UCcnc Plasma features

[vmax549]

Robert DIVOTS are either an advanced Science Degree OR a combined work of black magic and magic monkey dust. There ARE tricks to the trade (;-)

BUT I think that you will find it is a LOT easier to deal with them in UCcnc and its EXACT timing of events . NOT like mach3 where it was a best guess method.

ALSO KEITH has some tricks up his sleeve in his NEW THC card that may deal with them as well (;-)

I would reach over and give him a head slap him to get him to hurry up BUT he is on teh almost exact opposite of teh planet. I can't quite reach him from here.

(;-) TP

[Beefy]

Hi Robert,

the way an oxy acetylene cutting torch works is the small outer circle of "spikey" flames heats the metal until the surface is just starting to melt or very close to it (something like that). Then the big hole in the middle of the nozzle is the oxygen hole and when the oxy starts coming out the metal is literally on fire. It's nothing more than heat and a continuous supply of oxygen that causes the metal to rapidly burn (not melt) away. So likewise air has roughly 20% oxygen and you get the same effect when the plasma torch is turned off but to a much lesser degree and of course it's not enough to keep itself going so the metal stays "alight" for just a short time after the torch power is removed. Bit of a sidenote but when I was in industry we were always told about how oil could spontaneously ignite in the presence of pure oxygen. I could never figure that one out because the triangle of fire requires fuel, oxygen, heat (where's the heat). But that's what we were always told.

My suggestion of a smaller relay was more for consistency of time to de-energise. For example if say a big dopey relay and a small fast relay had variation times of 20% then 20% of 3mS is a lot less than 20% of 15 mS. Funny thing about the relays, I just bought two 5v reed relays yesterday. When I get around to it I'm going to hook them up and test the deactivation time. They should be even faster than the PCB relay I use on my hole centre marking PCB. If you use one make sure both the current AND power handling capability of the contacts is more than enough. You also want longevity so the relay doesn't want to be stressed to its' max.

Don't know if you haunt Plasmaspider, but the divot issue has came up over there numerous times and a few have got rid of them with synchronously switching the torch off a short distance before the end of the cut path. The good thing about UCcnc with it's excellent M10/11 (laser output) and excellent M205/206 (THC ON/OFF) you can even disable THC just before you give an M11 to turn of the torch so loss of ARC OK will not halt motion in the last mm or two.

Also I tested the repaired version of the UCcnc install yesterday and they've fixed the glitch in the step pulse stream where you put an M205/6 and a change in feedrate at the same place in gcode.

Balazs also got back to me and said they are going to implement the changes I suggested so you can adjust THC Correction speed and Corner Anti-Dive settings while cutting. That is the ultimate method for fine tuning things in my eyes.

Keith

[vmax549]

OK I'll BITE why would I NEED to change the THC correction speed while cutting ?? Tuning in a system I set that values as HIGH as it can go without stalling the motors (stepper). That is as QUICK as it can ever go. WHY would i need to slow that down ??

Same with Corner dive But I can almost see where that MIGHT??? be useful. But once dialed in here I have never changed it (;-) Most times you ONLY lock out on corner dives is for very sharp corners and in very sharp corners you are NOT there long enough to enact a change.

Please SPlain it to me.

(;-) TP

You SHOULD already be able to do both of those functions (;-) for testing .

(;-) TP

(;-) TP

[Beefy]

Ha ha, I knew you’d be replying to this one Terry. Just had a day out with the family and knew I’d be responding to you when I got home LOL. We’ve had our different ways of thinking on this one in the past. I actually did explain one of these reasons in our emails some time ago by the way.

OK, first of all, it can only be a good thing to be able to adjust these during real cutting. Even if that was just for initial tuning, that beats the hell out of adjusting the setting offline, then doing a test run afterwards. And then if it wasn’t quite how you liked it, stop everything, adjust again, then do another test cut. That gets back to the radio example I mentioned earlier, you tune a radio station with the radio ON. These are settings that you can set at varying levels so why do it the hard way when there is an easier, faster, and more accurate way.

I used to have to tune the PID speed governor reactions on 1 MW gensets. We did that while the set was running because you can see immediately how it reacts when you adjust something. I’m applying that same thinking to tuning THC reaction and Corner Anti-Dive reaction. You can do test cuts and adjust these settings until you are happy with the reaction. You also get to see exactly at what point things go wrong and you can back of as little or as much from that point.

A beginner might get the settings wrong but then he has an option of adjusting it during the cut until he sees things “settle down” (jackhammer explanation coming), rather than stop in the middle of a cut, adjust things then have to do a mid cut restart and hope he got it right. I just can’t see anything bad about having the OPTION to adjust these settings in real time. May not be absolutely necessity but I see it as a creating a much more user friendly plasma system. I want people to flock to UCcnc because it looks like it's going to be my new plasma software/system so the more user friendly it can be the better.

THC CORRECTION SPEED & WHY IT MAY BE BAD TO BE TOO FAST

1. Reason #1 – THE JACKHAMMER
So why don’t I just set it to warp speed as you say ??
Well, many have had the problem of the jackhammer Z action when under THC control. That is caused by either one OR a combination of two things, either TOO FAST correction speed or TOO SMALL voltage dead band.
But basically the jackhammering / oscillation or whatever others call it, is caused when the Z responds to an out of range voltage condition and moves in the correcting direction. It gets to the deadband voltage point but it cannot stop in time to stay within that dead band area. Thus it shoots through the dead band and goes outside of the other side of the dead band. Then the process repeats but in the other direction and hence the jackhammering and the wavy / scalloped cut edge. Many have had this problem when Z is too fast. I am one of them.
This is corrected by either REDUCING the THC correction speed so it can stop within the voltage dead band area, OR increasing the span of the voltage dead band. The latter basically means you are sacrificing cut height accuracy simply so your Z can go faster, but many cuts don’t even need fast THC correction speed. I will even explain soon why on thicker steel you may need to REDUCE correction speed from your normal setting.
Mach3 did not have acceleration / deceleration on its THC moves but UCcnc DOES !!! This overshoot / jackhammering action is a common problem on Mach3 systems despite the Z supposedly stopping instantly (no accel/decel). Now seeing as UCcnc has accel/decel applied to THC moves the axis can NOT stop instantly when the THC UP or THC DOWN signal is removed. You’ll see from the logic analyser results I sent that after a THC UP or THC DOWN signal is removed the Z step pulses do not immediately stop, but decelerate to a stop. I am thus forseeing that THC moves in UCcnc MAY need to have accel/decel removed for this very reason but as yet I’m not sure.
This is another thing I want to test when I have my table connected up to the UCcnc system. I want to see if it is better or worse to have accel/decel applied to THC moves. If it turns out to be worse due to inability to stop within the voltage dead band, I’m sure Balazs will say it’s an easy fix to do THC moves without it. I bet it’s actually easier for them to not have it in the code.
Yes, accel/decel will allow faster Z moves but that’s not much use if the faster Z moves need time to decelerate to a stop and thus fly through the dead band. Maybe that is why Art decided NOT to have accel/decel applied to THC moves.

2. Reason #2 – DIVING IN VOIDS
I pondered this one for some time before a well known manufacturer of plasma systems did a write up on it and confirmed my thoughts. It applies at much lower cutting speeds. One type of anti-dive is VOID anti-dive, very different from CORNER anti-dive which is related to a reduction in cut speed. Void anti-dive is when the torch crosses over a void and the voltage rises to a level way above normal. Your anti-dive % setting freezes THC down movements when this abnormally high voltage level is reached and once the torch has crossed the void and is cutting metal again the voltage drops back down to where it was. But this is where the problem CAN happen when you have very fast THC correction speed when at a slow cut speed. The fast THC correction speed manages to react to the voltage increase and the Z moves down fast enough to PREVENT the voltage from ever reaching the void anti-dive setting. Thus the torch dives at the void area and crashes into the metal. It was explained that under these circumstances it is beneficial to reduce the THC correction speed, and that in any case when cutting at very slow feedrates there is no real benefit to super fast THC correction speed

So there you have two reasons why you don’t necessarily always want the fastest THC correction speed. You may only want enough to do the job well.

Not every plasma table user is very experienced and knows what do in advance. I see it as a great way to teach them how to use their table by getting them to do tests and watch what happens as they change settings while doing a test cut. I wish I’d had that facility when I was learning and someone had explained it to me. IF (IF IF LOL) I get my THC system up and running I want to teach beginners about the whole process and instruct them to do these test cuts and watch what happens when they adjust this parameter or that parameter during a cut. That’s another small reason I’d like this facility. I see so many plasma newbies learning about all this only when they have a problem and come begging for help on forums.
I myself got rid of my jackhammering problem by lowering the THC correction speed but the thing is I don’t know exactly how much faster I could have it. I just set it lower and then the problem went away. I would like it if I could do one simple test cut and find out very quickly where my jackhammering starts, then just back it off by a safety margin. That surely beats multiple test runs to find the highest speed before jackhammering starts.

Keith.

[robertspark]

Keith, thanks for your explanation, I know what Terry's point is as he educated me a couple of months back on the m3 forum.

Personally I'd like a bit of accel and deceleration. With thc, but by a bit I mean one or two steps (a small known multiple number), which after all are micro steps, just to aide in preventing stall by the tiniest amount, hence for me I'd like to know or programme my steps (units) and not units/s/s (although via calculation they are the same)

The other function which is thc speed relative to feedrate is easy to employ via macro, and I can see the point in that if you were cutting over kerfs or doing art type work.... the problem is, art work is not done in the thicker stuff at slower feedrate (is it most of the time).... hence I can see benefits, but not much occurrence of utilisation, plus it's easy to employ by macropump, as the set / target/ gcode feedrate only changes for g1, g2 and g3 moves.

You blokes have cut miles more metal than I, but that is just my pondering on the ideas.

Rob

[Beefy]

Hi Robert,

Balazs has explained to me that they did consider the matter of acceleration / deceleration applied to THC moves and to put it in a nutshell, he said it is for good reasons.

Can't wait for when I get my UCcnc system hooked up to the table.

Keith.

[vmax549]

HIYA Keith I have teh email and read it again BUT no one else got to read it. (;-) Now they have.

Teh jack hammer effect has only a little to do with the deadband. It is more that the stepper is outrunning the THC's ability(speed) to correct the voltage and it can overshoot before teh THC can react. Also remember the tthe up/down control is NOT proportional it is all or none so really a PID would not do much good.

I think that we will find that an accelleration for Up/down is not a good thing for THC control with steppers. IF teh Z moved a lot more than it does it would be great BUT teh Z most times only moves a very short distance and there is NO TIME for accel and deaccel because the event is really over before it gets started.
THAT was the main reason for Art/Tom to NOT use accel in developing Mach3 thc controls, There is NOT enough time . NOW if you are using servos with proportional gain then that is another story.

There IS a new THC that is coming out that TRIES to do proportional gain on teh existing Mach3 type THC function but I cannot see that it will work with steppers as teh needed UP/DOWN signals are not proportional it is like a light switch on/off no dimming.

The earlier Industry problem with Correction Speed was with teh EARLY analog systems that were SLOW to begin with . A digital control is PLENTY fast enough to freeze DOWN on a gap crossing negating any need or advantage to change teh speed. AND plenty fast to make decisions on when to restart down control . The old analogs could not react that fast. SO yes there were problem on fast correction speeds on slow gap crossings. The thc could not reacy fast enough to freeze down. Not so with digital.

That brings up you radio example a digtal tunner you cannot TUNE it is either on freq or NOT for the most part. An analog tuner yes you tune in by hand, slow process

Corner Dive is strictly a function of teh controller as it KNOWs what teh Velocity is at all times and can simply block DOWN. IF you start juggling correction speed THAT can effect corner
dive corrections and can effect THC reaction times and then around and around teh merry go round we go (;-).

Teh newest THC to come out attempts to do Corner dive AT the THC based on a velocity signal from teh Controller but WHY ??? Just to say you did??

Just some thoughts,

(;-) TP

NOTE: You do remember that you can already change those settings in UCcnc on teh fly for testing . We figured that out working with Setvolts adjustments.

[OlfCNC]

I found the THC control with accelerations much better than without it. I had my Z motor stall a few times with mach when it had to correct the THC and later i learnt that it was because mach does not apply any acceleration, it jumps the Z motor which is not good as my Z axis is not as light weight as much as it probably should be. With uccnc it never stall and still corrects fine on heavily bent thin sheets.
If this was a poll then I give one to have acceleration. ;-)

[vmax549]

NO Poll I do think it is great that you can set the rates as HIGH or low as you want. But it would not be a surprise to find it makes little to no difference with the average hobby machine. THC moves are micro compared to milling/router.

(;-) TP

[Beefy]

Hi Terry,

Teh jack hammer effect has only a little to do with the deadband. It is more that the stepper is outrunning the THC's ability(speed) to correct the voltage and it can overshoot before teh THC can react. Also remember the tthe up/down control is NOT proportional it is all or none so really a PID would not do much good.
Yes but the deadband gives a larger window thereby allowing a THC with less than super fast reaction speed. Whatever the cause, the end result is the same, and the solution is either increasing the span/window of the deadband or decreasing the THC correction speed, or a mixture of the two. Another thing we haven't mentioned is inertia and lack of gantry rigidity. Even if the THC is lightning fast with its' reaction time, and IF the stepper could stop immediately, if the rigidity of the Z axis and gantry cannot take the sudden stop of the torch (think water hammer in pipes when you turn the tap off), it can allow the torch to pass through the voltage dead band and that's another cause of jackhammering. Whatever the cause the fixes available are increasing the dead band and/or reducing THC correction speed.
I understand fully the difference between a basic THC UP/DOWN style system and a PID system. Have done for a long time.

That brings up you radio example a digtal tunner you cannot TUNE it is either on freq or NOT for the most part. An analog tuner yes you tune in by hand, slow process
Well I was referring to an analogue one, not the digital ones that automatically lock in to a frequency as they find it. The process is a lot faster doing it in real time (radio turned on) than turning it off, changing the tuning dial, then turning the radio on to see if it tuned in well. That's why I used that as a comparison.

Corner Dive is strictly a function of teh controller as it KNOWs what teh Velocity is at all times and can simply block DOWN. IF you start juggling correction speed THAT can effect corner
dive corrections and can effect THC reaction times and then around and around teh merry go round we go (;-).
I don't see how that can happen.
The system monitors the XY motion feedrate, NOT the THC Correction Speed. If your Corner Lock (Corner Anti-Dive) was set at 80% and your XY feedrate was set at say 1000 mm/min, then when the system finds the XY feedrate has dropped to 800 mm/min it simply ignores THC DOWN signals to prevent torch diving. Changing THC Correction Speed should have no affect on the reaction of Corner Anti-Dive at all. The THC Correction Speed is simply how fast the Z moves when it gets either a THC UP or a THC DOWN signal. The operation of Corner Anti-Dive is completely independent of this. THC reactions times are a functions of 2 things, 1). the Torch Height Control itself, i.e. the unit which actually measures the voltage and sends an UP or DOWN signal to the controller (e.g. UCcnc) and 2). the the CNC controller (UCcnc). From this point the reaction is determined by how long it takes from when the controller gets the THC up/down signal, to when it sends out Z step pulses. 1). & 2). are completely independent from Corner Anti-Dive operation.

NOTE: You do remember that you can already change those settings in UCcnc on teh fly for testing . We figured that out working with Setvolts adjustments.
Yes but Balazs told me you can't do that with Corner Anti-Dive. He said the only way is to adjust the DRO and click the Apply button. He then emailed me later and said him and Balazs#2 where looking into changing it so that it could be changed during a cut. He is aware that I know how to update DROs via either a basic "updown" DRO, or a macroloop monitoring user buttons, or the same with a plugin.

OlfCNC,

I found the THC control with accelerations much better than without it.
I've been pondering this one. I'm wondering if part of the reason could be that UCcnc has much faster reaction times than Mach3, i.e. after it gets the THC UP or DOWN signal. During testing I consistently got 3-4 mS from when a THC up/down signal arrived at UCcnc to when the first Z step pulse (movement) started. Unfortunately I have no comparison tests with Mach3. Plus if the Z has quite a high acceleration rate setting that is also helping it get "closer" to instant moves and halting. Yes with Mach3 you are forced to keep your correction speed quite low to avoid stepper stalling. Mine is at a whopping 15% LOL.

Keith.

[vmax549]

hIY AKeith I tried teh Correction speed trick via Plugin and it works just fine to adjust it (;-) At least I see teh numbers change. I guess they never really tried doing it?

Machines that tned to loos steps on Z are normally under geared on the Z feed and the lost steps tend to be from over speeding the motor where the torque band is weak.

IF you can POGO in Z that is as FAST as you can go for THC control of Z as you are already outrunning the THC responce.

I think the deadband width just delays WHEN the pogo will occur .

ALL in all I think teh UCcnc THC is great and teh work KEITH and Balazs 1/2 has done is great news for all of US. and a few MORE adjustments surely will not hurt .

AND with some of the new features of Keiths THC board you will really LOVE it. Can you say really controllable hole marking capability ?? How about being able to create a dot matrix marking function for writing on metal dot matrix style.

(;-) TP

[vmax549]

"" Corner Dive is strictly a function of teh controller as it KNOWs what teh Velocity is at all times and can simply block DOWN. IF you start juggling correction speed THAT can effect corner
dive corrections and can effect THC reaction times and then around and around teh merry go round we go (;-).""

YEP I MAY have spoke to quick on that one before I tested the theory (;-)

(;-) TP

[rooies13]

Good day Vmax

Did you come right with this or not?

IF there is anyone out there that understands UCcnc plugins I could use some help implimenting a Modbus Plugin for plasma control of the THC (Modbus RTU) and Hypertherm torch
( Modbus ACSII). I have a Modbus library but don't quite know what to do with it (;-)

(;-) TP

[vmax549]

I have been able to talk/read a modbus device in a VERY primitive way. Not good enough to do anything with. I am afraid the plugin is WAY ABOVE my understanding of programing. We will have to wait until UCcnc decides it is time to create teh Modbus plugin. They will know how to properly integrate it into UCCNC.

Sorry(;-) TP

[rooies13]

Okay, i also have no knowledge of plugin programming, so i`m of no help either.
No need to say sorry, i was just curious if you got it to work. Is there a UCcnc person on this forum we can talk to?

Dennis