Spindle index with proximity sensor = bad results

More
07 Feb 2019 08:19 #125910 by emcPT
I have several machines running linuxcnc, where all have spindle motors with encoders or encoders directly mounted in the spindle. I never had a problem with synchronised motion or threading.
My last retrofit have a proximity sensor on the spindle, where it is ON about 20% of a turn. I have seem posts where users use proximity switches, so I suppose they are suitable for this propose.

Machine is fully working except threading as the thread does not start always on the same position - it can start almost half turn away.

I have the proximity switch connected to a encoder index input on a 7i77 using non differential mode (one wire only). The idea that I have is that the proximity switch is slow or maybe because the ON status is larger than what is needed?

At the same time I think that it is only important the rise of the signal, but I can be wrong here.
Attachments:

Please Log in or Create an account to join the conversation.

More
07 Feb 2019 09:46 #125912 by emcPT
Some update.... it looks like the proximity switch is not constant... need to investigate more.

Please Log in or Create an account to join the conversation.

More
07 Feb 2019 10:21 #125914 by rodw
In some tests I did on prox switches mounted as home and limit switches, the repeatability was +- 0.3mm with the sensor passing over the target. I did hear that the prox sensors are more accurate if they approach the target head on but in my case they would crash while coasting to a stop.

So they may not be the best shaft sensor.

There are some nice looking encoders that are designed for the spindle to pass through which I had in the back of my mind if I ever built a lathe.

Please Log in or Create an account to join the conversation.

More
07 Feb 2019 17:20 #125935 by Mike_Eitel

Please Log in or Create an account to join the conversation.

More
07 Feb 2019 17:48 #125936 by emcPT
The problem was in my electronics. I used a voltage divider to get from 24V to 5V with 2.9K and 1K resistors. When the sensor was "off" instead of having the desired 0V on the input of the Mesa, I was getting about 1.6 V that gave false positives. I think because the 7i77 have some pull up scheme. Adjusting the voltage divider, no false positives, so all is working ok now.
The following user(s) said Thank You: rodw

Please Log in or Create an account to join the conversation.

More
07 Feb 2019 19:34 #125946 by PCW
The 7I77 has 2K pullups on its encoder inputs so you need to have a low value divider (or use a NPN prox and the built in pullup to 5V)

Please Log in or Create an account to join the conversation.

Time to create page: 0.408 seconds
Powered by Kunena Forum