Electrinic Digital indicator

[Phil]

Well after looking at Televue prices of there digital indicator for there focuses i was shock at the cost. The indicator and cable would set me back a shocking price of $1300. The data cable itself is $369 for a cable. Now i think this is a bloody rip off. So i got online and work out i could do the same thing for get this about $200. I found these guys really good http://www.tresnainstrument.com/digital_indicators.html?gclid=CMGHr sXSoJcCFQsQagod7gTkJA
So i now have a digital read out on my focuser and comuter for allot less then it would cost from the shops that sell these things for televue. The resolution is 0.001mm/0.00005"

[Phil]

more photos

[bmitchell82]

MINT JOB :D I just love it when people "stick it to the man"

ISS SAYS NO TO RIP OFF'S

yay for us:eyepop:

[Phil]

Good on you Brendan. Nice to get a reply.
Phil

[mick pinner]

thanks Phil, l found a local shop selling these units and ordered one. a great idea.

[bojan]

Excellent work.
And definitely NO to ripoffs :-)
However, do you really need 0.001mm resolution?
Electronic caliper with 0.05mm costs around $50 or so. And it has RS232 (interface must be added though but I think this is just low-voltage RS232 level converter)
Also I saw mechanical gauges (looks similar but no electronics, just gears inside and round dial with pointer under the transparent cover), for similar price.

[AndrewJ]

Gday Bojan



Most of the cheap calipers with RS232 out use a funny Chinese protocol and arent "simple" RS232 at set bauds ( which is a pain ) ie
http://www.shumatech.com/support/chinese_scales.htm
I have hacked into a std vernier caliper, and can read it with my PC via a parallel port logic analyser, but its not real practical for a focusser. A dial unit with RS232 out may be the go if it uses the same protocols.

Andrew

[wasyoungonce]

.001m = 1mm..which isn't that accurate in terms of this.

;)

Well ...someone had to say it...:D

I know..I'm just being a smart arse!:whistle:

[bojan]

Sorry :-)

Andrew, fair enough :-) and thanks for the very interesting link, I did not know about this.
Anyway, even $200 is excellent compared to k$ "normally" asked for...

[Karls48]

Here is the link to schematic and code to implement RS232 connection to cheep Chinese digital calipers http://www.compendiumarcana.com/caliper/ (http://www.compendiumarcana.com/caliper/)
I did not try to build it myself yet.

[bojan]

Karl,
Now THAT is wonderful :-)
Thank you... :-)

[Merlin66]

I've got a few old Mitutoyo micrometer heads lying around but never thought of connecting them to an RS232.
Do you know the part# of the small 4 pin(?) connector used, also what's the software to display the readings?

[bojan]

Quick googling returned this.. it seems those things are plentiful:

http://zone.ni.com/devzone/cda/epd/p/id/4751
http://www.cnczone.com/forums/showthread.php?p=474310
http://www.microridge.com/gageway_sm_series.htm
http://www.microridge.com/gageway_mx.htm
http://us.100y.com.tw/PNoInfo/44232.htm

[Merlin66]

Bojan,
I found most of these references but they require additonal $$$$$ hardware/ software to talk to PC.
Does anyone/ has anyone found a source of cheap connectors which will allow the Mitutoyo verniers to plug directly into a PC? ie 4pin to 9pin serial? and associated fread-out freeware software???
I did find www.windmill.co.uk (http://www.windmill.co.uk) interesting?!!!

[bojan]

That is tue..
I will try to see if something else could be modified to fit into the mitutoyo slot...

[DJDD]

Thanks, Phil. Looks good.

Please excuse my ignorance but did the software and cable come with the digital indicator?

also, is the resolution that is listed on the vendor's web page correct?
resolution = 0.001 mm
accuracy = 0.01 mm
?

if they are is it correct that accuracy is 10x higher than resolution?
and that the resolution is 1/1000 of a mm?



cheers,
DJDD

[bojan]

Accuracy and resolution are two different beasts..
Accuracy means how far away you are from REAL value (compared to etalon or standard).
There is a third parameter which should be listed but is not, and that is repeatability of reading, because people mix it with accuracy.
Basically, if repeatability is good (that means, always the same reading whenever you measure the same thing) it can be calibrated if you have a standard.. then accuracy becomes the same as resolution

[allan gould]

Now at Xmas sale price of $98

[DJDD]

thanks, bojan.
I am familiar with precision versus accuracy, which is what you refer to as the "third parameter", I think.

however, i guess i had the thought experiment:

reading = 10.001
but accuracy states that true value could be anywhere from 9.996 to 10.006. Therefore, a precision of 0.001 is not helpful. In this case I am equating Acccuracy to error, which may not be strictly correct.:shrug:

[bojan]

You are quite right, in your way of thinking.
But if you have a standard (or something you know it is exactly 10.000) then you can determine what the error or gauge is, and account for it in any future measurement. Then, a precision (I would rather use term "resolution") of 0.001 will become useful.

However, in our case, I believe the stated resolution (number of digits) is more a marketing thing.. so not very useful in reality.

[robgreaves]

I bought an excellent 4 decimal place digital DTI on Ebay from the States for about $60. Very accurate and repeatable, and has an RS-type data out port too. Could easily be fitted to a focuser to measure travel.

I got mine for the day I actually get round to building a Foucault tester.

Regards,
Rob

[DJDD]

ok, so i think it is just a matter of semantics?
Accuracy = accuracy but resolution = precision? if so, then it must be marketing.


regards,
DJDD

[bojan]

Yes.

[Karls48]

The digital calipers I got from eBay have stated
Resolution 0.01mm
Accuracy +_ 0.02mm < 100mm, 0.03mm > 100mm
Repeatability 0.01mm

Ken, I don’t think that you will find software only solution to connect calipers output to computers RS232. I think that there will always have to be microprocessor to translate non-standard communication from the callipers to standard RS232 protocol. But, I may be wrong. Electronics bits are about $10 worth, but PCB is a problem. How I wish I had a one of those fancy small CNC routers to make one off PCB’s.

[bojan]

Well, it does not have to be RS232.. but it has to be level translation (because of 1.5V battery, digital levels inside calipers will be determined by this.. The rest could be done in software.

Unfortunately, these days neither Dick Smith nor Jaycar have anything useful to do one off PCB...
Couple of year back I was using photoresist spray (positive) distributed by them. (I think it is carried by RS Components, but the can costs is $25 or more. However, this spray can must be fresh, and since it is not used very often, it may be already unusable if stored for too long out of fridge)
The layout I would plot (or laser print) on tracing paper or transparent foil and use this to expose PCB sprayed with photoresist. Holes I would drill later, using guides (holes in the pads)
Today, they have only Dalo-pen (very thick hand-painted tracks) and PCB's with negative photoresist (Expensive.. and which means plotting or printing musy be done in negative - tracks are transparent, and no-copper areas must be black, which is a problem, this kind of plot is not generated by laypoout software packages).
I used to use one of those routers in NEC, but OHS came in one day and banned it's use (toxic dust and moving spindle - both very dangerous according to them.. ).

The viable method for PCB prototyping today may be printing directly on copper by means of modified ink jet printers (or printers that can print directly on CD's), and curing (at high temperature) ink, which has to be pigmented (not die-based) .
This layer of roasted ink offers sufficient resistance to etching solutions (like FeClO3, or H2O2+HCl)
I plan to try this at some stage..

[Karls48]

Bojan, I have done all that myself in the past. Plus very long time ago, sticking the adhesive artwork tracks and the pads on transparency to produce “master”. I have tried transferring (with hot flat iron) laser printed artwork directly on the cooper and then etching – it did not work. Most of my prototype circuits I build on various Vero bards. But as I’m getting older – I don’t see that well any more, my hands are not as steady and I found that I do make too many mistakes when build circuits on Vero boards. Also some circuits may not work as expected because of board layout.
Your idea of using inkjet printer is a interesting one, they become pretty cheap nowadays.

[bojan]

Karl,
It is not my idea :-)
I wanted to plot directly on copper (I have flat-bed plotter that can be used for this) but plotter has issues.. like drivers for windows, which are very bad. They do not take into account how ink dries so pen tends to be clogged very quickly.
I have DOS-based PCB editor and excellent driver for pen-plotter, but it works with DOS-based schematic editor.. so I do not use it any more, but I should..
So, I went on googling and found out that people are printing directly on copper with Epson printers (because they have permanently mounted piezoelectric heads, and some ink cartridges can be refilled and electronically "recharged". with pigmented inks (I said in my post above the ink must be die-based, but it is quite opposite, sorry..)
And I bought suitable one on ebay for $5 last year, it is now waiting to be modified and hopefully, used :-)
Anyway, one of interesting links is this one:
http://techref.massmind.org/techref/pcb/etch/c84-st.htm
http://techref.massmind.org/techref/pcb/etch/cx4200-vs.htm

There is even a discussion group at Yahoo:
http://tech.groups.yahoo.com/group/Homebrew_PCBs/
Have fun :-)

[bojan]

I am wandering now if IR laser can be used for curing ink? And which type of ink?
Then, I could use my pen plotter with windows drivers since there would be no clogging.. Just paint the ink all over copper, "plot" with laser (instead of pen) and wash away non-cured area of ink to expose copper that can be etched.. voila... :-)

EDIT:
There is company in Dandenog VIC, they do printing with UV curing inks, I asked for the price of that ink and when they replied "$800+ for 2dcl bottle".. I gave up :-)

[AndrewJ]

Gday Bojan

/

Not so easily, If its the chinese protocol,then its a 40kHz non std clock
Its messy without a PIC or the likes in between to process it
as you have to match two parallel channels of Clock and data

As to PCBs, there is a very cheap and foolproof method near you :-)
I tried those sprays and NEVER got any to work

Then i tried "Kinsten" brand presensitized boards and havent looked back
They even have double sided
Local Melb supplier is in Glen Waverley
http://kalex.net.au/Text/1098426142937-5357/pC/1098424907968-1192
4"x6" boards are $3.00 for phenolic or a whopping $5.90 for fibreglass

You just laserprint a std mask on clear film, cut a section of board to the size you want ( and save the rest )
peel off the protective plastic, expose under UV ( i use 10sec in the sun )
Wash off with draino and etch
Easy peasy

Andrew

[bojan]

Hmm Thanks for this Andrew :-)
Yes, I was using something like that long time ago (and that was in Europe, it was German product I believe) , and it worked perfectly (with layout plotted by penplotter on tracing paper or plastic foil).
I never managed to find it here though.
Will try asap! :-)