[rlwoodjr]

A while back I bought a mill. See link for mill pictures:http://www.cnczone.com/forums/bridge...rigid_ram.html.

Since it had encoders attached to the steppers I decided to try and make a DRO. I still need to make a box to put it in.....nothing fancy just a zero function at this time. However I did leave the serial pins open so I can add functions if I decide to later. The main chip is the Atmega328 ( same as the Arduino). The chip is actually programmed with the Arduino Uno. I attached the current code in a text file.

[rlwoodjr]

put it in a box and mounted it to the mill....

[ksanalytical]

Very nicely done.

[zafarsalam]

Excellent job. What's the resolution of your encoders. And till what speed can you read the encoders without losing any pulses. I am beginning to explore arduino for my cnc projects, specially it's interaction with Mach3. Right now I'm just a novice with arduino and yet to make a useful app with it.

Zafar

[rlwoodjr]

Excellent job. What's the resolution of your encoders. And till what speed can you read the encoders without losing any pulses. I am beginning to explore arduino for my cnc projects, specially it's interaction with Mach3. Right now I'm just a novice with arduino and yet to make a useful app with it.

Zafar

Thanks!

I am running the 500 cpr at 4x so I get 2000 ticks per revolution. the mill is 5 tpi, so I get 0.0001" per tick. Max speed is 120 IPM. If I did the math right that means a tick every 50 microseconds (20 kHz). The interrupt for the micro-controller takes less than 10 microseconds...probably about 5 microseconds. I would guess that you could get up to 50 kHz without to much trouble.

[zafarsalam]

Thanks!

I am running the 500 cpr at 4x so I get 2000 ticks per revolution. the mill is 5 tpi, so I get 0.0001" per tick. Max speed is 120 IPM. If I did the math right that means a tick every 50 microseconds (20 kHz). The interrupt for the micro-controller takes less than 10 microseconds...probably about 5 microseconds. I would guess that you could get up to 50 kHz without to much trouble.

Is it Turbocnc on your machine PC?

[rlwoodjr]

Is it Turbocnc on your machine PC?

Yes it is. Boots in a few seconds...

[zafarsalam]

Yes it is. Boots in a few seconds...

Last time I used Turbocnc, it was losing position on arcs and didn't have constant velocity contouring. Then I switched over to Mach3. That was almost 6 years ago.

[rlwoodjr]

Last time I used Turbocnc, it was losing position on arcs and didn't have constant velocity contouring. Then I switched over to Mach3. That was almost 6 years ago.

That is too funny. I do not have any problems with arcs, but the constant velocity always sounded good...not so much for the mill, but I also have a router.

I like DOS in the shop.

[Bubba]

Yeh, Dave K says he has CV working on V4.2. Now if we can just get him to release it!
I use Turbocnc on my mill and router and soon (I hope lathe)!
ditto the dos in the shop.

[rlwoodjr]

Yeh, Dave K says he has CV working on V4.2. Now if we can just get him to release it!
I use Turbocnc on my mill and router and soon (I hope lathe)!
ditto the dos in the shop.

Dave has said that for awhile...any news on when it will be done?

[Bubba]

He hasn't said lately, had some correspondence Christmas time but nothing on that subject.

Wrote wishing a happy new year with a "nudge", but no response.

[zafarsalam]

That is too funny. I do not have any problems with arcs, but the constant velocity always sounded good...not so much for the mill, but I also have a router.

I like DOS in the shop.

Mine was 4.0, I guess. Maybe the position loss was with backlash enabled. It was a known issue then. Anyways, I am happy with Mach. I've controlled a variety of machine tools with it like, mills, machining centers, turning centers, turret punch, pipe bender, oxy-fuel cutter etc. Yet to do a router though.

[ycroosh]

This is pretty cool. I started working on a similar project, although I intend to convert mine into a rdiculous overkill that can interpret eithe G-Code or Excellon files (my mill isn't CNCd yet).
Do you have any plans to add more features?
What LED drivers did you use? (MAX72xx or something else?)

Reagards
Yuriy

P.S. I will probably "opensource" my project on my blog, but I don't want to clutter the forum with the URLs yet

[rlwoodjr]

This is pretty cool. I started working on a similar project, although I intend to convert mine into a rdiculous overkill that can interpret eithe G-Code or Excellon files (my mill isn't CNCd yet).
Do you have any plans to add more features?
What LED drivers did you use? (MAX72xx or something else?)

Reagards
Yuriy

P.S. I will probably "opensource" my project on my blog, but I don't want to clutter the forum with the URLs yet

Thanks, I have thought of making a complete gcode interpreter and step/direction generator, but I have to many projects....not enough time. This board has serial connection so it could me modified later.

See attachments for the schematic and pcb layout, I originally had more resistors to limit the current to the led's, but since they are only on 1/6 of the time I had to remove them to get the display to brighten up. You can see I am using just the atmega to drive the lcds.

[ycroosh]

Thanks for the details. It just dauned on me the that the axes don't need to talk to each other, so one MCU per axis works just fine . Is ATMega copind with the multiplexid pretty good? (i.e. no flickering)

Thank you
Yuriy

[eaglezsoar]

Thanks, I have thought of making a complete gcode interpreter and step/direction generator, but I have to many projects....not enough time. This board has serial connection so it could me modified later.

See attachments for the schematic and pcb layout, I originally had more resistors to limit the current to the led's, but since they are only on 1/6 of the time I had to remove them to get the display to brighten up. You can see I am using just the atmega to drive the lcds.

Thanks for the uploads and a very interesting article. Since this is labeled as
open source, would you be willing to post the g-code file for the PCB? It looks like a double layer board, is this true?

[rlwoodjr]

Thanks for the uploads and a very interesting article. Since this is labeled as
open source, would you be willing to post the g-code file for the PCB? It looks like a double layer board, is this true?

Single side board with jumper wires on top. See attached gcode files. M for mill and D for drill.

[rlwoodjr]

Thanks for the details. It just dauned on me the that the axes don't need to talk to each other, so one MCU per axis works just fine . Is ATMega copind with the multiplexid pretty good? (i.e. no flickering)

Thank you
Yuriy

I adjusted the delay to optimize the look until you cannot see any flickering.

[eaglezsoar]

Single side board with jumper wires on top. See attached gcode files. M for mill and D for drill.

Thank you for the files! It is rare to find someone who will go out of their way to share with others. Thanks again!