Page 3 of 3

Re: New low cost DSP laser controller board

PostPosted: Sun Aug 21, 2016 6:56 am
by BenJackson
educa wrote:I searched the manual, but cannot find how low 1 PPI pulse takes ? Usually this is around 3ms, but how much is it on the DSP board ? And it is configurable ?

Here's what I've learned:

  • PPI can be enabled or disabled. If disabled, laser is on continuously.
  • If enabled, 50 PPI is the minimum (enforced after you hit "ok", so you might not notice)
  • Duty cycle of the laser output is always 50%, regardless of pulse rate
  • PPI is not "pulses per inch". It is pulses per distance, and mostly scales linearly with speed.
  • It does not scale linearly with the PPI setting. It does go up and down vaguely with PPI setting, but there must be other limits. For example, at 25.4mm/s (1 IPS) and 1000 PPI, I got 333 PPS instead of 1000. But setting 500 still gave me fewer (not straight cap at 333) and 250 gave me fewer, but not half as many.
  • Pulses can get much shorter than 1ms. I didn't try to find the limit, but I saw frequencies >2.5kHz, and duty cycle is always 50%.

The good news is that I was able to bump my top speed and acceleration for my X axis way over what I could get on my PC with linuxcnc. I had to bump my current over what was recommended in the E5 hookup diagram, but that's pretty generic. I don't really know what the pololus were set to, because the tiny pot on those boards is inscrutable.

(Since I had the scope hooked up anyway, I looked at the laser drive signal while engraving, and it looked nonsense to me, but I have yet to let the E5 actually fire the laser)

Other notes:
  • OUT1 seems to be an active high (5V) laser output (so just inverse polarity of the laser TTL signal)
  • OUT2 seems to be an active high (5V) "job in progress" signal. It doesn't activate for manual movement or laser firing, but stays high for the duration of a job.

Re: New low cost DSP laser controller board

PostPosted: Sun Aug 21, 2016 10:13 pm
by BenJackson
Another note on OUT2 (which appears to be +5V during a job): On a cold power up, it goes high for 296ms at start (almost immediately). This doesn't appear to be an electrical anomaly. It's got a nice sharp rise and fall and goes to 5V. The laser PWM signal goes high for the same period. Luckily the TTL (active low laser drive) goes high a few ms earlier, which is one of the reasons I think this is under SW control.

It doesn't happen when you hit RESET on the front panel. EDIT: Sometimes it does happen when you hit reset.

If you connect a relay to this signal, that is plenty long enough to energize the relay, unfortunately.

Re: New low cost DSP laser controller board

PostPosted: Tue Aug 23, 2016 6:11 am
by BenJackson
After testing PPI with paper, wood, and cloth, my basic conclusion is: PPI can work for wood, but probably not for paper or cloth.

The basic reason is that you cut wood very slowly, which means reasonable pulse counts (like 500 PPI) have sufficient pulse durations that you get the desired effect, and the slow speed means there's not much variation of speed in curves or corners. It's fussy to tune, because PPI and speed still interact, but once tuned, you can cut whatever shapes you want.

For cardstock I found that it was possible to get good settings for rectangles, but those didn't translate to more complex shapes. The varying speed of the cuts in curves etc caused the pulse widths to vary a lot which changes the power considerably.

Re: New low cost DSP laser controller board

PostPosted: Sat Aug 27, 2016 4:02 pm
by BenJackson
The engraving algorithm works really well. It's smart about skipping blank space, both horizontally and vertically. The engraving quality looks good. And I've easily been able to drive my X axis 2x as fast as I did from a computer (where I ran out of pulses/sec).

One tiny flaw: The "box" button that traces out the job area does not take into account the blank areas of the engraving. I actually haven't tested whether blank areas are considered for "beyond size", although it's possible, since "beyond size" knows about overshoot for engraving.

Edit: Forgot something else: I can't get the USB thumbdrive feature to work. The E5 clearly recognizes the difference between no thumbdrive and thumbdrive present, and it complains when the drive has an invalid format, but it will never find work files on the drive. The LaserCAD software automatically truncates the filename to 8.3 when writing as a file, but maybe there's another step? I searched the LightObject forum and found similar questions with no answers.

Re: New low cost DSP laser controller board

PostPosted: Sat Aug 27, 2016 4:26 pm
by educa
Ben, if the dsp controller can let your X go a lot faster, then I guess it shoots the laser a lot faster too.

However. If you calculate a laser pulse at 3 milliseconds, then you can only shoot 333 shots per second.

If I take a size of 0.005" as dot size, then 333 dots is 1.665 inch. So theoretically you can only shoot then at speeds 1.665inch per second. That is extremely slow for engraving.

Do you have any means to check (logic analyzer maybe) how long the dots are pulsed ? I guess that the laser is pulsed in pulses far shorter than 3ms then ?

I'd like to find some more info about that actually, since I will try to engrave from an arduino due (84Mhz)

Re: New low cost DSP laser controller board

PostPosted: Sat Aug 27, 2016 7:41 pm
by BenJackson
It does shoot the laser really fast when engraving. I didn't turn on PPI for engraving, and I have no idea what it does in conjunction with engraving.

I did put a scope on the laser pin while cutting with PPI, which is what I described a few posts back. The duty cycle is always 50%, so the pulse duration is just half the period. You can see I saw the period go all over the place, from very high to very low.

Re: New low cost DSP laser controller board

PostPosted: Sat Aug 27, 2016 8:41 pm
by BenJackson
WARNING: CUT BOX MODE CANNOT BE CANCELLED. There's a menu option to cut out the bounding box of your work. I tried using just to make a quick test cut while adjusting the exhaust fan. Once started, it will not stop for STOP, ESC, PAUSE, or OPEN PROTECT. Luckily my door is wired through the laser PSU's water protection, so opening the door prevented the laser from firing. But this is a dangerous feature, and I would recommend avoiding it. What did stop it was RESET, and of course I could have turned off the main power.