Smartstep / linuxCNC / novamill control

scr · Post by **scr** » Fri 18 Jan , 2013 18:32 pm

Hi all,

not sure where best to post this - after a day or so of reading and searching for the same topic, thought I'd take the plunge. Apologies for cross posting / if I've asked something that's already been answered.

Machine is a 1995 Novamill - serial 50615C. Acquired from Dave at Global Surplus in Halifax.

I am hoping to control the mill from linuxCNC. It's free, open and powerful. I originally (perhaps naively) thought that I'd be able to connect the parallel port on the PC to the port on the side of the control box, find out the pins that connect to the X, Y, Z, axes, and end stops etc - and get things running.

I had thought of getting some new stepper drivers, but it seems a waste to ditch the control box.

I have found this post - that details making up a cable connecting a parallel port to certain pins on the 96 way connector on the 'Baldor Nextstep' board.

viewtopic.php?f=4&t=2264

Before I do anything foolish ... do I remove the board on top of this, then make up a connector to connect to the aforementioned pins? As far as I understand it... (simplifying things) I need to bypass any of the Denford control stuff that interfaces with old Denford software, and get the PC's parallel port talking / controlling the stepper drivers, end stops, milling head, and so forth. Is there another post that I've missed that details this stuff? Has anyone used linuxCNC with the Novamill? Should I use other software? I have EMC linux installed and would be happiest using that if possible.

Any advice would be appreciated - if I need to move this post to the Novamill section of the forum, or similar, please let me know.

Very impressed with the level of detail and help available here on this forum - a true indication of Denford quality and service!

Images included for clarity ...

Ta
Stuart

angel-tech · Post by **angel-tech** » Fri 18 Jan , 2013 19:20 pm

i've put loads of these onto mach3.

scr · Post by **scr** » Fri 18 Jan , 2013 20:44 pm

UPDATE

I removed the top board and de-soldered the original wires from the 25-pin chassis connector on the control box case.

Connected the X,Y,Z axis step, DIRs and GNDs as stated here viewtopic.php?f=4&t=2264

I used some nice jumper/connector cables to connect directly to the massive header - easier than sourcing a 96-pin connector!

Connected to LinuxCNC stepper test program - and, it lives!

Found that the axes didn't quite like the velocity of 1500 mm/s and accel of 150 mm/s2

Values of 150 mm/s and 150 mm/s2 seemed to wrk well - though each axis seemed to differ in how well it handled higher speeds. More testing required.

Now to work out the spindle control.

!

angel-tech · Post by **angel-tech** » Sat 19 Jan , 2013 0:36 am

radio control servo leads are good for the 96 way connector.

Post by **Denford Admin** » Mon 21 Jan , 2013 10:49 am

This post may lead you to some information on the spindle control:
viewtopic.php?f=54&t=3522
I can't remember if anyone has got the board to work from Mach or LinuxCNC but I can't see why not...it looks like it just needs a 5V PWM signal to set the 0-10v (isolated) analogue output.

angel-tech · Post by **angel-tech** » Mon 21 Jan , 2013 12:31 pm

I had it working of sorts, i had full control of the spindle using step output from Mach, but it was reversed. 0 commanded gave full revs and full speed commanded gave me 0 revs.
I needed some relay control, so i went to diycncs sp4dc board.

scr · Post by **scr** » Thu 07 Feb , 2013 19:15 pm

Angel-tech, can you give some more advice on where the PWM 5v connects to the board between the top board and the spindle drive board? The one with the isolated analogue 0-10v circuitry?

Post by **Denford Admin** » Fri 08 Feb , 2013 9:51 am

That info can probably be found here:
Baldor Nextstep Step and Direction pins - 96way connector
viewtopic.php?f=4&t=2264
viewtopic.php?f=15&t=2119

scr · Post by **scr** » Mon 10 Jun , 2013 10:56 am

Hi all,

so, thanks to user RNR107 I now have my mill up and running.

Decided against computer spindle control - settled for an on/off switch and pot for speed control. This seems a lot easier than the computer controlled option!

Both the switch and pot were wired directly to the Sprint control board:
- switch across inputs 4 + 5
- pot across inputs 1, 2 + 3 (middle pot pin to input terminal 3 on sprint board)

My next problem is with calibration of the software control (linuxCNC).

I am looking for specs on the (original) stepper motors:

- step time (how long step pulse is on) - currently 5 micro s
- step space (minimum time between step pulses) - currently 5 micro s
- direction hold (how long direction pin is held after change in direction) - currently 20 micro s
- direction setup (how long before direction change after the last step pulse) - currently 20 micro s

after some test runs (I went with default settings to simply test the mill) and leaving the machine at idle static position for 30 mins or so, the exposed 'Y' stepper is quite warm.
Would I be right in thinking that incorrect stepper timing settings might be increasing this warming / heating up of the steppers? I would have expected them to get warm, but not so much at idle / static position.. ?

Anybody have any ideas?

Cheers

Post by **Denford Admin** » Mon 10 Jun , 2013 12:21 pm

Stepper motors do get warm and even hot when at standstill.
They take the same current whether moving or not - some intelligent drives can reduce the current when at standstill, but I'm don't think the Baldor drives have the ability.
The timing spec's you are looking for will be related to the drive, not the motors.
Some cheaper electronics can miss signals if they are too short, or happen too quickly. Have you put new drives on our used the Baldor's ?

DavidJ · Post by **DavidJ** » Mon 09 Sep , 2013 9:53 am

Angel tech, can you please advise which pin you connected to for PWM speed control ?

I can resolve the incorrect min/max issue by inverting the signal (in software or hardware).

Information