Adding a rotary encoder to a large Rotary Table

Hi all,
I need to add a degrees readout to a large rotary table.
The table is around 1 meter in diameter, and I can only add the sensor to the outer rim, kind of like a trundle wheel.
My initial idea was to just setup a rotary DRO with a rubber wheel running on the periphery, but then I though what about the gearing?
The table is so large that I cant have a 1:1 wheel, so how to do it?

I dont mind doing a bit of electronic jiggery pokery, but Ive not done anything like this before...

all ideas and help gratefully recieved :)

Dave

There must be some kind of mechanism that turns the table? An encoder could be fitted to this, the only problem is that backlash may exist.
I recently fitted a servo to a 1.5m dia. air bearing table, in this case there was an adjustment in the gearing to take up any backlash.
A Maple Systems 2 line display was used for the readout.
Al.

Forgot to mention that I cant modify any of the drive train, it is manually operated, and interlocked with other items.
Currently the positioning is done by bloke counting to 10 then letting go of the button, and this adds a certain amount of variability into the process (some what understated...)
The positioning doesnt have to be accurate, as long as we know where it has got to IYSWIM.



Dave

Can you wrap a cog belt around the outside, and use it like a rack to drive an encoder?

What does the drive train consist of exactly, rough description?
If there was any milling or something that resembled a gear tooth or rack gear on the underside etc, a proximity sensor could be used?
Al.

What are the target resolution and accuracy?

Is continuous rotation required?

Ken

You can wrap the table in magnetic scales (the kind that is like tape) and use a standard magnetic dro sensor to pick up the ticks. You'll get very good resolution with this method.

And if you are machine ferrous material you will get a million whiskers.

The Hall effect sensor would be the better choice, the sensor material just has to be ferrous metal, you can even get non-ferrous type also.
Al.

Another couple of options is if an encoder is used, rubber friction wheels specially for encoder drive can be obtained, if the ratio was worked out correctly, you could use the marker pulse as a 1 pulse/degree counter, the other method may be a bit tedious and would entail drill and tap for a Allen head machine screw every degree and use a proximity switch as a counter.
Also it has to be designed as to whether the table rotates 1 turn or continuous, the electronics would has to take account of this.
Al.

Encoder Products Company offers encoders with a rubber wheel and a spring loaded mount which would work well for this application. I use one to keep track of linear feet on a winding application. If you use a counter/timer readout you can use its scaling functions to divide the encoder down so that it reads in degrees.

I am using this encoder: Model TR1 Tru-Trac

And this counter: Rate/Total/Batch Meters

I think that would get you pretty close to a working solution.

Matt

Lots of good info so far, thanks :)
I started typing this last night, a bit more detail, not sure why it didnt submit...

The Table top is 25mm thick, 1.2 meters in diameter, and plain disk on both faces. Its mild steel, with a thin stainless steel cover over it.
The drive system is sealed inside the housing, and I have no access to it, but its a DC motor and I presume a worm/wheel affair.
Wrapping a toothed belt round might work, but still has the gearing problem of 1 rotation of the table is many of the encoder spindle. I dont think it would be much better than a rubber contact wheel?

The target resolution is better than 1 degree, 0.1 degrees would be ideal.
The table can only rotate in a clockwise direction. I need to be able to cope with that without getting to a travel limit on the encoder, as it can never 'unwind' if you see what I mean.

Dave

Dave

If you use a counter for your readout, like I linked to, you can scale the encoder pulses so that 478 encoder pulses (for example) will display 1 (degree) on the readout. You can have a decimal point in there too so it would be accurate to a tenth of a degree. This counter will generally keep counting higher and higher. You could use a single hall effect sensor as your reference mark and the signal from this could reset the counter every rotation. When the hall effect sensor was almost activated, the readout would show 360 degrees and then it would be reset to 0 degrees once you hit the sensor. This would provide an absolute angle measurement and would limit cumulative error to that found in a single revolution.

Matt

with that diameter, the epc or equiv rubber wheel encoder sounds perfect. what u dont seem to grasp is that u can buy a display that scales the encoder pulses to display any units you want. so about any encoder rubber wheel thing (say 1000 or 1024ppr) will work as long as ur display is scalable. newall in UK sells a SA100 simple display that is scalable, so do folks like EPC, danaher controls, and red lion. $ 300 for encoder with rubber wheel plus $ 200 for display and u r good to go. if u need specific part nos & prices as examples pm me and I will send or just google it all.

Mike,
cheers, Ive googled the SA100, and will give newall a call after the weekeknd :)

I think the rubber encoder is the way to go, and Ive found a uk source.

I had a feeling this should be simple, I was just out of my depth. :drowning:

Dave