Reading and fixing Chinese scales

[irving2008]

The scale is (I think) just a PCB in (another guess) a sinusoidal shape. A source of the scale 'tape' would be VERY useful; It could be just stuck to the side of a mill table, for instance, and be read by a close fitting head (no bulky scale of PITA fixings).

The scale isnt a tape but a thin PCB with a lattice comb finger arrangement where the spacing follows a sinusoid...or at least thats one type, there's a nyumber of variations on the theme. Its clear that the PCB is manufactured in long lengths, I'm guessing ~32" then chopped up to suit. The trick I suppose is creating a long accurate dimensionally stable PCB.

[irving2008]

Is it possible they fit them to stainless because it is non-magnetic?

I remember, back in my days as digitizer tech support, the manufacturer accidentally printed a menu overlay for a capacitive tablet using conductive ink. Oops-Splatt :heehee:

I got the impression that the technology wasn't friendly if messed with.

Did you try insulating the scale then sticking a cap to ground to de-jitter it? They don't seem to bother much with decoupling.

Robin

There's about 1uF of inbuilt capacitor. The scale is powered all the time its just the display thats turned on and off, at least on these ones, so they read the same when you turn them on as when turned off... as long as you dont move them. I've seen a number of articles that suggest putting 100nF + 10uF when powered remotely and no battery in the compartment. I like Bill's idea of a local reader pcb and localising the power arrangement, that will get rid of many glitches. I've sketched a design using a small number of SMT parts that can be fitted to the existing reader PCB instead of the display case. I think a small bit of ali milled to suit will act as a suitable casing.

The existing case is definitely built down to a price. It took several attempts to get the display to work again after I dismantled it because the positioning of, and pressure on, the zebra connector seems to be critical. and the screw holes are marginal at best... in fact the 4 retaining screws for the PCB have all but one pulled through the pcb so i flowed a little solder in the holes to tighten them up.

I am debating whether to grind the calipery bits off or try to use them as a mount...

oh the other thought I had was making the PIC do calibration to improve accuracy over the long runs... some form of look up table of offsets per cm

and another thought: what about compensation for temperature? does the PCB expand enough to warrant it?

[BillTodd]

the other thought I had was making the PIC do calibration to improve accuracy over the long runs... some form of look up table of offsets per cm

Certainly do-able. Of course, it does require a calibration standard of must higher accuracy than the scales (thinks: I have a magnescale strapped to the side of my lathe that's good to a few microns over its entire length).

Though frankly, for the RF30 type mill I'm using them on, they're probably good enough as they are.

what about compensation for temperature? does the PCB expand enough to warrant it

If, as you suggest the pcb is a set of disconnected copper fingers then I guess it is compliant enough to flex with the SS carrier.

I am debating whether to grind the calipery bits off or try to use them as a mount..

They cut easily with a friction saw but watch the temperature,as you don't want the glue to come undone. I drilled mounting holes with a carbide pcb drill.

Were you thinking of making a stand-alone display device or using a PC?

[irving2008]

Certainly do-able. Of course, it does require a calibration standard of must higher accuracy than the scales (thinks: I have a magnescale strapped to the side of my lathe that's good to a few microns over its entire length).

Though frankly, for the RF30 type mill I'm using them on, they're probably good enough as they are.

Ditto for my MD30, the 12" claims 0.04mm across the full length, 0.01mm 0 - 100mm. Since nothing I make (to date) is bigger than 100mm....

Were you thinking of making a stand-alone display device or using a PC?

I was going to use 15 0.8" led 7-Segment displays and a couple of driver chips and a PIC, the displays are only 50p each and the PCB for it is simple enough. Doesn't warrant the hassle of a PC. Depending on the PIC I might add a USB port so a PC can read them too.

[BillTodd]

You might be able to save yourself some effort (I hate PCB layout) by using a EPOS display like this with serial input and large VFD

Or if you fancy tinkering...

http://cgi.ebay.co.uk/new-2-line-by-...item2c52e20dc2

[irving2008]

Hmmmm.... theres an idea...£4.99 isnt going to break the bank...

No doubt if I made it work somebody'd want to copy it


<edit>
Oh Google is a wonderful tool ... These are manufactured for Fujitsu (ICL as was) and are probably a RS232 serial interface (EPSON emulation) with a 12v 10W supply requirement. Shouldnt be too hard to figure it out. Only 9mm character height tho if these were the customer facing ones, 11mm otherwise..

[bogstandard]

Having played about with these for a few years on manual machinery, and spent grossly too much on them, attempting to get them to be stable, accurate and reliable, all I can say is ...

If they are going to be mounted anywhere they are liable to be close to swarf, dust or liquids, forget it.
They will be continually breaking down. I was lucky for a week to go by without having to strip one down and get it working again.

I have gone completely over to glass scales, both on lathe and mill, and the only place I use one of these is for upper z axis reading on the quill of my mill, well away from the things mentioned above.





I found that when running from an external supply, I played about with different values, and found I had to use a 22uf cap across the battery connections to prevent the display from jumping about to whatever reading it wanted to give, that was caused by interferance from the machine motor.



I bought a few of the spare cables (just the little rubber connector and 1mtr of cable to bare ends) about 5 years ago, and were about £1 each. From here, but they don't sell the cables any more, but they do sell the display boxes at a very reasonable price.

http://www.arceurotrade.co.uk/Catalo...gital-Readouts

My display was rigged by a friend to run from a wall wart, which got rid of battery power out of the display and read head.
To make the plugs more stable, a short length of toothpick or plastic rod under the four contact fingers keeps them much more rigid in the read head.

Been there, got the tee shirt AND baseball cap.
I wish you all the best in your efforts to cure the problems that have baffled a great many of us manual machinists for way too long.:cry:

Bogs

[irving2008]

Thanks for your input. I was planning to mount them underneath the table where they would be somewhat protected. It may be that its all a waste of time... but i like to experiment for myself

I have read several examples of people having success with these(or so it is claimed), sufficient for me to 'have a go' If it doesnt work its not cost me a lot...

[bogstandard]

I wasn't trying to put you off doing it (except to save your sanity), just giving you some hard earned pointers to be wary of.
If I think on tomorrow, I will see if I have some of those cables knocking about still, but don't hold your breath, I had a massive throw out about 18months ago when I rebuilt my shop. If I find any, I will PM you.

This might help you in the quest to protect your scales.

http://start-model-engineering.co.uk/workshop-news/

Enjoy your journey


Blogs

[irving2008]

Thanks for that. The idea is very similar to what I had in mind but as I dont plan to use the actual readouts I had already identified some upvc U- and L-channel which I was planning to use as the protective covering.