My cheap Chinese laser experience...

[Neon Forests]

In article <13nlli2sjr5r57c@corp.supernews.com>, john jardine
<john.jardine@idnet.co.uk> wrote:

"Neon Forests" <moog@wildblue.net> wrote in message
news:010120081351574620%moog@wildblue.net...
[...]
Keep in mind that the Rolands have steppers with around 38 ohms on
their coils.
One of the stock motors in the engraver is about 3 ohms per coil.
A huge load.
So I'd probably be apt to replace the motors or add a driver stage and
treat the old drivers as pre-drivers, but the whole PWM/chopper thing
is the first to get looked at.
[...]

Maybe it is the other way round, in that the Roland chopper is fixing the
current at say a constant 200ma feeding it's design 38ohms load, hence a
motor power of 1.5W but that same 200ma with a Chinese 3ohms only gives
120mW motor power!. The PSU voltage should have virtually no effect.
Anyway, I hope one day to have my own death ray.
Best of luck with the project (and please do a video of it cutting or
marking some plastic .

You might have something there, though I think they're getting more

current than that.
Hard to say as yet, but I have to break that rail soon to put on some
new connectors anyways, so I'll be able to get a good averaged value.
I'll also be able to see where the current goes when it's asked to go
slow. It may well be overlapping pulses and it's getting eaten up
there.
I don't know... yet.

2nd reply....

Cutting 10 mil plastic (and if you mean .01 inches vs millimeters), it
hasn't had much problem.
It'll go through 1/4 plexiglass if you leave the beam on it for about 3
seconds, but that's a long time for a cut.
If 10 millimeters, I'm not sure it could go through anything that thick.
At least practically- I can burn a hole through a 1" piece of plastic
if I leave it on for 8-10 seconds, but that's not going to help with
much cutting.

I've had no problem going through paper, cardboard, cork, many
plastics, nylon, vinyl, polystyrene, PVC and those types of materials.
It's really up to the material and whether it absorbs 10600 IR light.
Metals hardly care at this wavelength and power, but most organics and
petroleum based materials burn or cut readily.
Making a new gasket for one of my desoldering pumps right now..

K

 

[Neon Forests]

In article <13nkssnk5damscc@corp.supernews.com>, john jardine
<john.jardine@idnet.co.uk> wrote:

Looks suspiciously like the stepper driver motors are being powered down
(too early) to a 'holding torque only' condition about 1/2ms after an X or Y
step pulse has ended (IC134).
A thought only, but maybe linking out R147 and R160 could keep motor power
applied all the time, thus can deal with the slower step rates at full
torque.
Personally, I'd be inclined to drive the motors directly from 2 external
stepper drives fed directly from that printer port. You could then set the
motor current at will. Seems lots of CNC progs available that will translate
HPGL (or G code) to printer steps under your direct control.
Or even, do the translate prog yourself with aBreshenam line and (maybe)
circle draw routine. Most CAD progs out there just seem to generate 'move
to' and 'line to' commands and ignore 99% of the available HPGL or G code
possibilities.
Having said that, I haven't a clue about Macs! and these things simply may
not be possible

John,

Thanks for your input.
Well appreciated.
You may also be correct.
I haven't had time to really see what's happening.
I don't know if the drivers can handle a more continuous load over time.
I'm keeping them cool, but there's obviously limits to what current
they can pass either when clocked or continuous.
Since the motor control also receives a chopper signal (normally 20Khz)
and I'm slowing that down as well, I know there's homework ahead of me.
I believe I still need the chopper happening because I don't think the
drivers can handle a full linear load.
Keep in mind that the Rolands have steppers with around 38 ohms on
their coils.
One of the stock motors in the engraver is about 3 ohms per coil.
A huge load.
So I'd probably be apt to replace the motors or add a driver stage and
treat the old drivers as pre-drivers, but the whole PWM/chopper thing
is the first to get looked at.
I would also like to bump up the voltage to normal spec.
The plotter electronics want 28 and is getting 27 and while that's not
much, it could be the difference that's needed here.
I should also add that the laser's supply is regulated, so it's giving
27 volts no matter what.
The Roland uses an unregulated 28 volt line which reads around 40 volts
unloaded.
So it may well be better at providing pulse transients with a little
extra current at first.

Either way, thanks for your interest and suggestions.
It's a hobby... though I admit that most of my hand tools are now
labeled nicely. ;-)

Kevin

 

[Neon Forests]

In article <13ni96t2is1eke3@corp.supernews.com>, john jardine
<john.jardine@idnet.co.uk> wrote:

"Neon Forests" <moog@wildblue.net> wrote in message
news:301220072118486731%moog@wildblue.net...


http://www.synthfool.com/laser

An interesting read!.

The Roland should respond to something like "VS 1,1" added to the top of a
HPGL print file, (ie Velocity select #1, for pen #1). Is this not an
option?.

Thanks for reading.

It does respond to VS commands as low as 2 and that definitely helps in
slowing it down, but it's not quite slow enough.
I have to do multipasses on some materials to cut them.
Also I'm pushing the limits on the whole current vs torque issue.
As I slow down it down, it has a bit of a problem with torque or
overlapped stepper motor pulses (I haven't investigated yet) and so has
repeatability issues.
But I'm just a hobbyist here and it's a good start.
I may end up ditching the motors and electronics and starting off with
something better, but it was more along the whole "can it be done?"
thing than trying to end up with a piece of reliable production
machinery.

Kevin

 

[Neon Forests]

http://www.synthfool.com/laser