Advanced Search

Very cool project PedPEx! It looks like we had a lot of similar ideas for our interpolator boards!

The reason for the RJ45 pinout I use is that it allows for using standard ethernet cable, where the twisted pairs in the cable will be matched to each signal pair (SIN, COS, REF, PWR/GND). But in my design, I made a small polarity mistake, and the zero/reference signal polarity is opposite to the other signal pair. It would be better if pins 4 and 5 were swapped in my pinout. Signal-wise, it makes little difference, but looking at ethernet cable wire colors, the zero signal is the wrong way around compared to the rest.

And as to why I don't use M23 instead of R45. I agree that M23 is a very nice connector. And pricey. I don't exactly recall the prices, but certainly above 20€/connector. On the other hand, shielded ethernet jacks/plugs are not given away for free either. Another reason for me is also that I use normal cable glands for my (heavily modified) electrical cabinet. Gettting RJ45 connectors through cable glands is not all that nice, but easier than M23. Or you could use RJ45 "bulkhead" connectors instead of normal glands.

If I make another version of my board, I'm pretty sure I will change the design to use transimpedance amplifiers to convert the 11 µApp current signals to 1 Vpp signals. It would make measurements extremely much more convenient, and also low pass filtering easier. At the same time, the boards could be used with 1 Vpp scales by simply leaving the transimpedance amplifiers out (my current design can also use 1 Vpp scale). And I would most likely move over to surface mount for pretty much all components.

There does seem to be quite much interest for sin/cos interpolators.

In other news, I have my machine reassembled and up and running. I will post some more info about this once I find some time.

Hallo 
Bei mir Eckt es wieder mit meiner Linuxcnc . Ich wollte die Werkzeug Längenmessung Installlieren. Habe das Video von Talla genommen und alle Einstellungen gemacht aber leider kein Erfolg. Ein Kollege der nicht von mir weit Wohnt hat es auch versucht auch ohne Erfolg. Danach hat er das Gmoccapy Innstalliert ist zuerst auch nicht gegangen da immer wieder Achs Positsion Fehler gekommen sind. Nach wechsel des Netzteil (hatte 500mV) auf 2 A ist der Fehler nicht gekommen Wechsel wird durch geführt. Nur hätte ich gerne meine Oberfläche (Probebasic) was muß man ändern? Es gibt bei diesen Programm schon eine Einrichtung für die Werkzeug Vermessung, nur scheint diese nicht Freigeschaltet zu sein. Ein Viedeo gibt es auch nur habe ich keine Einstellungen HAL und INI gesehen. Kann mir eine Einstellung der Der Dateien zu Vefügung stellen?
Für eure Hilfe wäre ich Dankbar.
Gruß Hans

Upload the pictures on the forum, not on 3rd party sites as they may (and will) contain mallware/tracking stuff.
Thank you.
-
I do have, new newer used THK rails sitting in the shop, namely:
2 of 780mm long 35mm wide rails with 4 of long and narrow carriages, roughly 150mm long
2 of 360mm long 30mm wide rails with 2 of short and narrow carriages
More info if you think you can use them, and might be able to ship to you at the end of August.
I do have 2 more of used THK45 rails and 2 carriages 1 long and short, rails one is 1250mm and the other should be around 700mm. Pretty rough condition but no play.

Hello,

I'm on LinuxCNC 2.9.4 on Debian. I want to invert the jogging sense of the X axis on a lathe, so that when I rotate the pendant wheel clockwise, the X axis moves in the X- direction. I just changed this in the vc-p1as.hal:

net jog.0.cont-minus  vc-p1as.jog.0.minus  => halui.axis.x.minus halui.joint.0.plus
net jog.0.cont-plus vc-p1as.jog.0.plus => halui.axis.x.plus halui.joint.0.minus

(On the first line I changed to .plus at the end and in the second I put .minus at the end)

This works only for the Velocity mode. In Step mode, the clockwise rotation moves the axis in the X+ direction.

How can I change the X axis direction in Step mode?

One year ago the X axis was moving in the direction I wanted in both Step and Velocity modes (probably on 2.9.2, I don't know anymore). I don't think this is related, but I moved from a parallel port to a mesa card. With only a very basic knowledge of the HAL, I commented out the halui. signal section in the taig_lathe_v2.hal file to make the pendant work.

I'm attaching the relevant files.

Thanks

7I76EU pin names are identical to 7I76E names

Can you post your current hal/ini files?

I wanna note: It looks kinda professionally done, but its not. I'm in the luck of having a father in law that owns the place (he is a mason by profession) and i can use the place and the incredible useful overhead crane. Without that it would be impossible. For other projects i used the good old engine hoist in my garage, but for this machine it is just to small.

I will update as soon as there is any progress done.

1. Need to solve the guide rail problem
2. Ways need to be reground and scraped. Including the gibs.
3. Ballsrews need to be serviced
4. Milling spindle needs to be repaired
5. CLEANING

Yeah, sorry about that, made a quick change in the V2 version, and had a typo
that broke 7I96 input 0

EDIT:

Note that the 7i96sd_mpgp1 firmware has a different pinout on P1 and not a close match
to the common wired MPGs. I just fixed and re-uploaded the 7i96sd_inmp1v2 firmware
if that's more appropriate (and the  7i96sd_inmp1v1 firmware was always OK )

 

Hallo liebes Forum,

Ich habe die Lösung gefunden. Ich habe meine 7i96S mit dem file 7i96sd_mpgp1.bin von der mesa homepage neu geflasht.
Jetzt funktioniert alles wieder wie erwartet. Mein Ziel ist es mein Handrad auf P1 anszuschliessen, das müsste mit diesem file auch gehen.

Vielen -dank

Hello PCW,

I found a solution:
I flashed my 7i96s with the 7i96sd_mpgp1.bi file from the mesa homepage. Now everything works as expected.
My goal is to wire my MPG to P1, I think, this bin-file is for realizing this too.

Thanks a lot

Update:

So i took the machine mostly apart.

It is really dirty. I don't know what the pre-owners did with it, but dried up coolant is basically everywhere. Its yukky. It takes a lot of time to clean everything.

Linear guide rails on the Z Axis -> Scrap ( 2 out of 5 blocks are wiggling in the Rail) The rail is some special THK rail and is only sold as a set by Mazak. I requested a counter offer from a THK dealer... even more expensive.

The Y axis box ways are done. The lubrication system malfunctioned and the way wipers failed over the years. Already brought it in for regrinding. After that it needs to be scraped, which i am able to do. I lack a lot of experience in scrapping but its a "simple" box way and no dove tail ways.

Same goes basically for the X axis. Lube system got plugged up by failing way wipers and even the hardened ways got damaged. -> Regrinding.

Ballscrews:

Z Axis ballscrew is in good condition, but has no preload anymore.
X Axis ballscrew is in good condition, and still has preload.
Y Axis ballscrew feels a little bit crunchy... the balls need the be replaced and checked again.

Milling spindle:

The taper is shot and needs to be reground. I assume the bearings are also in no good condition. This repair also needs the be done professionally. 

The good news:

I fixed my temporary encoder capacitor solution with a proper super cap. I even upgraded it to a 1.5F

Good Morning,

I've installed Probe Basic and continue to have an issue with parameters not being written to the .var file. Anytime a value is updated via Probe_basic ui, a "File not Open" error shows and no changes are made to the .var file. Parameters entered via Gcodes work fine. For instance, SET TOOL TOUCH OFF POS generates "File Not Open" error, but MDI g30.1 works fine. The sim version works fine. 

Terminal shows: emctask.cc 68: interp_error

I am running:
Debian 12 Bookworm
LCNC v2.10
Current version of Probe_basic and QtPyVCP
installed sable APT version 

Any help would be greatly appreciated. 

Thank you,
Sean