Tweaky:

 By the way, it occurs to me its pretty simple to get this arduino to convert a 100hz pwm to 5Khz pwm.. So with one of these
Mach3 could easily be converted to allow for 100 power levels... just by adding a cheap arduino as a pwm up-convertor..Hadnt thought of
that before for some reason..

Art

Hi Guys:

   I jus finished hooking up the new control panel. It now takes a 50hz PWM and converts it to 5Khz PWM. So now I have Mach4
so I can type an Sword from 1 - 100 and get 0-100% power in 100 steps. Thats was job 1. I also have an override for that by a pot on the panel so I can override the Gcode program on the fly by turning the pot like a brightness pot to control the output.
   Also, rather than actually turn on the laser, the PWM , when sensed, just tells the laser to be prepared to fire at that power, if PWM is not sensed
but the Spindle on command goes high, the panel then uses just th epot for power. SO when Darwin ( or mach3) commands Spindle on, the laser gets
ready. The panel is also wired to the X and Y step and direction lines. ( you have to hook up 5 wires from the panel to the breakout baord.).

   SO now, when an step is sensed on X or Y, the laser will trigger off a shot of n pulses set by the user on the panel. When it exhausts those n pulses
it stops. Each time it sees a step it resets the counter, so if you move slow or fast, the power per step is the roughly the same, this should make grey scales easier as in this method speed is not as important.. more as I do more testing...

Art

Hi Art,

It’s all sounding very exciting and as you say, that system should certainly make greyscale working a lot easier to implement.

Just something, mentioned before and to keep in mind, which only affects laser operation and makes no noticeable difference to spindle operation…

Mach4 makes the changes to the S word (and the subsequent PWM output) considerably faster than Mach3 and as near as I can measure, changes occur up to 50mS before the completion of the preceding line of Gcode (the actual period in mS being dependant on the execution time of this Gcode line).

In the attached pic. of one of my earlier test samples, there are two changes of S word occurring when switching from an ‘engraving’ to a ‘cutting’ toolpaths and in both instances the increase in PWM pierced the work (shown by the arrows) at the end of the ‘engraving’ toolpath.
This fooled me a bit to start with but thanks to help by Craig and Scott adding a delay in the Mcode switching macro’s resolved the issue. Just how this can be resolved when using an axis direction pin for laser switching I have yet to discover but it is all part of the great learning experience.

I would just like to take this opportunity to offer a big “thank you” for setting such a low price for the Darwin pre-buy license. Keep up the good work my friend. 

Tweakie.

Tweaky:

>>In the attached pic. of one of my earlier test samples, there are two changes of S word occurring when switching from an ‘engraving’ to a ‘cutting’ >>toolpaths and in both instances the increase in PWM pierced the work (shown by the arrows) at the end of the ‘engraving’ toolpath.
>>This fooled me a bit to start with but thanks to help by Craig and Scott adding a delay in the Mcode switching macro’s resolved the issue. Just how this >>can be resolved when using an axis direction pin for laser switching I have yet to discover but it is all part of the great learning experience.

  Cant happen in this method, no laser is emitted when turning on the PWM, the panel gets the new power within a few ms.. but even that time doesnt
matter as radiation only happens as the step pin pulses. The step pin, resets the pulse counter which then allows radiation only for a set number of pulses
and then stops..unless another step pulse comes along. The next step pulse resets the counter and radiation starts again..or continues as it may have still
been shooting as a result of the last pulse. 

  In other words you can turn on the spindle , go for a walk, come back and then jog, the laser cuts only during motion.. Its my hope this cures many ill's
in terms of early trails and such.. mainly because you can enter a spindle delay.. a few ms will do it, and your power will always be proper no matter
if your in Mach3 or 4. This panel is designed to work as easily on M3 as M4 or EMC.. all you need is the 5 wire hookup to the breakout board.

 I hope to turn on power this week and test rather than just look at numbers, so we'll soon see if Im barking up the wrong tree.

Art

Hi Art,

In my mind, I know you are right but I will be interested to hear if you run into any issues with timing when changing the PWM duty cycle between low to high levels within a running Gcode.

Tweakie.

Tweaky:

 lol, yes as we all know the path to hell is paved with good intentions.. ( and idea's..).

  Ill show in a video how it all works out..

Art

Hi Guys:

(Copied from printer port forum.. wrong spot )

 First laser engraved photo done with Darwin. Darwin sent 40 levels of power ( of 100 possible) via
serial dataclock transmission while engraving at 7000mm/min.

https://plus.google.com/u/0/photos/118339423221700707274/albums/6054479007323314049/6054479008542443554

 Theres a video on the page as well. Im still just developing a method for darwin to do engravings or pin
punching.. I find surprisingly that my servos have a terrible following lag only noticable because of the lasers ability
to fire instantly on any step command at any power. SO if I cut in both directions, I get a "Carrot on a stick" problem
where the laser fires on the step pulse, but the servo doesnt get there for another 10mm's of travel.
  So Ill either have to put in a following error calculation for slow servo's.. or replace them if I wish to engrave in both directions.
Cutting isnt so much an issue. The photo shows quite an impressive quality for laser on melemine , and at least verifies that
adding a simple arduino, and hooking up 4 wires to the breakout board allows a large amount of laser control..

Art

Great work Art, the project is all starting to take shape now you are into the actual 'power on' testing stage 

Tweakie.

Hi Guys,

This is the little sub-panel fitted in it's place on the controller board. I used a couple of spots of cyno to fix it's position on the IC's before making the through connections to the main board. The function of these two switches will be software programed to control such things as ‘last setting store and retrieve’, 'PRF frequency', 'menu selection' etc. What started out as a simple project just keeps growing 

Tweakie.

Hey,

Do you have a build file like the J48 Controller for this addon? I got my hands on a 30w J48-1S-LSI laser. We were throwing it out and wanted to project it instead of landfilling the laser.

Thanks Tweakie

have a look here this is what Tweakie, Art and a few others are doing

http://gearotic.com/ESW/FavIcons/index.php?board=16.0