[dgapilot]

Here is a simple and inexpensive index pulse option. Use one of these http://www.lynxmotion.com/Product.as...8&CategoryID=8
mounted close to the spindle and have a part of the spindle covered with a light colored material (I used paper labels on mine). The calculations for the size of the flag didn't work for me, so I just kept increasing the size of the flag until I got a stable pulse.

Here is another solution for the index pulse http://www.cnc4pc.com/Store/osc/prod...roducts_id=129
Mine actualy has the older version of this board with an optical switch rather than the one they use now.

[Ron111]

Alan,
I did find my previous threads concerning automated threading and spindle sync pulses.
One in on the 9X20 Harbor freight and the other is the "Lathemaster 9 X 30 CNC adventure" if you look at post #40, #41 (PDF attachment file), and post #47, you can see How I did it. I threaded many times using this set up.

I will be out of town for the next couple of weeks (so if you have questions, it will be a while before I can get back to you)

Good Luke,

Ron

[dgapilot]

Ron,
your attachments aren't showing up!

[acondit]

Thanks, guys, but I am not doing a single line detector. I am building a 100 line plus index encoder. When run in quadrature mode this will give 400cpr plus index. EMC2 can now support a single line sync signal, but, since I started out to use the multi-line encoder I want to finish it.

My encoder wheel and detectors are already mounted on the lathe and working.

The problem that I ran into is that my BOB uses 470 ohm pull-ups on the opto-isolator input and my detectors are using a 560 ohm pull-down circuit. This results in the the detector signal being held higher than the logic 0 voltage all the time. Thus the parallel port pins are never seeing a low going input signal.

My solution is my small converter/inverter board which will work with the BOB's pull-ups and will be driven by the PNP outputs of my Detector boards through a 10K resistor to the base of an NPN transistor. The emitter of the NPN transistor is hooked to ground and the collector will connect to the opto-isolator input/470 ohm pull-up.

I have already redesigned my detector boards to use NPN transistors instead of the PNP transistors that they now use but I don't want to go to the expense of having a bunch made until I verify the operation with my small converter board.

Alan

[acondit]

OK, my persistence paid off.

My converter/inverter board solved the problem. As soon as I hooked it up I could see the signals on the computer. Then it was just a case of telling EMC2 which pins were which signals. Even the little speed bar in my pyvcp panel works.

Now I just have to learn how to use the threading codes to cut threads.

Alan

[acondit]

Here is a video of my lathe cutting one of a set of parts for my CNC router.

[ame="http://www.youtube.com/watch?v=PIg2ON_bbhY"]YouTube - CNC-BearingStandoff2ndOp[/ame]

I wrote a gcode routine to perform the same first op on 8 consecutive pieces. It faces off the end, cleans up the entire piece, cuts the bearing boss and then parts off the part and pauses for the shaft to be shoved in to the tool to cut the next part.

The second op starts with the first piece in the chuck, it faces it off to length, cuts the shoulder and then pauses for the next part to be inserted.

These parts will be the bearing standoffs for the y-axis bearings on my revamped router.

Alan

[acondit]

I hooked up my Logitech Attack 3 joystick to my lathe controller. I have the jog working but I haven't configured most of the buttons yet. It has 11 or 12 buttons besides the joystick.

EMC2 has a all of the necessaries for hooking up a joystick to a mill. It is basically only setup to handle the X and Y axis. So, when I hooked it up to my lathe I changed so that the z-axis moves when the handle is moved side to side and the x-axis moves when you move the handle forward and backward. Basically standing in front of the lathe like in manual operation, the axii move in the direction that you move the handle. Then I set up the trigger for finefeed, thumb1 for slightly faster feed and thumb2 for rapid jog.

I will probably set up a button (or two) to act as a slider for spindle speed (once I get the spindle speed control setup). I wonder if I could figure out how to use the scroll wheel to control spindle speed?

Another button will probably be setup to control a mister-coolant system.

I am a little reluctant to setup a button for the spindle on off. I don't want to bump a button and have the spindle turn on or off unexpectedly. Maybe if I build a more complete console then that might be a reasonable thing to do. Then I could also include a "machine-on" button as well and maybe a couple of buttons for home-x and home-z. Program "run", "pause", "resume" and "stop" would also be a useful buttons.

I guess I will have to sit down and map the various options to see what I have enough buttons to do and what is safe to do.

Alan

[123CNC]

Alan,

Are you using steppers and parallel port with EMC2 or servo's and third party hardware (e.g. pci PICO systems, etc.)?

Just curious about your hal files and interfacing the joystick.

A few of us may be further amused if you post a video of your joystick jogging and thumb trigger operation.

Nice work.

[acondit]

Alan,

Are you using steppers and parallel port with EMC2 or servo's and third party hardware (e.g. pci PICO systems, etc.)?

Just curious about your hal files and interfacing the joystick.

A few of us may be further amused if you post a video of your joystick jogging and thumb trigger operation.

Nice work.

I am currently using steppers and parallel port. I have several servos but I don't have any servo drivers.

I just basically followed the directions in the wiki here. Then changed these lines in my custom_postgui.hal file:

net joy-x-jog halui.jog.0.analog <= input.0.abs-x-position 
net joy-y-jog halui.jog.1.analog <= input.0.abs-y-position 
net joy-z-jog halui.jog.2.analog <= input.0.abs-z-position

to this:

net joy-x-jog halui.jog.0.analog <= input.0.abs-y-position 
net joy-y-jog halui.jog.1.analog <= input.0.abs-z-position 
net joy-z-jog halui.jog.2.analog <= input.0.abs-x-position

Then I set the speeds to be 0.5, 5, and 50. Increasing by a factor of ten.

I will try to get a short video in the next few days. It is a little hard to take a video holding the camera in one hand, keeping it focused and use the other hand to jog the lathe. I'm not all that coordinated.

Alan

[dan_the_welder]

I just took a look through your build log and I was admiring your very pro looking circuit boards.

Where did you get them done?

Very nice work all round by the way.

Dan

[acondit]

I just took a look through your build log and I was admiring your very pro looking circuit boards.

Where did you get them done?

Very nice work all round by the way.

Dan

Dan,

I did all the schematics in Eagle. Until version 5.2 of Eagle, I had been using Osmond-Quartz on my Mac to actually lay out the PCBs. All of my circuit boards up to the little inverter/converter board I had made through BatchPCB.com. The last PCB (my Inverter/Converter) I used Eagle and PCB-GCode to create the gcode and then I milled it on my CNC Router.

BatchPCB charges about $2.50 per square inch for boards, too high for production but pretty good for prototyping. Turn around time is probably too long for anyone doing commercial development but I think it works if you are patient.

Thanks,
Alan

[acondit]

Well, a delrin nut may be fine for a router but it really wasn't working well for my lathe. I just couldn't hold much better than a couple thousandths. I would get it set right on the money and then at the final diameter trying to do a finish cut I would find it off by 1, 2 or even 3 thousandths. I think the nut was just flexing too much for reliable cutting of precise diameters.

So I made a new nut out of cast iron. I got it done and it is installed but I haven't done any testing yet. It feels like it has no backlash. I was thinking about cutting it oversize and fitting it close with moglice, but I think that I will try it like it is for now.

The switch to a 10tpi screw doubled my max velocity from 7.5 to 15ipm on the X axis. With 4:1 gearing, 200 step motors and 10 tpi, I have 8000 full steps per inch ( I am running halfstep so it is really 16000 steps per inch). So, I have 0.000125" per full step. Hopefully this will prove more accurate than with the Delrin nut.

It is really wierd how these machines are a mix of metric and imperial measurements. The mounting screw is 6mm but the boss has to fit a hole bored in the crossfeed that is 0.375". Why not 10mm (0.3937")?

The first picture shows the partially completed delrin nut (now replaced). The second picture is the partially completed cast iron nut. The third picture is after the mounting hole has been tapped 6mmX1.

I forgot to take a picture of the threading operation. I drilled a 0.375 hole lengthwise through the nut and then I used the G76 gcode to tap a 10 tpi starter thread out to about 0.450. Then I made a tap out of a length of the acme threaded rod and used it to tap the nut to final diameter.

Alan

[acondit]

Someone asked me to post a drawing for my z-axis motor mount. Since my mount bolts to the existing holes for the leadscrew end bushing block, you would need to verify those measurements on your lathe before using. These lathes do not seem to be produced with highly repeatable measurements for all parts. Some parts are fairly uniform but others seem to have been produced with hand tools.

The first pdf is for the bearing block and the second is for the stepper motor mount. The length of the standoffs that you use depends on the coupling that you use.

Alan

[jackfrost]

Very noce Job. I love the workmanship

[acondit]

Very noce Job. I love the workmanship

Thanks for the compliment,

Alan