Rotary table digital readout

[petetwissellParticipant @petetwissell]

I have one of the 6″ rotary tables of generic design which can be mounted on edge as a dividing head.
For various reasons, it would be advantageous for me to have a digital readout for the table angle.
I had expected to find suitable kits readily available, as they are for linear readouts – but nothing came up in my searches.
I found a rotary encoder with remote display available, but the encoder is a 37mm high box which would require lifting the rotary table high off the mill bed and reducing available part and tool height capacity.

So – I had a look around to see what I could adapt. What I found was an inexpensive “digital level / angle finder”, the small box type, with a tilting display – various brands available…

This was actually something I had for several years, but it never worked correctly after I dropped it. With nothing to lose, I set about dismantling it.
Inside, the ‘heart’ of the device consists of a pendulum and an inductive encoder mounted on a pair of PCBs.
The front half of the box houses the two AA batteries and the display, which is connected to the main PCB by a short ribbon cable. In the photo I cave cut the ribbon cable and battery wires to seperate the encoder board out.

The pendulum is attached to the encoder rotor with an M3 nut (on the back side) and is easily removed.

Mechanical installation to the rotary table requires a simple aluminium ‘slug’ to press into the bottom end of the table’s hollow spindle, with an M3 hole tapped centrally.
To mount the PCB I cut a disc from some 2mm rigid plastic sheet (I happened to have some clear sheet, which is exceedingly brittle!). The disc is cut out to clear the smaller of the sandwiched pair of PCBs and holes drilled to mount the larger PCB with M2 screws and nuts.

The resulting assembly protrudes around 4mm from the underside of the rotary table, so I am making a simple rectangular aluminium sub-plate, recessed where required to clear the encoder PCB and to create a channel for the ribbon cable connecting to the display. The display will also be mounted on the sub-plate.

The origianl ribbon cable has 8 conductors connecting the encoder PCB to the display. I replaced this with a 10 way ribbon cable, which also carries the connections from the batteries in the display box to the encoder PCB.

The on/off switch is mounted on the encoder PCB. I attempted to create a remote switch, but without success. I suspect the switch on the PCB in not simply a normally open ‘push to make’ type.
I have decided to go for a slightly Heath-Robinson solution, with a pushrod in a groove in the sub-plate to operate the switch.

I’ll post more photos soon. I just need to machine the sub-plate.

[Russell EberhardtParticipant @russelleberhardt48058]

Interesting project Pete.  The 0.1 degree resolution might be a bit limiting though depending on your requirements.  The calibration on the handwheel is at least an order of magnitude better although what the absolute accuracy is I don’t know.

Russell

[petetwissellParticipant @petetwissell]

Hi Russell,

For most of my purposes, 0.1 degrees is sufficiently accurate. I will still be able to use the handwheel dial for ‘fine tuning’ if necessary, but the readout will give me the ‘ballpark’ position and gives the facitilty to zero the display at any point.

Pete.

[petetwissellParticipant @petetwissell]

This is the arrangement of the encoder PCB, mounted the my plastic disc and connected to the half of the original box which houses the display and batteries. The whole lot is sat on the aluminium base plate.

The underside of the rotary table with aluminium plug, tapped M3 for mounting the encoder rotor.

And the finished article.

Under the display can be seen the end of the pushrod which operates the on/off/zero button on the encoder PCB.

[petetwissellParticipant @petetwissell]

I’ve just been running some tests.

At first, I thought something was wrong with the encoder or its installation.

For the first 100 degrees or so of rotation, the readout aligned perfectly to the handwheel dial.

Over the next few turns of the handwheel, the readings drifted apart until they were 1.5 degrees different at the 180 degree mark.

Continuing round back to the start position, the reading came back into alignment.

I repeated the test, but this time used an additional means of measurement – a digital protractor – to see what was going wrong.

As it turns out, my readout is correct, but there is a lot of wear in one area of the worm wheel in the rotary table.

This discovery may explain some issues I have had in the past.

Although the readout is accurate to only 0.1 degrees, it is a lot better than the dial reading through the heavily worn gear.

[petetwissell]

On petetwissell Said:

…Although the readout is accurate to only 0.1 degrees, it is a lot better than the dial reading through the heavily worn gear.

An important observation!  Fitting a DRO to my mill transformed it by eliminating human error, but, because it detects actual position, it also compensates for wear and maladjustments, a very good thing.

I’ve not bothered to fit a DRO to my lathe because the screws and nuts are in good condition, but they will wear eventually.   Then a DRO makes sense!   (There are other good reasons for putting a DRO on a lathe, that they don’t apply in my workshop is just me.)

I have a rotary table stepper project that relies on the worm and gears being in good nick.   Your Wixey idea could improve it. By detecting when the worm isn’t accurate, the stepper driver could be programmed to compensate.

By the by I have a jaded opinion about second-hand rotary tables!   Condition is everything and some of them had a hard life.  Even in the best, the worm and gears are rather delicate, so if the table is used to hack metal quickly, they are likely to be worn.   A few years ago I found a youtube where a chap bought a big Chinese rotary table to cut traction engine wheels.  Didn’t last long!  Assuming the problem was it being made of Cheesium, he bought an expensive western made table, which lasted only slightly longer.  An expensive mistake: neither table was designed for heavy cutting.

Industry don’t expect tools to last!  My lathe book suggests 3 to 5 years before hard working lathes are replaced “beyond economic repair”.  A rotary table of the best make might only last a few months if thrashed.   Tool condition depends on usage, anything between ‘the need justifies the end’ brutality and occasional light cutting done by a skilled operator.  The second is an ebay bargain, the first isn’t…

Dave

[Clive FosterParticipant @clivefoster55965]

The pull wire sensors work well in this sort of application when run in a shallow slightly helical groove to avoid overlap at the end. Scaling is a potential issue but one of the modern programmable microcontroller beasts will handle that just fine.

I believe a design for a pull wire sensor was published in ME or MEW some years ago.

I wonder whether the flexible tape type magnetic “scales” could be made to work. Obviously two, suitably offset, reader heads will be needed due to the inevitable gap at the end. However carefully the junction is cut its never going to be insignificant at this level of accuracy. So long as the basic drum radius is adequately large the system should cope with the tape being on a curved surface rather than flat.

Clive

[Les Jones 1Participant @lesjones1]

A few years ago I did some work on connecting a remote readout to a wixey angle gauge. Here is a link to what I found out about them.

http://lesjhobbies.weebly.com/hacking-the-wixey-type-angle-gauge.html

There may be some information that is useful in this link.

Les.

[John HaineParticipant @johnhaine32865]

Some time back on the “old” forum Joe Noci described using some magnetic sensing chips for rotary measurements. Analog Devices also make chips that can support precise angle measurement.

[petetwissellParticipant @petetwissell]

Thanks for all the pointers.

There are, of course, many ways to achieve digital angle readout and I could have gone for a higher resolution industrial encoder and written my own software.

However, I wanted to create something using readily available, low cost parts which could be replicated by anyone with no special skills required.

 

[Author]

Posts