Spindle rotation sensor

Hi all-

Having (finally) gotten my head around basic HAL connections, I now would like to work with the sensors that are in my spindle. The spindle has a pneumatically-controlled drawbar, and provides the following:

1. actuator in home position (ie. not in contact with the spindle, so OK to spin)
2. toolholder clamped
3. unclamped (ie. no toolholder present)
4. spindle stopped

Part one will involve using various logic components to let HAL decide if it's safe to actuate the drawbar and release the toolholder. (In other words, sensors #2 and #4 are both true). And if it's OK to start the spindle, etc. I'm pretty confident that I can manage the basic logic connections, but where I'm a bit stumped is with rotation sensor.

The rotation sensor generates a logic TRUE twice per revolution, I suppose with some sort of slotted disk/opto sensor. Anyway, the issue is that depending on the final resting position of the spindle, this sensor might be TRUE or FALSE. So what I need is something along the lines of:

"watch the spindle sensor for some length of time, and if its state hasn't changed then the spindle is stopped"

I'm guessing that a timedelay might be involved, but that's about as far as I've gotten...! I haven't been able to find a HAL component "out there" for this purpose, and to be honest I'm equal parts excited and terrified at the thought of creating one.

Any pointers would be greatly appreciated.

thanks,
Pat

Thanks, Andy! I was successful in adding an encoder, and I have a HAL meter displaying the speed of the spindle.

Now the question is: what lowpass filter will settle down the displayed speed?

I find that the reading jumps up and down (with about a 1s period) by an amount proportional to the speed...around +/- 500 at 5000 RPM...! Not exactly the accuracy that I was hoping for! I tried adding these lines to my hal file, but so far haven't seen any changes:

loadrt lowpass names=lowpass.spindle
addf lowpass.spindle servo-thread
setp lowpass.spindle.gain 0.06

I have read the lowpass man page, but I have to admit that I find sentences like this rather humbling:

The digital filter implemented is equivalent to a unity-gain continuous-time single-pole low-pass filter that is preceded by a zero-order-hold and sampled at a fixed period.

I have a rough idea what most of the phrases in it mean, but somehow I'm having trouble turning that into code!

Thanks again for all your help.

Patrick

Hi Patrick,

You have a accuracy about 10% on your spindle rpm's with the input of a encoder.
Are you using Mesa?

What kind of photophotonic encoder are your using?

Thank you both for the quick responses.

I tried a gain of 0.001, and it still jumps up and down about 500 RPM every second.

To answer the first question, I am using a 7i76, but not its encoder input. Instead I have the output of the spindle rotation sensor wired into a GPIO, and then I have that "HALwired" to an encoder component, set to counter mode. Maybe I should run the signal from the spindle into the 7i76 encoder input? Should I expect that to perform differently? And if I do, what signal is the Mesa encoder output called?

Whew...!

Patrick

ps - HALwired instead of HARDwired...! For some reason this really cracks me up. No doubt I've been banging my head on this for too long and I should probably give it a rest.

I don't know what is physically inside my spindle, but whatever it is it generates a logic TRUE twice per revolution. My guess is a disk with a hole in it and an optical sensor.

Patrick

That's a good question. I'll check and get back to you.