Spindle Speed Encoder

[tblough]

Centroid control is looking for differential (RS422) quadrature encoding. That will be at least 6 wires + 2 more for the index.

Two wires will be for power and probably 3 more for A/B/Index if it's single ended. That's easily converted to differential output with a cheap IC. However with just a single row of holes, what you have is probably not quadrature output, but a simple pulse and that won't work with the Centroid encoder inputs.

I'm guessing you have two pairs or wires for two led light sources, and two pairs of wires for two photodetectors - one for speed sense and one for the index.

[Threedj16]

Yup, I figured this encoder wouldn't work for the acorn. But it is ashame to tear it out. For it's time, this was pretty advanced on a tiny lathe. Looks like maybe 100 pulses and one index per rev. Pretty funny though, as this would blow away what most people still use with Mach 3.....LOL.

[Muzzer]

I have the same issue, as I posted at the weekend (no reply yet).

There are differential line drivers that should do the trick like this: http://www.shop.cncdrive.com/index.php?productID=259.

Datasheet here: http://www.cncdrive.com/downloads/Diff_ ... manual.pdf

Are there other alternatives that Centroid could recommend?

[martyscncgarage]

Nope, you need to refit the spindle with a timing belt and pulleys running 1:1 and use a standard line driver encoder if you want accuracy when threading, tapping and Constant Surface Speed.

You are here because you wanted a better control, take the time to do it right, and as all the industrial CNC machines do. Standard line driver, differential encoder. Centroid recommends a 2000 line encoder on the spindle, in quadrature, that is 8000 counts per revolution.

Marty

[Muzzer]

Hmm. You wouldn't be standing in front of my machine with a belt and pulley in your hand for long. The head is a solid cast iron thing, full of gears with a large power drawbar on top, so not much chance there I'm afraid. It was a major job just to get sensors onto the bull gear. If I can't generate a decent signal from what I've got, then I won't have a spindle encoder signal, as simple as that.

8000 counts is surely way more than you should need for rigid tapping, even if it is a well proven solution. I've got 48 teeth, so would end up with 192 counts. Anyway, as I don't have much of a choice I'll let you know how I get on!

Murray

[martyscncgarage]

Just got the new lathe in, man it's beautiful for it's age. Anyway, anyone have an idea about the usability of this old optical encoder? Looks pretty state of the art for the 1980s. It has at least 4 wires coming off, possibly 2 more for power (hard to tell until i can take it all apart).

I think the part with open slots were for the zero index. They don't look broken out.

Thanks!

20180302_175238.jpg
20180302_175249.jpg

Suggest you start your own thread in the ACORN FORUM (TOP OF MAIN PAGE)

No, your optical sensor will not work if you want reliable and accurate spindle control.
Must use standard differential line driver encoder. 2000 count recommended (8000 count in quadrature)

You are using industrial controls, likely because you want that performance. You will need to use the proper devices to ensure optimal performance of your machine. You'll be glad you did.

Marty

[Threedj16]

Yup, already have one picked out from Automation Direct which meets all criteria and started my own thread.

Thanks