Gallery, Projects and General > The Design Shop

DIY DRO for lathe or mill

(1/5) > >>

AdeV:
A week or so ago, a shiny new DRO unexpectedly(ish) arrived at my workshop.... OK, not that unexpected really; it's for my old Edgwick lathe. It's also not the story for this thread - I'll document the fitting, and some of the hoops I had to jump through, elsewhere, when it's closer to being done.

My original plan, before the Chinese manufacturer (Ditron - via AliExpress - good crowd, based on my experience) heavily discounted the DRO readout for me, was to just buy the scales, and use an Arduino to read them. Initially, I'll be using the supplied DRO head, which is actually surprisingly usable.... but long term I still want to make my own, with the features I want in it...

Also - due to impending arrival of first child - I'm about to lose most of my workshop access privileges (and time!), which means much more time sat around at home with a relative paucity of tools... so I plan to fall back to plan "B" and use my software skillz to write the DRO head software for the DRO I want. And it also occurred to me, that you, gentle reader, might have some ideas of what makes a good DRO for you that I haven't thought of....

So, the overall plan looks like this:
 - Each axis will be connected to a cheap Arduino clone (Mini Pro, probably, as I have a bag full of them). One per scale, because with the 1 micron scales I've bought, they can't reliably keep up with more than one - and also, the system needs to track both A & B channels fully, in order to get the 1 micron resolution.
- Each axis will communicate its scale position to a Raspberry Pi, which will be running the DRO software
- A 7" touchscreen will provide the primary interface.
- I may also add some real buttons, if it turns out that regularly jabbing a touch screen with oily fingers doesn't work too well...

So far, I've "built" (the word is massively over-used in this regard) a one-scale reader, which appears to work fine. Yes, you can trick it if you move the scale too quickly.... the only part of a normal lathe/mill I can see this potentially being a problem is on a quill-mounted DRO, where it's possible to move it faster (I think - haven't tried) than the Arduino can count. I'm wondering about the carriage too, I can move that pretty quickly on the lathe. If it is an issue, there are faster boards out there than the Arduino.

Anyway.... with the basic principle established, now's the time to start designing the features. So far, I have:

Readout:

- Obviously, the basic "N" axis readouts are essential. In addition to the usual switchable decimal inches/metric; I think I could display both simultaneously (primary and secondary, secondary being smaller). Not sure how useful that would be.

- For our American friends, I wonder, would an inch fractional readout be useful? Maybe down to 64ths, with the secondary readout showing the usual thousands or tenths.

- Ability to re-name/move around each axis in software. So if you don't like the conventional "Z-Y" on a lathe, you can rename them "Carriage" and "Cross-slide", for example. Or "Tailstock". Or whatever Tailstock is in your first language.

- Nice big numbers that are easy to read. Actually, the Ditron scale is excellent in that regard, I will shamelessly copy it. But also allow the user to configure the colours.

- All-axes reset from a single button (with undo). Again, the Ditron partly does that (you can undo a zero press), but to zero all four axes requires four button presses. And really, who has the time for that, eh?  :scratch:


Features:

Some ideas below for things the DRO can do.

- The usual Chinese DRO functions: Bolt pattern, 1/2 function being the two that instantly spring to mind.

- The ability to combine two axes. e.g. on a mill, the Knee/column "Z" axis, could be paired with the quill "z" axis, to give a unified read. So now, move either the table/column OR the quill, and your "Z" reading moves as appropriate. On a lathe, you could combine the main carriage + compound slide axes so you can accurately bore to a certain depth using the compound, even if you accidentally moved the carriage last time you measured it (don't ask me why I bring THAT specific example up!). It also occurs to me that, with a smidgeon of trigonometry, you could accurately cut a depth even if the compound is at an angle, if you can type that angle in somewhere. Or have a rotary encoder on the compound slide too! :D

- Targeting with audio alert: By which I mean... let's say you want to bore a hole exactly 1" deep. Rather than watch the DRO, wouldn't it be nice if it beeped at you as you got close to your target depth; the beeps getting faster & eventually becoming a solid tone as you reached depth. Almost exactly the same as reversing sensors work on cars... Just means you don't need to keep watching the DRO instead of the workpiece.

- Switchable diameter/radius mode (for lathe cross-slides). No more having to remember to half the depth of cut to hit a specific diameter.

- Tool library (I think most Chinese DROs implement this in some form. I've watched This Old Tony going through his DRO functions, and I reckon I can make it an order of magnitude easier to use). This won't be much use on my lathe, as my toolpost is not particularly repeatable. Plus I don't have anything like enough tool holders.

- Angular readout: If I ever find a cheap dividing head on eBay (or maybe make one?) that I can experiment with fitting a rotary encoder to... it would make gear cutting a more straightforward operation. That said, if one is going to that much trouble, maybe it'd be easier to just make an "cnc-esque" dividing head with a stepper motor & remote control. Still, it'd be cool if the DRO could control it, wouldn't it?

That's all I can think of right now... but I should have a decent amount of time to incorporate any scope-creep that might come along; and if there's any feature you can think of that I could sensibly add.... feel free to suggest it.

I'm also happy, once the basic principle is proven out & the software is at a point where it could be considered useful, to build a DRO unit for any madmodder, for the basic cost of the parts & postage. i.e. no labour & no fee for the s/w, if anyone is interested? Note that at this time, I have pretty much zero idea how much the parts will cost! Although I do expect the screen to be the most expensive bit, at around £45.

PS: I'll also do a proper (and I mean, proper) write up of the plans and designs for all the parts (except the scales themselves), so you could build your own from the plans if you preferred). The S/W I will make freely available under one of the open source licences.

WeldingRod:
Being able to type in a specific number is super valuable!  I do it all the time after I measure something on the lathe.
I can't see any machinist being interested in fractional readout.
Putting a scale on a cross slide is quite a trick...  I did manage it on my old lathe :-)  angular encoding would be even more miraculous.  I could imagine a super thin pcb pattern glued to the carriage with a caliper type capacitive reader.  Hmmmmm
Saving the position on power loss is a perk!

I could imagine either a visual bar graph approach indicator and/or a separate flashing light.  Maybe a button for "I'm going back here a bunch of times".  Maybe three separate "here" locations?  Choice of approach direction or default both ways?

I like your list!

Sent from my SM-G892A using Tapatalk

awemawson:
I’ve had DROs on my Bridgeport for ages from BM Electronics that use a stainless wire wrapped round a spring loaded drum passing over a spool with an encoder. All this enclosed in a small die cast box.

 One of the calibrations they offered was for angular measurement by wrapping the external end of the wire round (say) the degree drum of the top slide.

I have to say I have been very impressed with them, especially as they were the cheapest DRO that I could buy at the time , this was before the Far East flooded the market.

BillTodd:
Not to put you off but... (BTDT using VB5 and my own microchip PIC interfaces)

Have a look at Touch DRO for andriod (works fine even on old tablets like HUDL etc) - it uses bluetooth or serial to connect to the scale interface devices (arduino etc.) and will save you a whole pile of grief writing software (at least until you see exactly what you want from the DRO)

For a Pi  you might want to look at Python and Kivy - I'm currently using it for a automated test system project , and once you get your head around the thing , it works nicely (doesn't require the Debian desktop overhead - see picture below)

One thing I really liked on my VB5 DRO  that I haven't seen on other DROs is analogue displays of fractional inches (and simultaneous display of inch and metric) . If you work in both systems it's so much easier than switching scales.

Sea.dog:
I’ve had DROs on my Bridgeport for ages from BM Electronics that use a stainless wire wrapped round a spring loaded drum passing over a spool with an encoder.

That's interesting, Andrew. I was considering something along the same lines when trying to work out how best to fit a DRO to my Boxford. What worried me about that solution was the effects of temperature change on the accuracy.

Navigation

[0] Message Index

[#] Next page