a closed loop system for a different CNC system for creating dimples. It completely took control of the torch on signal, external to the CNC controller. Plasma current was sensed with an accurate fast acting hall effect based device. Relays were fast acting. Timing was accurate and repeatable to a few milli-seconds, confirmed with logic analyser testing.
My dimples were repeatable EVERY time, and I could adjust to make them cute and tiny, or big and messy, regardless of whether air flow was on, and I marked over 700 of them on one job.

I don't think air quality, nozzle being hot/cold, etc will have anything to do with the problem. I reckon it's all a timing and/or sensing issue. A repeatable / precise method of detecting current, and precise timing of torch off, are the two main factors I believe are what's needed.

For the Arc OK of the plasma cutter you are relying on it's accuracy and repeatability. For the precisely timed duration/turn-off of the torch signal, you are then relying on the capability of the CNC controller. After that you are again relying on the plasma cutter for repeatability in the timing to turn the torch off when the Torch On signal turns off.

That is why I took control away from the CNC system, to eliminate as many variables as possible. Luckily my Powermax 1250 seems to have good repeatability with the time to switch the torch off.

I've never used the dimple marking yet since moving to Qtplasmac so after reading this thread I'll have big round eyes, watching in anticipation, when I do eventually try it LOL.

phillc54 wrote:

The plasmac component does not use Arc OK for spotting.
It was removed a long time back as many users were finding that by the time an ArcOK was received then they were getting too large a crater.
It is time based to a resolution of one thread cycle (in most cases 1mS) plus a voltage threshold for timer start may optionally be used.

 

Thanks for the update Phill. My Linuxcnc user guide must be way out of date LOL.

So could you clarify, if I set the voltage threshold at zero, would timing start from when the Torch On signal is given ?

Thanks Phill, I'll have a read of the latest documentation too.