pyvcp tachometer

Is it possible to have tachometer with only an index pulse ?
ie: the spindle control is manual.

A pulse can be seen on pin 10 but I'm unsure how to make the software connections
or whether this is even practical using only parallel port.

btw: the spindle has a belt driven quad encoder fed to a DOS program but a 1 ppr disk is being
added. I believe 1 ppr will be needed for future thread cutting ?

Hi

You can certainly have a spindle rpm meter with just an index signal.
That is all you need.

This is pretty much what you need in your hal for the encoder

and this in your custom_gui.hal for the RPM meter

This index signal will be fine for threading too, but you will require a separate pulse (encoder A channel) to thread properly.

I have a hall effect switch for my index and an optical switch and 120 hole disc for pulse on my lathe, which works fine.

(only Mach tries to kid people that you can thread with just one reference per revolution)

regards

Yes that worked. Thank you.

I'm thinking of ditching the belt driven encoder and going with the more common multi-hole disk.

2 questions:
What's realistic with software counting ? 100 holes ? Does it matter ?
Excluding the disk, are there bolt-on electronic kits available or do most people roll their own ?

(okay that's more than 2 questions)

Hi

I'm thinking of ditching the belt driven encoder and going with the more common multi-hole disk.

If you have a working belt driven encoder, I would keep it if possible.

I had one but it was fried, along with many other components in a power outrage at some point before i bought it.
Thats the reason I went to other indexing methods

This is my set up, as posted previously.
www.linuxcnc.org/index.php/english/compo...imit=6&start=6#14832

What's realistic with software counting ? 100 holes ? Does it matter ?
Excluding the disk, are there bolt-on electronic kits available or do most people roll their own ?

There are no kits I am aware of, but the components are simple and quite easy to put together.
This is the circuit I use, kindly supplied by Andy Pugh some years ago.



I have a 120 hole disc, which is approaching the limit of practicality in terms of disc diameter and the number of holes that can be drilled along its circumference, for the space available to me. The holes are 3mm and 3mm ish spacing between.
I chose to work in multiples of 60 for sheer convenience of calculation, both with spindle velocity and the initial PCD calculations for making it.

I also hypothesised that if all calculations within software as to the exact position of the spindle were in numbers that divided exactly into 360, the accuracy must be better.
Whether this is true I don't know, but it threads very smoothly.

regards

i'd go with ArcEye's suggestion , iv'e used various sizes of disks , from 48 hole to 120 , the 120 hole does appear to work better , or at least shows smoother running

ArcEye wrote

If you have a working belt driven encoder, I would keep it if possible.

I'll give it a try.

The manual page for encoder reads "phase B is ignored" for unidirectional threading with software counter.
So only index and phase A need to be connected in Hal ? In what situation would phase B be used ?