Advanced Search

Ok- so the first thing you suggest to try would be single ended.  Is just setting the jumper DOWN really all I need to do in this case and the axis should (ideally) behave like before? Or will there be more to it?
Well, I will simply try to jumper it to TTL and see if I can run the axis single ended properly or not.

Years ago I remember such "differential converters" on Uhu´s (DC-drives) - these where small and pretty inexpensive parts. But I just knew them for converting A+B channels to A/A* and B/B* , did not need the index in this times....
If I could use such converters to get a differential Index signal out of a single one and so solve the issue I would be fine with it.
Unfortunately I can not simply exchange the encoders as these Fanuc encoders seem to be part of the servo motor and the rear end bearing of the motor.

So I will first try to run the axis with the 7i33 jumpered to TTL, see how it behaves with the single A/B signals and if I can use the index signal then.

Thank you
Sigi

 

Maybe we can combine them?

I need some help here again. I am having a weird issue with my machine. I first thought it has nothing to do with Riocore, but now I am not so sure anymore.

Posted here: forum.linuxcnc.org/12-milling/57122-mach...-programm-repeatable

Oliver or someone with experience, could you explain a bit more what the settings in the Stepper dialog under Step mean?

The Step signal for my machine is inverted. I tried to remove the inversion from the step signal, but than the Stepper don´t move at all. Does that make sense??

Aciera wrote:

Can you post your modified kinemaitics file and your ini file?
 

Here's all my files.  I've included .ini, .hal, and kinematics comp file.  

OK, will get the hal file.

I did most of my work on big iron. Watching what happens with an AC vent hitting the top of a machine with a laser will drive you crazy.
Good luck!

Yes, it’s a rolled screw it has error from the manufacturing process. Thomson claims +\- 0.004 per ft / 300mm.

Though it could be mathed out in the SPU if I was using UCCNC or MASSO.

I'm happy to report that building the 2.9 branch (incorporating the merged PR mentioned above) from the LinuxCNC github solved the problem - the mdi-command-NN pins show up and function as expected.  Thanks for the fix!

-Rob

Nevermind, I forgot how to correctly read and comprehend English.

I would use an RC filter if in a high noise environment

Also not sure from your description,  whats 24V and whats 5V
no 24V signal should be anywhere near 7I92T I/O pins

Can you do a:

halcmd show all | grep home > beforehome.txt

With LinuxCNC launched and enabled but before being homed

and

halcmd show all | grep home > afterhome.txt

With LinuxCNC launched and enabled after being homed (well at least home button pressed)

and post both berforehome.txt and afterhome.txt here?

Ahh, looks like the index is single ended, maybe that's the issue.

Unfortunately the 7I33, each encoder input is all differential or all single ended

So one thing to try would be to select single ended instead of differential
encoder mode one the 7I33. This is not ideal as it less resistant to noise
but may be OK.

If setting the 7I33 to single ended encode mode works, there is also
the possibility of using custom firmware that puts all the index inputs
on one 7I33 encoder connector  so you could have differential for A,B
but TTL mode for the Index signals.

Another possibility is to add a TTL to differential converter for the index signals
Mesa has such a thing, but its really just 1 chip on a board

(26ls31 or one of the many equivalents)

No, I also watched the Z- Index and spindle index (not as brief as X) but watched the pins in HAL-Show and they do not change either.
Thats why I thought I made a mistake in the HAL.
The 7i33: otherwise the motors rum smooth as the spindle does and are perfectly controllable (motor and spindle)
The encoder cable- most likely- I will check this asap.
Interesting is that the encoders have Ua1 and Ua2 both also inverted, but the Index Ua0 not (also spindle encoder)

 

OK So it looks like the FPGA card is OK.

This leaves the 7I33 and the cable

Do the other index inputs work?
 

HAl and ini files attached.  We are retrofitting linux cnc to work with a Bostomatic 3 axis milling machine using a mesaboard.  We have accomplished a lot however we cannot seem to figure out why our machine does not follow the homing procedure.  Jogging works, we have remapped M6 successfully... we even made a bust using the machine.  However when the home all button is pressed (using AXIS as gui), the only movement that occurs is the home offset or home (parameter) however many units they are set. We used halshow to reveal the signals as a way to hopefully diagnose the problem, but we have found is that some signals are not present and they should be.  For example the pins with the verb homing do not get signalled, but those that end in homed are set after the home-all button is pressed. 
I have read the Homing Configuration in the docs many times through and we have essentially exhausted every combination of home_search_velocity, home_latch_velocity...etc.  The configuration we want to use is one where latch_velocity has an opposite sign than home_search does and we are using the index, and our soft limits are being shared as the home switch. (home_ignore_limits = YES) .  I've been thinking the only solution is to write a custom homing module using homecomp  or an existing one perhaps cia-402.   A response either suggesting other fixes or how to use homecomp would be extremely helpful.  Thank you