Demystifying Spindle Sensors

Hello,
I am trying to convert a Denford Orac to LinuxCNC. I had it running ok on Mach3 until I tried to thread. Hence I am now trying to move it to LinuxCNC. The problem is that I cannot find a straight plain and simple tutorial on how to setup spindle sensors.

I will be using some of these:
i.ebayimg.com/images/g/2VgAAMXQLoZR24W-/s-l300.jpg
(building 3D printers, I have tons of these on stock)
which I hope they can pulse fast enough: 2000RPM x 100 slots.

I routed a simple disk with 100 holes that will attach to the back of the spindle bore. The disc has one longer hole that can be used as 1 pulse per revolution.

As I understand, I will need to mount 2 sensors in the 100 holes per rev area. I will also need to mount a 3rd sensor in the 1 hole per rev area. Wouldn't counting full 100 impulses on the other sensor be the same information as 1 pulse on this sensor (which would both equate to one revolution)?

100 hole is fine for an orac ,
i used 2 sensors on my orac , as index and the other as channel a , you dont need channel b , but if you have fitted one thats fine .
setup is straight forward , the complication comes in checking your rpm and scaling this to actual rpm etc .
it took me some time in adjusting the readout to be actual speed , with not too much runout when changing between forward and reverse when threading , so you dont end up hitting the shoulder etc .
i just used the origional orac sensors , but adjusted the pullup resistors to get a clean signal as i dropped the input down to 5v ( just made life easier ) .
what breakout board are you using ? ,

Well.. I see I spelled it wrong: I made a 50 holes disc, I was counting divisions of 360 degrees needed to get 50 holes.

Anyway, threading works fine. I only tried the sample included with LinuxCNC on a random piece of aluminium tubing:

and I got this:

I guess I needed a little bit more RPM to get a higher surface speed.

Do you still have the original (Swedish ?) motor? Does it lack power while threading as it is only 1/2HP? Does it get hot at low RPM as the fan at its back doesn't cool it properly?

I tried using the original Orac sensors, but i killed the board that was powering them while I was testing things out. I couldn't find any schematics for it, so I had to put them aside.

One thing I don't understand: you have 50 holes on phase A. Hence, you already know when you have a full revolution (50 holes have passed). Why would you need another index pulse to count revolutions?
100 holes on phase A means 2 revolutions, etc.

I used this breakout board:
img.alibaba.com/img/pb/382/733/750/750733382_553.jpg

Yes, I was looking through HAL scope and at the 2000RPM that the Orac can handle, I only have about 20 pulses out of the 50 holes. Is it my sensor, or the parallel port? Also, the wave becomes wider on the active pulse (high), while the width of the low part is very tiny. This even if the hole represent 3.6 degrees opening.

What I am stuck at, is the spindle RPM display. I added that panel that is supposed to show the RPM, but it is about 1.33 ratio lower then the expected RPM. Any particular part of the HAL file I should look at?

20000
(at the moment I am not at quadrature, just A and Z)

Regarding the spindle speed display - isn't it supposed to read it off the sensors? The displayed value is about 1.33 ratio smaller then the actual speed. That is even for low speeds.