[Momentz]

how is a air valve easier? doesnt that mean buying a compressor and having it controlling the air and all that jazz? wouldnt something electronic be easier? i dont know much about running air actuators, just thought the central locking locks would be easier or using something similar

here are the links that didnt work before

http://www.smp-sign-systems.com/prod..._Messer_e.html
http://www.globalimaginginc.com/equi...ng/Tools.shtml

[fatal-exception]

guy2b1,

Check here for info on setting up Mach3: http://www.machsupport.com/forum/ind...82.0/wap2.html

[Manuel1020]

Hellow every body. This is my first message in the forum. I am writing from Spain (excuse my mistakes in english).

I have a router and I built an A axis for tangential cutting. You can see it in the attached image.

It is very similar to the fatal-exception design but with direct conexion between the 1Nm NEMA 23 motor and the cutting axis.

Now my contribution to this thread: An automatic homing for the A axis. As you know to start cutting the knife should be aligned with the X axis.

This can be done manually every time that you start a cut or use my solution. In the image you can see a green disc of 5cm diameter and a small electronic circuit perpendicular to it. The circuit contains an optocoupler H21A1 (Phototransistor optical interrupter switch) with U form and two small resistors and the U hugs the external part of the disc. The optocoupler has an infrarred emisor in one part of the U and a capteur of infrarred light in the other part. When the plastic disc is between both there is no coupling and the output is 5 volts. I made a small perforation (0.7mm) to the disc and when the perforation arrives to the U the infrarred light arrives to the capteur and the output is 0 voltios. I send this output to the breakout board and to the Mach3 and works equal to a limit swith for xyz axis. Every time that do a homing in Mach3 the A axis looks for the perforation and stops there.

Now my firsts experiences with tangential cutting and then a request.
Mach3 works very well in tangential mode with G1 G2 and G3 codes and you can see in the image a small peace of corrugated with four circles joined by a rectangle cutted with it and with the knife that appear at the bottom (a x-acto knife) The problem is that after several cuts the knife brokes because the point is to weak for the effort. I tried then to mount another knife (the one that appear in the image mounted in the axis) but then the knife cuts in advance (appx. more than 1mm for each mm in depth). First result: The angle of the knife should be less than 20 degrees and this cannot be acomplished just with drag effort: An oscillating knife have to be developped¡¡¡

I think that can be done with a small DC motor, a cam and a spring, but I do not know how. Can any body help me?.

[fatal-exception]

Hi Manuel, thanks for posting and welcome to the forum. Your A-Axis looks good. I'm sure it works well. i like your idea for homing the axis. This is something that I didn't really think about. I just thought that I would manually home the zero once per session, and it will maintain it's position while everything is powered up. Can I ask how accurate and repeatable the homing system you built is? Within a degree or two?

Yes, the knife shown might be good for very thin materials where it won't overcut too much, but not so much for thicker materials. Is it possible to do cuts in thicker materials with multiple z moves? This would be a bit less stressful on the blade. I've also thought about grinding my own blades out of old 1/4" end mills (3/8" shank). I have about 200 of them laying around.

I've decided to build my design as it stands, although if I really want oscillation, I should be able to put it in after the fact by boring and reaming the 5/8" output shaft in my design and having the oscillating shaft slide inside of the main shaft. The motor, cam, and spring would then be located above the pulleys. That's the thinking anyways.

I'll get some pictures tonight of my progress so far on the assembly.

Paul

[Manuel1020]

Hi Paul

Thanks for your reply.

I will answer first your second question: Yes, it is possible to do cuts with multiple z moves. I did three z moves to cut the corrugated (4 mm thicker) shown in the my previous message image.

Now the accuracy question. Because I do not have any dial gauge to measure the accuracy I did as follows: I disable the automatic zero when homing feature in Mach3 and I home the A axis and put the reading manualy to zero. Then I start doing several A moves in both directions with the keypad and afterwards home again the A axis and look for the new reading.

The results after 20 trials was: 11 times I got a new zero, 6 times I got 0.36 degrees and 3 times 0.72 degrees. The .36 and .72 values are a logical result and not a random one because I am working with 5 microsteps in the stepper driver, 360 dregrees divided by 1000 (200 steps * 5) gives 0.36 degrees rotation for one microstep and .72 for two microsteps.

As consequence I think that the accuracy is better than 1 degree. In your case, with a 1:3 reduction and a better design than mine, you could get an accuracy less than 15 minutes arc.

Manuel

[vger]

I've just read this thread. Some very good thoughts here. I've done some mat cutting by hand and seen it done at the local craft store using manual mat cutters. Doing it with a CNC would be really nice.

Diving the cutting blade into the material at an angle could be done via several methods.

Electricly with a solinoid.
Pnumaticly with a small air cylinder.
Very small motor and lead screw.

With each of these I would think the actuator should rotate with the cutter which would mean most likely a hollow shaft spindle driven with an offset motor (stepper or servo). A slip ring arangement could carry power for the solinoid or motor / lead screw method or a simple swivel for the pnumatic method.

One question I have... with the vibration... would one want the vibrating movement to be in line with the cutting edge of the blade in a sawing action? Or should it move the blade in line with the cut direction?

For linear drive sawing action, would a voice coil drive method work? Perhaps even with a stationairy coil and moving magnet attached to the blade. Just run off of 60 Hz for 3600 strokes/min (3000 for 50 Hz). Or to move the blade in line with the cut direction, perhaps the head actuator from a hard drive to provide a rocking action.

Just some thoughts....

Steve

[fatal-exception]

Thanks Manuel, I just wanted to get an idea for how repeatable an optical system like yours would be. I guess the last question I have is how far from the axis of rotation is the window in the disk, and approximately how large is the window? I'm just wondering if I could make a small disk that would fit into the 'block' in my design and still be large enough to repeat within an acceptable range.

Hi Steve. Yep, all those methods would work, although I think originally, the developers of Mach3 were going to try to put the 'plunge at an angle' into the software.

I'm pretty sure that the oscillation takes place perpendicular to the work, ie, sawing action. Electrical would be an interesting way to actuate the sawing. It would remove a few mechanical parts anyways. Would there be an easy way to adjust the amplitude?

Anyways, I think I will put in the means to saw. Here's what I have in mind. The oscillation will take place using a teeter totter design, a dc motor with a cam on one side, and a fork on the other side. There will be 2 shafts, the outer one will rotate, the inner one will rotate and oscillate. The fork will move the inner shaft up and down. The inner shaft will be keyed to the outer shaft some how. Maybe the voice coil idea could be used in place of the motor and cam?

Paul

[Manuel1020]

Paul, I have thinking about how Mach3 solves the limits. It works as follows: Moves the stepper motor in one direction until the computer receives a high (or low depending on the setup) signal. Then stops and starts moving in the opposite direction until the signal comes again to low (or high) and then stops: That is the limit position.

In our case, a rotational axis, is the same. Mach3 starts rotation in clokwise direction until it finds a high signal (a hole or just a cut that allows the infrarred light to activate the phototransistor), then stops and starts moving in the opposite way until the signal comes again to low. This opposite movement is done of course microstep to microstep and then the accuracy will depend first of all from the microstepping. Assuming that you have no microstepping the minimum movement will be one step and the accuracy will be 360/200 = 1.8 degrees.

Now, how appart from the centre of the axis could be the hole or the cut? When more distal the better, assuming that there is microstepping or reduction (like your case) because more long is the length of the arc to be run. I think, but I do not have proves, that with 3 cm diameter will be enough for a good precision (less 1º). To this 3 cm you have to add the space for the electric circuit only to one side of the disc. You can see the space required for the H21A1 looking the Fairchild datasheet.

I hope this explanation will solve your doubtes.

I agree with you: The oscillation must be perpendicular to the surface that is being cutted.

Could you, please, after thinking in depth your solution for a perpendicular oscillator with a DC motor and cam illustrate it with some figure or cad plans or... in order to understand it. I am not an experienced man in mechanica as yours. I worked programming computers since them appeared in the 70's and now I am retired and I am learning and playing with routers, mach3, drivers, steppers and all these stuff.

[vger]

Paul,

Adjusting the amplitude on a voice coil driver might be a bit diffficult. The free air amplitude would drop as soon as resistance was encountered by the knife. Some older voice coil hard drive head actuators used an LVDT as positional feedback. A power amplifier drove the voice coil with 2 inputs, one from the LVDT and one from a DAC. Change the output on the DAC and the amp would drive the coil till the LVDT signal canceled out the DAC signal. A 12 bit DAC would give you 4096 positions (cylinders) on the head movement. Newer drives use "servo information" written in specific locations on each track of the disk. That is read by the head and then the voice coil drive is adjusted accordingly. For amplitude control, I suppose one could attach an accelerometer to the voice coil and use it's output as an amplitude signal and vary the AC drive level accordingly as a vibration calibrator works.

Using a simple camshaft drive with a DC motor might be the easiest to implement (like a prison tattoo gun), although changing the amplitude would involve changing the cam.

Steve

[Rene Nuñez Paz]

Hi, Manuel
Please, what type of stepper are you using for your tangential knife..

Thanks in advance

Rene

[Manuel1020]

Hi Rene

The stepper is a NEMA23 1Nm power working with 5 microsteps per step (1000 ms per revolution) directly attached to blade holder.

Manuel

[Rene Nuñez Paz]

Thanks.
Regarsds

Rene

[fatal-exception]

I've been working on the design a bit more. I have adjustable oscillation now built in. Although it's added quite a few more parts, I think it is doable. The way it's shown here, the oscillation can be adjusted from a minimum of .03 to a maximum of .08. The nested shafts are designed for a maximum stroke of 1/4". If the cam wheel was designed with a larger offset, more stroke could be accommodated. I'm not sure of the effect of stroke as it pertains to cutting, so this will be a bit of an experiment to start off with. (Maybe .01 is all that's needed???)

I suppose the motor could be replaced with anything that vibrates.

The only iffy part is the outer shaft. I'm not sure how easy it will be to drill a 1/2" hole down the middle for the 4 1/2" length. I guess I'll find out. I'm then going to press the brass bushing into both ends and then ream the final 3/8" diameter. Both the inner and outer shafts are going to be made from drill rod.

The teeter-totter will be made from steel.

It would be nice to have a way to lock out the oscillation in case there's material that doesn't cut well using it. Still have to give this some thought....

[Bristol, CT]

Interesting project. From my experience, the missing item is that the 45-deg blade need to be slide into the mat using a force (selenoid) at the same angle. A z-axis force will break or bend the blade. This may be the so-called W-axis that was referred to earlier.

[leonardojc]

Hi, i have a problem with the Lift Z setting, i trying differents values but never lift the Z axis. the A axis is rotating very nice. but the Z only lift when jump to other vector.
anyone can help me??

Thanks

[Manuel1020]

Leonardo
En General Config tienes que marcar el eje A como Angular y luego en la pantalla de Settings especificar en Tangential Control el angulo a partir del cual quieres que suba el eje Z y a que valor de Z tiene que elevarse. Yo utilizaba 30 grados y 2.5mm y me funcionaba.
Por cierto, ¿Utilizas un eje Z oscilante o fijo?
Manuel

[leonardojc]

Manuel, muchas gracias por responder, te cuento, ya tengo asi configurado todo, pero sigue sin levantar el Z en los angulos. Solamente se levanta cuando va de una trayectoria a otra.

el eje Z, va a ser Fijo.

[Manuel1020]

Leonardo
Estaba equivocado. El valor de Z que se pone no es la coordenada a la que tiene que ir el Z sino lo que tiene que subir frente al valor que tiene en ese momento. Por eso si has puesto que vaya a Z cero, no hara nada. Haz un programa para hacer un rectangulo y ponle 100 o mas para subir Z. Ejecuta linea a linea en simulacion y veras que subira y bajara el eje Z antes de cambiar por ejemplo de 90 a 180 grados el eje A. Por lo menos a mi me funciona asi.

[Rene Nuñez Paz]

Hi all,

Agree with Manuel, since the value displayed on MACH3 ZLift is related to Z must be raised as to rotate the blade where the blade angle spinning higher than those indicated in the parameter LIFT TANGENTIAN ANGLE CONTROL,

On the other hand wanted to see / propose if might be possible to lower / raise the blade instead of the Z axis use a solenoid to be activated with a macro style
when Z = -1 then select an output pin to activate the solenoid and this case the Z axis we would leave without setting in Motor Tuning.

I appreciate your comments.

regards

rene

English = Google Translator


Hola a todos,

Estoy de acuerdo con Manuel, ya que el valor ZLift que aparece en MACH3 es el relativo a cuanto debe de subir Z para que la cuchilla gire en los casos en que el angulo a girar la cuchilla sea superior al que se indique en el parametro LIFT ANGLE de TANGENTIAN CONTROL,

Por otra parte queria consultar/proponer si podria ser posible que para descender/elevar la cuchilla en vez del eje Z se use un solenoide, que se activaria con una macro del estilo
que cuando Z= -1 entonces active un pin de salida para activar el solenoide y este caso el eje Z lo dejariamos sin configurar en Motor Tuning.

Agradezco sus comentarios.

Saludos

Rene

[Manuel1020]

Leonardo
Se me acaba de ocurrir que quizas no has activado el Tangential Control de Mach. Tienes que ir a la pagina de Settings y pinchar en el titulo "Tangential Control" y el indicador luminoso a la derecha del titulo comenzara a parpadear. ¿Es eso? Si no es asi dame tu email y te enviare el fichero de setup de la fresadora de cuatro ejes que utilizo yo.
Manuel