Shop OBEX P1 Docs P2 Docs Learn Events
New User - Industrial Control Use Question? — Parallax Forums

New User - Industrial Control Use Question?

88incher88incher Posts: 14
edited 2007-05-28 20:20 in BASIC Stamp
Hi there,

We are looking for a method of controlling the torch height on a CNC Plasma Arc Cutting Machine, the Common way to do this is to measure the arc voltage (between cutting head and the job)·and use a motorized (stepper or servo) slide to move the torch towards or further away from the job proportionaly to the arc voltage.

Typlicaly the arc voltages run at around 300vdc at open circuit and anything from 200 - 100vdc when the cut is started, so presumably some sort of voltage devision device is needed.

Would the Basic Stamp be suitable for such an operation? (i.e voltage sensing and motor control?)

Any suggestions or help would be much appreciated.

All the best,
James
«1

Comments

  • Mike GreenMike Green Posts: 23,101
    edited 2007-05-23 22:39
    A Basic Stamp would be fine for the job. If you use a servo motor, the Stamp can control it directly. If you need to use a stepper, you will need some kind of motor driver. Parallax sells several that work with the Stamp.

    Measuring the arc voltage is a little tricky. You can use a precision voltage divider to reduce the voltage to a more reasonable level for measurement (say 0-5V). You probably will want to isolate the arc voltage from the rest of the circuit because of the transients and the high voltage. You'll probably want a small isolated DC to DC converter to produce 5V for an analog to digital converter (like the ADC0831 - see Parallax's store) and some optoisolators to talk to the converter.
  • 88incher88incher Posts: 14
    edited 2007-05-23 22:51
    Hello Mike,

    Wow, that is possibly the fastest reply I have ever encountered on a forum, many thanks for the info. I had guessed that these stamps would be suitable, just wasn't sure how to impliment the system. Unfortunatly my engineering background is very much more mechanical then electronic, so we have a great plasma profiling table but it's going no where fast with·out this torch height control.

    Next question, as we are based across the pound (UK), who are the·best people to approach for Parallax products?

    Many thanks again,
    James·
  • 88incher88incher Posts: 14
    edited 2007-05-23 23:07
    SO so sorry, I've just thought a little harder about this, and realised that we use··High Frequesncey Arc Initiation (HF Start) which will kill most things electronic placed across the output and ground. Maybe another solution would be to use a proximity sensor, is that feasible?


    James
  • skylightskylight Posts: 1,915
    edited 2007-05-23 23:22
    I realise you are wanting to measure voltage of the arc here but is it possible to go about it in a different way and measure temperature at or close to the arc point on the bed (could be done with optical thermometer)? I presume the torch moving up and down is to control the arc and therefore cutting temperature, is this right?
  • 88incher88incher Posts: 14
    edited 2007-05-23 23:25
    Someone just suggested the Ping))) module, but I think I may have deleted their post by mistake, so sorry. Unfortunatly due to the harsh envoronment that the sensor will be placed in I don't think the ping will do the job, also the suggested gap between cutting nozzle and job is something in the order of .040" to 0.120"

    Again many thanks, and any more suggestions - Great!



    James
  • 88incher88incher Posts: 14
    edited 2007-05-23 23:29
    Hi there,

    No sorry, I think I have been mis-understood, measuring the arc voltage is a way to determin the arc length (which is what we are trying to control) the actual voltage is not inportant to us, we we are just using that as an idea to gauge the arc length (or distance between the cutting tip and the work) which is the important factor.

    Basicly when sheet or plate is loaded onto the machine it is never perfectly flat which means the torch needs to raise and fall to keep a fixed distance away from said sheet or plate.

    James
  • skylightskylight Posts: 1,915
    edited 2007-05-23 23:46
    then an insulated rod touching the work surface moving up and down vertically controlling a pot with basic stamp reading the value using RCtime calculating distance and operating a servo should be conceivable.
    Another non contact method could be measuring light bouncing back off the surface though this would need calibrating when different materials were used due to reflectivity of said material. Could a laser pointing to a guage on a wall somewhere be used as the sensor?

    Post Edited (skylight) : 5/23/2007 11:54:34 PM GMT
  • agfaagfa Posts: 295
    edited 2007-05-23 23:48
    Just out of curriosity. What type of system control are you using?· PLC?· What determines the initial torch height, is it just the gauge of the plate then the height is never adjusted for that piece?· You mention a common way to measure this.· Are you not using it?· and if so, can't the·programming on the system be adjusted to change at predetermined intervals?
  • 88incher88incher Posts: 14
    edited 2007-05-24 16:43
    Hi all,

    It must be a non contact method, as bits of steel / metal slag get blown all over the place in this process and a feeler could get stuck or catch on the surface, also as small parts are cut they have a tendancy to tip up, again possibly causing the feeler to crash.

    The gantry table it's self is controled using a closed loop stepper system driven by dedicated CNC software on a PC, this bit is no problem and completely sorted now. It's just the torch height that's bothering me. The course torch height will be controlled by the PC system just as the other two axis will be but the fine dynamic height control needs to be stand alone in reality.



    In theory it should work something like this.....

    1) PC moves the torch Z axis (height) to a safe working distance above the job (say 1" above) (at this point the torch height control should be in it's home position and inactive)

    2) PC·moves the torch to the first cut position using the X,Y axis.

    3) PC moves the torch down to a set height above the job (ready to strike the arc, approx .300" above, this is not a critical distance so can be done with out sensing)

    4) In the CNC program an·M Code (CNC Machine code) is processed by the pc CNC software to turn the torch on (strike the arc) (M code 03 typicaly will do this by causing a relay to close)

    5) At this point the torch height control will know the torch has been started (by the closing of the relay) there will be a short dwell (Typicaly 0.5sec) while the arc is established and then the torch height control will kick in, moving down automaticaly (using a seperate Z axis slide of it's own) to a pre determind height and maintain this height (by mesuring from the torch head to the job)

    6) PC moves the torch through its lines of cut (using X,Y axis) all the time the torch height control is keeping the torch at the set height.

    7) PC reads a M05 code in the CNC program (turn torch off) and the torch height control moves back to it's home position.

    There we go, bit of long post but thinks it gets across what I'm trying to achieve?

    So any more suggestions greatfully received.

    James
  • skylightskylight Posts: 1,915
    edited 2007-05-24 20:24
    it sounds like you may need some form of surface mapping taking place before the work is done, couldnt a probe of some sort trace the cutting route first and store the results to the z axis movement of the CNC program then the cutting head should raise and lower when needed to? how that would be achieved i'm not sure as im no expert on programming in that language but someone on here may have the answer?
  • 88incher88incher Posts: 14
    edited 2007-05-24 20:52
    Hi Skylight,

    Intresting suggestion, unfortunatly the other reason for torch height control is that as the material is cut it tends to warp due to heat and this also causes torch height variation. Still nice idea.



    James
  • skylightskylight Posts: 1,915
    edited 2007-05-25 07:36
    is the material in thin sheet form? if so could some form of vacuum underneath keep the material flat this would also aid in keeping the material still.

    Or perhaps some form of rapid cooling after/before the cutting torch that minimises warping?
  • 88incher88incher Posts: 14
    edited 2007-05-25 09:20
    The sheet/plate that is cut could be anywhere between 1mm and 19mm (approx 0.40" and .750")
  • WingtealWingteal Posts: 2
    edited 2007-05-25 23:29
    88, I have a small plasma cutter and am a novice Stamp user. I understand the need to optimize the arc length over an imperfect work surface. It sounds like your cnc cutter moves the head and not the work. Either way, why not employ a simple mechanical strategy and let the plasma head 'ride' on the surface of the work? You could construct a carriage that rides on ceramic slides or wheels. The initial arc culd be struck at .300" on the Z axis and then after a delay, the head could be released to ride within a selected carriage height. Spring suspension could be added. Beyond the plasma headheight issue, I can imagine that some type of arc anaysis, perhaps in conjunction with the voltage output, would be useful for cut optimization and energy savings specific to particular materials being cut. I believe that there are generic sensors that can gather data on the optic wavelengths and heat generated by the cutting arc. Surely this has been accomplished by the major manufactures and is searchable. It is an interesting project and I hope you report on your progress.
  • bennettdanbennettdan Posts: 614
    edited 2007-05-26 01:08
    88incher,
    Have you used a current meter and seen if the current is about the same when the cutting head has the propper arc? if the machine uses a steady current as it is cutting then you might be able to read the current with a shunt resistor and then have the stamp make the changes to the height of the plasma cutter head with a stepper motor or servo.
  • Kevin WoodKevin Wood Posts: 1,266
    edited 2007-05-26 06:09
    Everything that you are trying to do should already be handled by your PC based CNC software. If it's not, that might be your problem right there.

    Adding a seperate, independent source of motor control seems like a bad idea, Basic Stamp or any other chip, due to the likelyhood of conflicts between controllers.
  • bennettdanbennettdan Posts: 614
    edited 2007-05-26 07:44
    Hey,
    ·· You might try one of these http://www.sparkfun.com/commerce/product_info.php?products_id=242#·and a AD converter chip.
    ·
  • 88incher88incher Posts: 14
    edited 2007-05-26 09:31
    Wingteal said...
    88, I have a small plasma cutter and am a novice Stamp user. I understand the need to optimize the arc length over an imperfect work surface. It sounds like your cnc cutter moves the head and not the work. Either way, why not employ a simple mechanical strategy and let the plasma head 'ride' on the surface of the work? You could construct a carriage that rides on ceramic slides or wheels. The initial arc culd be struck at .300" on the Z axis and then after a delay, the head could be released to ride within a selected carriage height. Spring suspension could be added. Beyond the plasma headheight issue, I can imagine that some type of arc anaysis, perhaps in conjunction with the voltage output, would be useful for cut optimization and energy savings specific to particular materials being cut. I believe that there are generic sensors that can gather data on the optic wavelengths and heat generated by the cutting arc. Surely this has been accomplished by the major manufactures and is searchable. It is an interesting project and I hope you report on your progress.
    Nice to hear we're not the only ones doing this kind of project. It's very important for us to keep the torch height at it's optimum as this is the bigest factor in avoiding a heavily angled cutting kerf. We really are trying to avoid any contact methods of height control as it's not the commercialy accepted method. We are actually looking at measuring the arc voltage again, I·think we can get a voltage probe before the HF if injected.

    Thanks for the input,
    James·
  • 88incher88incher Posts: 14
    edited 2007-05-26 09:38
    bennettdan said...
    88incher,
    Have you used a current meter and seen if the current is about the same when the cutting head has the propper arc? if the machine uses a steady current as it is cutting then you might be able to read the current with a shunt resistor and then have the stamp make the changes to the height of the plasma cutter head with a stepper motor or servo.
    Hi Bennettdan,

    Interesting sugestion, unfortunatly the plasma power source uses a drooping slope charicteristic (i.e constant current and variable voltage) unlike a MAGs power source which uses constant voltage and variable current.
  • 88incher88incher Posts: 14
    edited 2007-05-26 09:40
    Kevin Wood said...
    Everything that you are trying to do should already be handled by your PC based CNC software. If it's not, that might be your problem right there.

    Adding a seperate, independent source of motor control seems like a bad idea, Basic Stamp or any other chip, due to the likelyhood of conflicts between controllers.
    Hello Kevin,
    Problem we have is that once the system is developed it needs to be possible to retro fit on to other peoples machines that eather have no or malfunctioning THC controls. Our software will handle the job but most peoples won't!

    James·
  • bennettdanbennettdan Posts: 614
    edited 2007-05-26 15:36
    You might be able to use the Infared module I posted in my last tread...I dont know if the heat from the torch will be a problem..
  • kelvin jameskelvin james Posts: 531
    edited 2007-05-26 16:08
    Out of interest, is there a problem using voltage feedback after the hf arc start? Couldn't you just monitor the hf control circuit to determine when it is off to enable the voltage sensing?
  • 88incher88incher Posts: 14
    edited 2007-05-26 17:07
    kelvin james said...
    Out of interest, is there a problem using voltage feedback after the hf arc start? Couldn't you just monitor the hf control circuit to determine when it is off to enable the voltage sensing?
    Hi Kelvin,

    That's a very good point, I don't think that it's a constant HF curcuit, the only problem with HF is it's ability to jump large air gaps (which is kind of the point of using it) I'm not sure how yuo would isolate it from the voltage sensing circuit untill it had dropped out?

    James
  • 88incher88incher Posts: 14
    edited 2007-05-26 17:08
    bennettdan said...
    You might be able to use the Infared module I posted in my last tread...I dont know if the heat from the torch will be a problem..
    It's a good idea, but I just fear that it is such a harsh envoronment that it may fail quite quickly.
  • bennettdanbennettdan Posts: 614
    edited 2007-05-26 17:39
    Well the sensor will read from like 30 Cm away which should be out of the harsh environment and they are industrial sensors that are in alot of industrial sensor you see in factories. I just dont know if the arc will throw off the reading but I dont think they will fail due to the environment it has a lens to protect it pretty good if you had a arcshield to keep down on the light and heat from the cutter inbetween the sensor and the torch head it should work fine..
  • Kevin WoodKevin Wood Posts: 1,266
    edited 2007-05-26 19:39
    88incher said...
    Problem we have is that once the system is developed it needs to be possible to retro fit on to other peoples machines that eather have no or malfunctioning THC controls. Our software will handle the job but most peoples won't!

    Just to let you know beforehand, I'm not trying to rain on your parade.

    It seems that the solution to the problem is to upgrade people to the latest software, or fix the malfunctioning controls.

    Here's the way I see it - if you have seperate controllers that can adjust the height of the plasma cutter, you need to maintain very good communications between them to avoid contol conflicts. When you add in all of the necessary checks and balance to the new system, it starts getting complicated.

    If the whole idea behind the main controller being able to do it's job is that it knows precisely where everything is at all times, adding a second controller means that there needs to be a way for the main controller to adjust itself with what basically becomes a real-time variable program offset. Without this, what's to stop it from crashing the cutter into the cutting surface, etc.

    So (again, this is the way I understand it) you are no longer looking at a simple BS2+sensor+motor controller to compensate for the main controller software's deficiency. You need to build a system with a lot of real-time feedback between the primary and secondary controllers.

    If that is the route that you do take, where that type of feedback is imperative, I would suggest that you upgrade from the BS2 series to either the SX or Propeller chip. Both have a lot more horsepower than the BS2, as well as having much better real-time capabilities. The SX has an interupt system and 75MIPS performance, the Propeller has highly deterministic computing via 8 cores with a common scheduler.

    I guess my point is that all of the motion should probably be handled by one control program, even if you have a seperate BS2+sensor obtain a height/voltage reading and pass it to the control program.
  • 88incher88incher Posts: 14
    edited 2007-05-26 20:45
    Kevin Wood said...
    88incher said...
    Problem we have is that once the system is developed it needs to be possible to retro fit on to other peoples machines that eather have no or malfunctioning THC controls. Our software will handle the job but most peoples won't!

    Just to let you know beforehand, I'm not trying to rain on your parade.

    It seems that the solution to the problem is to upgrade people to the latest software, or fix the malfunctioning controls.

    Here's the way I see it - if you have seperate controllers that can adjust the height of the plasma cutter, you need to maintain very good communications between them to avoid contol conflicts. When you add in all of the necessary checks and balance to the new system, it starts getting complicated.

    If the whole idea behind the main controller being able to do it's job is that it knows precisely where everything is at all times, adding a second controller means that there needs to be a way for the main controller to adjust itself with what basically becomes a real-time variable program offset. Without this, what's to stop it from crashing the cutter into the cutting surface, etc.

    So (again, this is the way I understand it) you are no longer looking at a simple BS2+sensor+motor controller to compensate for the main controller software's deficiency. You need to build a system with a lot of real-time feedback between the primary and secondary controllers.

    If that is the route that you do take, where that type of feedback is imperative, I would suggest that you upgrade from the BS2 series to either the SX or Propeller chip. Both have a lot more horsepower than the BS2, as well as having much better real-time capabilities. The SX has an interupt system and 75MIPS performance, the Propeller has highly deterministic computing via 8 cores with a common scheduler.

    I guess my point is that all of the motion should probably be handled by one control program, even if you have a seperate BS2+sensor obtain a height/voltage reading and pass it to the control program.
    Hi Kevin,

    Don't worry your not 'raining on my parade', the whole idea of this thread is to get ideas on how to produce an effective and THC by any way possible,·we have nothing at the moment so have no perticular path to force. I'm just chucking out as many ideas as I can to gauge peoples responces, as I've not got an electronics background I thought this may be the best way to start.
    What would be the best way to clean up this post? Should I re-write a·detailed brief of what our objectives are?

    Cheers,
    James
  • kelvin jameskelvin james Posts: 531
    edited 2007-05-27 03:11
    Nothing wrong with a learning experience for anyone, whether it is specific to their objectives or not.

    From my understanding, the hf arc start is just that, it starts the arc and is over. The hf starts a pilot arc in the torch tip, and runs until a constant arc (current)is sensed to the material being cut, and then the hf generator is shut off. It is kind of like a pilot flame for your furnace burner, but has the ability to shut off. I found this nice web page describing the function and more

    www.centricut.com/TA_StartingProblems.htm


    I can't give you any detail on the isolated circuit needed, but Mike already stated the ADC idea needed to use with a stamp in a previous post. As far as i am concerned, there is no distance measuring sensor that can operate in a complete xy axis radius around the cutting tip that would work in that appicatiion. Only my opinion, so there may be some other options i don't know about.
  • skylightskylight Posts: 1,915
    edited 2007-05-28 06:37
    In my industry in the old days of using Arc lamps for TV lighting it was found that the best mode of·maintaining an arc·was an experienced operator, taking that experience and installing it into an electronic interface may end up being a complicated job?

    Certainly the selection of a suitable sensor seems to be a big hurdle, perhaps another way of approaching the problem is needed?

    The unpredictibility of the warping of the material being cut doesnt help, is this where the problem needs investigating?

    Coming up with methods to either eliminate or minimise the warping, allowing the cutting head to stay at a relative distance to the work surface.

    Could the underside of the material sit in a form of coolant? The actual cutting area will of course gain extreme temperature in order to cut but the trick being to rapidly cool the surrounding area to stop warping, could this be achieved? Would the cutting torch have the power to·be able to vapourise the contact area regardless of it being cooled?
  • Damien AllenDamien Allen Posts: 103
    edited 2007-05-28 20:20
    Wouldn't it be possible to monitor the current using a current transformer? There would be obvious differences between arc and no arc. Also with a bit of experimentation there must be a sweet spot current when you are at a perfect distance. I realise this may vary dependant on the thickness and type of material but again this shouldn't be a problem. This would also give you the isolation you need so the HF won't send your controller loopy.

    Damo
Sign In or Register to comment.