Laser engraver from plotter project log, with pictures!

[KTP]

Hi,

I am starting a project log to build a very low cost laser engraver out of an old HP7550A plotter and some other bits from printers and copiers. The motion of the carriage in the HP7550A pen plotter is extremely fast, and well suited for a flying optic mirror/lens system.

Here is the donor, obtained locally for $15. These plotters can be found at computer salvage stores and on ebay. The 7550A uses Pittman DC servo motors with quadrature encoders...perfect!

[KTP]

Ok here are some pictures of disassembly (my favorite part!). You can see the Y axis Pittman servo in one picture, the other servo which moves the paper is hidden. Note the very nice alumining timing belt gears and the wide timing belt...looks very promising!

[KTP]

Ok, at this point it looked like the motion assembly and paper feed were one giant unibody construction, but ahah! there are some hidden screws which allow the Y axis carriage to be separated, reducing the size drastically. After removing the belt, pen carrier and motor, the Y axis assembly was presented to the band saw for a bit more disassembly

On the plus side, I have increased the travel to over 14 inches by using the space where the plotter used to pick up pens!

[KTP]

Ok, so now I have reassembled the Y axis after trimming off some plastic that was in the path of the soon to be added lens/mirror. I added a aluminum angle to strengthen the assembly which makes up for the plastic I needed to remove. The whole axis is very lightweight (good) and extremely fast (great). It should allow for insanely high IPS travel rates which is nice when you are rastoring a large image with the laser. Next step is to hook it up to an AMC servo amplifier, Pixie P100 controller, and Mach 3 to see how it moves. I will post a link to a video of this.

[KTP]

Ok, as promissed, I have made a video of the carriage moving under Mach3 control using a Pixie P100 step/dir to analog controller board and a AMC 12A8 brush servo amplifier. I am quite pleased with the results, the Pixie P100 lets you pull every little ounce of energy out of the Pittman motor. The motor is surprisingly strong for it's size, and the top speed should be well above the 2640 IPM (44 inches per second) seen in the video. The Pittman motor/belt reduction drive assembly seems to have 4000 steps per inch of travel. I have set the Pixie P100 step multiply to 4, giving 1000 dpi resolution. Using Mach3 at about it's maximum step rate of 44khz, this gives the 2640IPM feedrate I used in the video. I have also attached a tuning graph running on the Pixiecom software showing the response of the Pittman servo mounted in the carriage. You can tell the motor responds very well to a 50 step impulse error in the test.

video: http://www.skyko.com/videos/HP7550A.wmv

[miljnor]

Awsome thread man will be following it with baited breath!

What servo amps do you use? (I see you are the maker of the Pixe100 nice job)

thanks
Michael

[KTP]

Thanks for the kind comments Michael!

The amplifier I am using at the moment is an Advanced Motion Controls 12A8D, which is a bit overkill for these smallish motors. I bought quite a few of them very cheap on ebay and they work well with the Pixie, so why not?

The laser I am going to use initially is a small air cooled Synrad J48-1 10 watt unit. It is RF excited, powered by a 30 volt 7 amp DC supply and currently is outputting 16+ watts at 95% duty cycle. I have included a couple of pictures of it cutting some 1/8" balsa stick crudely waved in front of the focusing lens by hand

[thuffner3]

If I may ask kind sir, Where and what cost is the laser you mention above?

Thank you
Neil

[KTP]

Hi Neil,

The laser in the picture is made by Synrad (www.synrad.com). This unit sells for around $2450.00 new from Synrad, but can be found used on ebay and other places for $800 to $1300 or so. Unfortunately, they are so rock solid and they last so many TENS of thousands of hours that they hold their value on the used market. Lucky me, I got this laser for free out of a scrapped semiconductor machine that was going to the landfill. Unfortunately, I didn't get the rest of the machine.

[KTP]

Here is a quick frame I threw together using 8020 extrusion. I think I may have overbuilt it a bit It has inside clear dimensions of 24 inches by 48 inches, though my Y axis only has 14 inches of travel at the moment. Building for the future I guess...

The x axis linear rails are mounted as you can see in the picture. I found these IKO rails and carriages brand new on ebay for $125, so might as well use them rather than round rail. They have VERY little resistance to movement, which will be nice. Low mass too!

There will be 3 cross bars bolted to the underside of the top platform to hold the engraving honeycomb table. I will fashion a vacuum/exhaust box out of 1/8" something and attach it below the honeycomb. The lower shelf will contain the RF excited CO2 laser, dc supply, motor controls, air pump, and computer.

[KTP]

Yesterday I ordered the pulleys, bearings and drive shafts for the x-axis. I already had 19 feet of XL (.2" lead) 3/8" wide belt, so will use that for now. It is neoprene with kevlar tension members. Getting what I wanted in pulleys was a bit of a puzzle. I wanted all the pulleys to have a 3/8" ID so I could use 3/8" shafting running inside the 8020 extrusion to link the two drive belt pulley systems together. Several pulleys were out of stock for 5 to 6 weeks at SDP/SI, but I ended up with four 14 tooth pulleys, one 15 tooth pulley, and one 42 tooth pulley. The x axis drive belts will ride on the linked 14 tooth pulleys, which will be driven by a brushless servo geared down through the 15 and 42 tooth pulleys. This gives a final ratio of 1 rev of the servo motor shaft equals 1 inch of x travel (14/5 = 42/15). The brushless servo has a 2000 line encoder, so resolution will not be an issue. The final question is will the brushless servo be able to drive the x axis at speed and with enough acceleration? Some scratch calculations give:

1.5 Newton-meter stall torque rating of the servo motor. Using a hopefully conservative 1 Newton-meter running torque and noting that the 14 tooth pulley pitch circle radius is about 0.5 inches, we get a force of 79 Newtons on the belt, or roughly 17.7 pounds of force. For the acceleration, noting that F=ma, and taking a total guess of the mass of the y axis gantry and the drive components to be 4kg or less (under 9 pounds) we get a acceleration figure of 19.7 meters/sec^2. Enough?

I like the try it and see if it works method better, especially when using cost saving "what I have on hand" components.

More to come.

[KTP]

Decided to go looking for more information on the brushless servo I plan to use for the x axis. I found the specs at:

http://www.apicontrols.com/newapi/pr...rbo_mbt23.html

The motor I have is MBT-N233-BCNC Actually, I have 2 of them Ebay $15 each brand new...glad people don't know how easy brushless servo motors are to drive...oops

Looks like the Continuous stall torque is 13.6 lb-in (1.537N-m) and the peak torque is 40.8 lb-in (4.611N-m). The rated voltage seems to go up to 230VRMS at a whopping 6380RPM, but I am going to use an AMC B25A20AC brushless amplifier and Pixie P100 so will be limiited to the 115VRMS input to the B25A20AC with max RPM of 3190. So really this justifies the 1 revolution of motor shaft to 1 inch of belt travel, since the maximum IPS rate I could theoretically achieve would be 3000/60 = 50IPS. Not slow, but not much more than what the high end commercial flying optic laser engravers can do. I know they have optimzied the mass of the moving components much better than I have, but I want to shoot for the best I can with available funds and equipment. But really I need to keep the mass of the y axis much lower than the 4kg I used in the crude calculations...slinging even 2kg back and forth at 50IPS is going to bounce the machine I fear. Time will tell.

Will be interesting to see how this motor performs. Can't beat the price.