How many have actually needed high speed THC

For general topics and questions that do not fit into any of the other categories or forums.
beefy
4.5 Star Member
4.5 Star Member
Posts: 1504
Joined: Fri Jan 18, 2013 3:19 am

Re: How many have actually needed high speed THC

Post by beefy »

tcaudle wrote:A Moped will get you to the store just as easy as a Mustang GT500. Maybe not as fast, maybe not as comfortably and perhaps a bit wetter in the rain but if your goal is to just to GET there then why spend the extra bucks?

Just a few observations though.
. You shouldn't be crossing over other kerfs/voids to begin with . That is Pilot Error . Only time you should cross a kerf is at the ending of a closed object as it approaches the beginning. That occurrence can be stopped using a Velocity Anti-Dive because the trajectory is going to slow down as it decelerates at the end,
.
I don't know a single Engineer or Programmer that does not think they can do a better job on a design or software than any other person. I think its part of the personality of those trades.
The moped / GT500 comparison is implying speed difference, which is a moot point if a slower THC can keep up with slower cutting conditions. It's only valid when a THC cannot keep up, as the cut speed and voltage rate of change becomes too fast.

Regarding crossing kerfs / voids. Operator error or not, nearly every decent plasma system has this facility built in, obviously it seems to be a required feature. As you said a great many cuts are closed objects and these have lead ins/outs, and therefore a kerf is crossed every time with these. So kerf crossing is a very common and basic part of plasma cutting and is therefore most often NOT Pilot error. You could also get someone using a piece of metal that has a small hole drilled in it, or whatever, and the torch path crosses this. But at the end of the day a top end high performance system should have it included (void lock) as a basic function.

Sorry Tom this bit has completely gone over my head:
That occurrence can be stopped using a Velocity Anti-Dive because the trajectory is going to slow down as it decelerates at the end.

Velocity anti-dive is normally set as a percentage of feedrate, and automatically turns off the THC when the feedrate drops below that percentage speed. I tend to term this "Corner Lock" as opposed to the very different "Void Lock" (kerf crossing). Due to deceleration when changing direction for instance the torch can be forced to slow down, and so the THC gets temporarily disabled until the torch accelerates out of the corner and regains a speed above the velocity anti-dive setting. So not sure what you are saying in the above statement and how it's applied to kerf crossing.

And not sure where you are going with Engineers/Programmers doing a better job on a design. I'm only trying to get some facts sorted so users are not led to believe they need a super fast high performance THC when they are doing jobs that simply do not require it. How good or bad any of us are with design has little relevance to what we are talking about.

I started this thread because I was wondering how far a basic THC UP/DOWN system can get me before I need a higher end PID controlled THC system. Now it seems to have turned into warnings of doom and gloom (if you don't have the latest top end THC) for cut quality and consumable life even when cutting at very slow speeds.
2500 x 1500 water table
Powermax 1250 & Duramax torch (because of the new $$$$ync system, will buy Thermal Dynamics next)
LinuxCNC
Sheetcam
Alibre Design 3D solid modelling
Coreldraw 2019
BTA Plasma
3.5 Star Elite Contributing Member
3.5 Star Elite Contributing Member
Posts: 590
Joined: Thu May 12, 2011 4:28 pm

Re: How many have actually needed high speed THC

Post by BTA Plasma »

Doom and gloom :lol: Please don't look at it that way. It isn't worth discussing if the information shared stresses you out.
tcaudle
4 Star Elite Contributing Member
4 Star Elite Contributing Member
Posts: 1353
Joined: Tue Feb 28, 2012 6:47 pm

Re: How many have actually needed high speed THC

Post by tcaudle »

Methinks He just wants validation of his theory that he does not need any more performance than a basic 10 year old THC design can offer and its wasted effort and just a way for vendors to make more money. You would think after years of talking about doing his own design that something would emerge that reinforces the Keep=it-Simple Concept.

So my response is : If you are happy with the speed and performance of your THC or of one based on the UP- DOWN feedback to the software, then by all means continue to use it. Thousands of the lower performance THCs have been (and are still being) sold over the last 12 years. I did the very first design of a simple low cost THC (later released as the THC300) and worked with Art at MACH to put in the THC logic into MACH and did all of the testing. That was way back at MACH1. At the time the parallel port was the main I/O and was pretty fast and gave acceptable response to THC feedback. The DTHC (used in the MP3000 and later upgraded to the DYHCII) was the first model to move all of the control and settings into the control program while later models expanded that with toolpath based dynamic settings, Stored presets and a knobless user interface to MACH3. Later there were full automation options cutting process and for the Hypertherm RS485 Cut current control. All really useless bangles if you just wanted to cut out simple shapes on flat material .

So here it is: If you are not doing any type of aggressive cutting at feedrates below 250 IPM and you can keep the material from popping up rapidly from stress relief and heat expansion, then you don't need a high performance THC. Several vendors of tables told customers they did not need a THC at all as a way to keep the initial cost down. That was true if all you cut was thicker material and were careful to make the metal level. Much like the infamous 2 x 2 plasma table, it was effective in a very narrow market.

I have no objections for someone to build or bay a system that is "good enough" for their needs. When we do designs on THC and motion controllers we look up-stream at the higher end commercial THC's and their specs and features . There is a reason they have evolved to those levels.
Rodw
4 Star Member
4 Star Member
Posts: 780
Joined: Sun Aug 21, 2016 1:49 am
Location: Brisbane, Australia
Contact:

Re: How many have actually needed high speed THC

Post by Rodw »

beefy wrote: Sorry Tom this bit has completely gone over my head:
That occurrence can be stopped using a Velocity Anti-Dive because the trajectory is going to slow down as it decelerates at the end.

Velocity anti-dive is normally set as a percentage of feedrate, and automatically turns off the THC when the feedrate drops below that percentage speed. I tend to term this "Corner Lock" as opposed to the very different "Void Lock" (kerf crossing). Due to deceleration when changing direction for instance the torch can be forced to slow down, and so the THC gets temporarily disabled until the torch accelerates out of the corner and regains a speed above the velocity anti-dive setting. So not sure what you are saying in the above statement and how it's applied to kerf crossing.
I think I have to side with Tom here. You don't cut a part out and encounter an existing void or cross an existing kerf. 40 years ago at college, I was quickly taught how to use an oxy and then I had to cut a 1" hole out of heavy plate for a tractor drawbar dynanometer. I still remember how surprised I was when the middle dropped out and I lost my compass point reference. I also remember how annoyed I was about the divot the torch left as it free falled down. Thats when you will encounter a void.

Having programmed a plasma Anti Dive/ corner lock that monitors XY speed, I see the GUI LED flick on and off quite regularly and I must admit I've forgotten about it as its just there in the background. But as Tom has said, any XY move (including an inside cut) will decelerate to a stop at the acceleration rate set in your software so Anti dive will kick in. This prevents those embarrassing divots I experienced with a hand held torch all those years ago. Its suddenly dawned on me, that anti dive is a much better term than corner lock for this feature. I'll have to update my GUI.

I suspect that if your Neuron has this feature, you won't need to worry about turning it on and off from within gcode.
beefy wrote: I started this thread because I was wondering how far a basic THC UP/DOWN system can get me before I need a higher end PID controlled THC system. Now it seems to have turned into warnings of doom and gloom (if you don't have the latest top end THC) for cut quality and consumable life even when cutting at very slow speeds.
Beefy, don't sell yourself short here. When t (time period) is held constant (which is what happens when we use a timer interrupt to sample our feedback signal (torch voltage), the PID formulas reduce to just a few lines of C code. The magic is in selecting the P, I, and D tuning parameters, not the mathematical algorithm. Thats where simulations and trial and error comes into play. But yes if all you are doing is cutting 16mm plate for an excavator bucket, the "bit bang" approach will get the job done.

I will say here that Jim is right based on my limited experience, Z acceleration settings is a pretty critical tuning parameter. I'd also say that a stepper motor can easily give you the velocity and acceleration required. I suspect that for the reasons mentioned by Jim that the 6600 mm/min and 2000 mm/sec/sec I can get out of my stepper is too fast! A mathematician I know did some calculations about the acceleration required to cut corrugated iron.
I was thinking about machine Z axis requirements for thin corrugated material.
Example:
1) thin material, mfg says use 240in/min = 4in/sec
2) sinusoidal height (Z) pattern of corrugation:
period: 2 in (distance between peaks)
height: 1 in (peak-to-peak, so Zmax=0.5)
3) time between peaks:
2 in/(4 in/sec) = 0.5 sec
4) frequency: f = 1/0.5sec = 2 Hz
5) Zamplitude:
Z = Zmax*sin(2*pi*f*t)
= 0.5*sin(2*pi*2*t) = 0.5 * sin(12.6*t)
6) Zvelocity = V = d/dt(Z)
V = -Zmax * 12.6 * cos(12.6*t)
Vmax = 0.5 * 12.6 = 6.3 in/sec = 378 in/min
= 160 mm/sec = 9.6 m/min
7) Zaccel = A = d/dt(V)
A = -Zmax * 159 * sin(12.6*t)
Amax = 0.5 * 159 = 79.5 in/sec^2 = 2019 mm/sec^2 = 0.2 g

So folks aren't kidding when suggesting that high performance
is a requirement for Zaxis motion when cutting thin materials
Pretty sure a basic NEMA 23 is up to the task and got it up to 2500 mm/sec/sec on a 50mm distance without loosing steps but I need more real world testing to confirm short move segments a THC is exposed to doesn't loose steps.
beefy
4.5 Star Member
4.5 Star Member
Posts: 1504
Joined: Fri Jan 18, 2013 3:19 am

Re: How many have actually needed high speed THC

Post by beefy »

tcaudle wrote:Methinks He just wants validation of his theory that he does not need any more performance than a basic 10 year old THC design can offer and its wasted effort and just a way for vendors to make more money. You would think after years of talking about doing his own design that something would emerge that reinforces the Keep=it-Simple Concept.

So my response is : If you are happy with the speed and performance of your THC or of one based on the UP- DOWN feedback to the software, then by all means continue to use it. Thousands of the lower performance THCs have been (and are still being) sold over the last 12 years. I did the very first design of a simple low cost THC (later released as the THC300) and worked with Art at MACH to put in the THC logic into MACH and did all of the testing. That was way back at MACH1. At the time the parallel port was the main I/O and was pretty fast and gave acceptable response to THC feedback. The DTHC (used in the MP3000 and later upgraded to the DYHCII) was the first model to move all of the control and settings into the control program while later models expanded that with toolpath based dynamic settings, Stored presets and a knobless user interface to MACH3. Later there were full automation options cutting process and for the Hypertherm RS485 Cut current control. All really useless bangles if you just wanted to cut out simple shapes on flat material .

So here it is: If you are not doing any type of aggressive cutting at feedrates below 250 IPM and you can keep the material from popping up rapidly from stress relief and heat expansion, then you don't need a high performance THC. Several vendors of tables told customers they did not need a THC at all as a way to keep the initial cost down. That was true if all you cut was thicker material and were careful to make the metal level. Much like the infamous 2 x 2 plasma table, it was effective in a very narrow market.

I have no objections for someone to build or bay a system that is "good enough" for their needs. When we do designs on THC and motion controllers we look up-stream at the higher end commercial THC's and their specs and features . There is a reason they have evolved to those levels.
Tom,

regarding your first paragraph. Thanks once again for trying to publicly ridicule me (it's not the first time you've attempted it). You are welcome to think what you want but my opening question was pretty simple. Instead of getting mainly simple answers to my question, I get what I consider to be some bullcrap shouting out that people NEED a fast modern THC regardless of what they are cutting.

I'm not the only one who's came back at you for jumping in threads protecting your sales.

Now for the rest of what you said, thank you for saying it plain and simple. At least that does seem to throw some crap on the BS that BTA said about needing a super fast THC even for slow cutting, otherwise your torch will be A FEW volts out and you'll get short consumable life and cut quality.
2500 x 1500 water table
Powermax 1250 & Duramax torch (because of the new $$$$ync system, will buy Thermal Dynamics next)
LinuxCNC
Sheetcam
Alibre Design 3D solid modelling
Coreldraw 2019
beefy
4.5 Star Member
4.5 Star Member
Posts: 1504
Joined: Fri Jan 18, 2013 3:19 am

Re: How many have actually needed high speed THC

Post by beefy »

BTA Plasma wrote:Doom and gloom :lol: Please don't look at it that way. It isn't worth discussing if the information shared stresses you out.
If I looked at what you said as "information" I'd view it differently. Unfortunately I view what you said as misleading BS, sorry :lol:
2500 x 1500 water table
Powermax 1250 & Duramax torch (because of the new $$$$ync system, will buy Thermal Dynamics next)
LinuxCNC
Sheetcam
Alibre Design 3D solid modelling
Coreldraw 2019
Post Reply

Return to “CNC Plasma Cutters General Forum”