Advanced Search

My suggestion is to use m19

Use halmeter or halshow to check the encoder counts at the output of the Mesa driver.

hm2_7i77.0.encoder.03.counts is probably the pin to check.

I don't think that the PP can do it (not as a first step, anyway). It doesn't "know" that it is a repeated step (and even more so doesn't know that it _will_ be a repeated step)

You can increase the minimum segment length which will reduce the file size (and will also often improve the machine speed) and you should allow the PP to generate arcs, where that is a setting as that will both reduce file size and improve fidelity.

In theory an intermediate tool (even applied as an input filter to LinuxCNC) could reduce file size. But at that point it may be too late, what do you gain? You alread have the huge file on the system.

Spotting the repeats (especially repeats with an incrementing variable) would be a rather "entertaining" bit of code to write.

TTL should work with Mesa 7i77, so check the hal file and make sure Y axis is using encoder.03.
Also, just in case, encoder.03 is the 4'th encoder connector on Mesa, the enumeration starts from 00.

It might help to change to wayland or xorg in case you already did.

Most likely overclocking won't help much. If you're isolating 3 cores, you could try isolating 1 or 2 cores instead.

I'm retesting my 7i95t configuration now... and I see the error.

JT

That is odd for sure.

JT

My guess is there is an offset on the Y axis from a previous program or the program that called the subroutine.

Well I have been researching the scales. It appears these scales LS486 are incremental scales 1Vpp output with a separate  "R" output for reference position. It looks as though there is a reference pulse at each end 35mm from the full measuring length on one side and 70mm ish on the other, due to the reader position.

Someone here has already tackled these scales so i will post over on that thread and see how he went with them.  linuxcnc scales index ref marks
 
This is the pin out on the schematic. I also found a pdf of the specs so that has helped. Not sure these can be a hard limits but can be used to home to set zero. It couldn't be to hard to add some hard wired limit proximity switches in as a safety?

The pulse on these sine/cos scales ref mark looks like it is only 0.5V, I am sure one of the input IO mesa boards said that the digital inputs could be set to something like 8bits resolution? hope I have the terminology right. So I should be able to program that pin to look for a set voltage difference to indicate zero etc? have I got that correct?
 

Thats enough for one post. Cheers Str8

Hi Chris
How can I use your MESSAGE_TYPE = nonedialog in Gmoccapy 3.4.6? (LCNC 2.9.2, Debian 12)
Cheers
Wayne

Hi
I use 7i77 for feedback control with encoder motor signal succesfully
Now I try use feedback with linear scale for Y axis
I wiring  TTL signal from linear scale A signal, B signal to Encoder 03 slot (TB4 terminal)
I can open Gmoccapy but when JOG DRO of Y axis not change i think problem from linear scale

I have question ,TTL can work with 7i77 ?


 

So....
the machine is runnung linuxcnc 2.9.3 now. I use cutom tabs with probe basic.
and will tackle lubrication and toolchange next.

for lube any suggestion which component works best?
For my lathe I have one, but it kind of lubes to much. cause it runs every start and also the timer runs when idle. its only for the ways, so i didn't bother when there was to much oil.
On the Picomax I would prefer if the timer would keep its value even on restart and I need two instances one for the Spindle and one for the ways.
Is there a suitable component already done?

was unnecessary the timer is integrated in the bube pump by a mechanical reduction of the motor.

Thanks

No joy there, But it would not be hard to make a conversion plate to mount a 57mm/60mm servo or stepper if you did want to go with a Mesa solution.
You might be able to retain the stall homing if you write a custom homing module.

I got around to doing some videos on how my Maho works. Not quite the calibre of Mark's. But there may be stuff of interest to some of you. There is stuff about the switchgear, transformers, contactors, breakers, gear change, etc that will be common. Mine has some older stuff on it. There is also a run through some G-code in the older format that Heidenhain used. It's on my Google drive here:

drive.google.com/drive/folders/1NVXGKyaN...cWKjsMfe?usp=sharing

Cheers,
David

Not sure which version your image uses but there was a fix for this in gmoccapy that seems to have missed the release of 2.9.2:
github.com/LinuxCNC/linuxcnc/commit/760f...de07ea08b3f00b09b57f