PDA

View Full Version : It's now official, it's a POS



John Stevenson
10-05-2011, 05:23 PM
So we work to the best of our abilities and use the best we can afford but there is always someone who has better toys.

So I have this POS Bridgeport and currently working on some parts for a company that makes CCM machines, not uses them, makes them. Now these guys think and work in microns, not imagined microns, not well I moved 1.6395475423" so it must be 1.6395475423" but real world microns.

As a test I took immense care in doing a 4 hole layout, took all the slop out, only worked in one direction and used a two year old Sino DRO system.

Hole layout was as follows:-

X Y
Hole A1 -62.69 -18.00
Hole A2 -62.69 -43.00
Hole A3 62.69 -43.00
Hole A4 62.69 -18.00

Every hole was dead nuts on before the short spotting drill was offered to the work. To all intents and purposes this should have been a classic text book exercise.

Results off the CCM were as follows :-

X Y
Hole A1 -62.8875 -18.1487
Hole A2 -62.9432 -42.8534
Hole A3 62.9436 -43.1674
Hole A4 62.6783 -17.8952

So it's now official, it's a POS.

Mind you how many machines could stand this test ? without access to one of these very expensive machines you don't have a clue.

J. R. Williams
10-05-2011, 05:30 PM
Don't blame the mill' blame the digital system.
JRW

DFMiller
10-05-2011, 05:32 PM
Hmm. That seems to be more error than I might have expected. Whats the spec on your DRO encoder?
0.003" per inch and 0.008" per inch error.
Dave

Kiwi
10-05-2011, 05:37 PM
And in half an hour when the tempreture drops ?:(

macona
10-05-2011, 05:44 PM
Try doing a set of bored holes and see how that reads.

jkilroy
10-05-2011, 05:56 PM
****, I can manually layout holes and pin punch them better than that. Some of those numbers are a quarter of an inch out!!!

John Stevenson
10-05-2011, 05:57 PM
BTW those measurements are in mm.
Funny thing is these holes are on steps in a big 6" diameter thick walled alloy tube.
The tube is 800mm long and I faced that off using a big milling cutter and a one metre long rule.
One step was 210 mm in and the CCM read that step as 209.9879

So the rule is better than the DRO :D

It's scary working for these people :rolleyes:

ahidley
10-05-2011, 06:06 PM
Sir John, Earl of Bridgeport & Sudspumpwater. MBE

If your DRO uses glass tubes for measuring then it is possible that you have gotten some cutting oil on the tube. Take the tube out and clean it off and I bet it will work fine again!

David Powell
10-05-2011, 06:10 PM
I made parts on a " First" milling machine which were frequently checked on a CMM. Read in metric the discrepancies frequently seemed frightening, but when translated into Imperial they generally sounded allright.( our general allowance was 2 thous unless otherwise stated) On only a couple of occasions did I need to remake/modify parts because what I actually got was outside the designers final allowances. I blame that fellow Bonaparte for creating lots of unnecessary worry!! Regards David ( still thinking in thous) Powell.

DFMiller
10-05-2011, 06:18 PM
John you are messing with me.
Your example used inches
"
moved 1.6395475423" so it must be 1.6395475423" but real world microns.
"
and your grip was in mm
Are you sure your DRO is not made of Brass?


:D
Dave

gcude
10-05-2011, 06:19 PM
You could head over to the PM site and make a down payment on this if it would make you feel better:

http://www.practicalmachinist.com/vb/machinery-sale-wanted/rare-sip-3k-jig-borer-mint-233339/

:D

jkilroy
10-05-2011, 06:34 PM
:D Sorry for my confusion, and my oh so American like insistence that everyone does everything the same way. MM's, whew, thats much better.

David Powell
10-05-2011, 06:42 PM
If you have to consistently work to the last thou or so, and especially if the concern is mainly the location of holes in relatively flat parts( ie baseplates) the tram of the machine becomes very important, especially when one hole is to be bored with a boring head, thus needing a long distance between quill end and work and another is to be reamed with a tiny reamer ( 60 thous or less), thus the end of the quill is much nearer the work than before.If the heavy handed fitters from the shop floor came and used " my " mill in my lunch time I usually found it slightly out of tram afterwards, not enough to be obviously noticeable but enough to give poor results on the cmm later. Regards David Powell.

Alistair Hosie
10-05-2011, 06:46 PM
John's dro is made of wood and is acurate to the nearest inch or so.I know I made designed and calibrated it for him with double sided tape and playdough and a main spring from an Ikea clock .Its's a little wonky as it is designed to allow for vibrations and is oiled daily with teatty milk. so there stop expecting too much when you know I am remotely involved.Alistair

boslab
10-05-2011, 07:44 PM
i wish i could work that accurately, i regularly work to microns, its easy, i look at the hole and say 'thats exactly whare i wanted it', 'shame the hole in the other bit supplied by the customer btw, is in the wrong place!
i can fix it with a cutting torch lol
i feel sad for your loss [of accuracy] jhon, get weaker glasses.
on another unrelated topic [but funny] my missus was mixing some cement for me, i tend to be a bit of a perfectionist as far as mortar mix is concerned, anyway i made the mistake of moaning about her accuracy while mixing, told her, 'its not stiff enough to lay bricks properly'
she retorted, let me pop in the house for some cement additive, i was caught off guard, plasticiser will make it worse you idiot, no she said, i'll use your viagra, its not going to be much use to you for a while, should stiffen up nicely, you orrible cow i told her, didnt help either.
you could try some [viagra] in the oil! should take up the play!
get a nice cincinatti or somthing!
regards
mark

John Stevenson
10-05-2011, 07:50 PM
Viagra, now there a thought, next time I go see the old man in the nursing home I'll get some.

They use it to stop them rolling out of bed at night.

uncle pete
10-05-2011, 08:05 PM
John,
Those are totally understandable readings. You have a British built Bridgeport. Over here we build our CMM's on North American built Bridgeports.:p

That was pretty interesting to see what the results were when measured with top quality equipment. I also wonder how your DRO would check out against some guage blocks. Since you have exact measurements for your part can you recheck your DRO against the part with a really accurate DTI? The measurements obviously wouldn't be to the accuracy of that CMM but it would get you close.

Pete

David Powell
10-05-2011, 08:23 PM
Further to my previous posts. I ran the " First " mill for 5 yrs before a big, for me and it . job broke out. I found that on fibreglass pallettes made for holding circuit boards my setting out in the X direction seemed to be consistently " out" by a couple of thous or so compared with brand new , allegedly correct items. To cut a long story short we eventually found out that the readout had never been " corrected" on installation and that 12" of X travel on the readout was actually 11.997", ie it and real measurement disagreed 1 thous in 4". No one wanted to do anything about it, so for the rest of the time i simply added a thou every 4". My Sino readout at home came with a sheet showing how well it measures., but if I had put it on crookedly then it would still be wrong. Have fun, work safe and blame Bonaparte! Regards David Powell.

lazlo
10-05-2011, 09:14 PM
As a test I took immense care in doing a 4 hole layout, took all the slop out, only worked in one direction and used a two year old Sino DRO system.

Did you ever wonder why all the mounting holes are off-center on Chinese machine tools? Now you know why :D

Scottike
10-05-2011, 09:40 PM
Sir John, If you get rid of that Bridgeport for a "real mill", what would you have to complain about?
"A life without grief is not life"

loose nut
10-05-2011, 09:43 PM
Did you ever wonder why all the mounting holes are off-center on Chinese machine tools? Now you know why :D


They are on center if you look at them with slanted eyes.:D

justanengineer
10-05-2011, 09:58 PM
Sir John I think its time you traded in that imposter machine for a nice stout K&T.

J Tiers
10-05-2011, 10:31 PM
How the hack do you KNOW after using that CMM........?

I got output from a CMM to match up a part to an assembly..... Sheesh..... all I got was a "cloud" of points...... we hadn't a clue what it all meant, so we grabbed digital calipers, and measured it up.... made the part to fit what we measured, and it fit.

One of the digis was 4" HF, the other a 6" Starrett, so you'd have good reason to doubt them both.

But thay were a lot more useful than the output I got from that %$#@! CMM.... it was a Faro, mebbe that is why.

MickeyD
10-05-2011, 10:48 PM
The only thing this proves is that John needs some new glasses so he can see to sharpen his drill bits correctly or he needs to quit buying Chinese drills. I was drilling a hole pattern on one of my Okuma mills that will hold +-.0002" all day long (gotta love those Japanese glass scales) and when I checked them the holes were all over the place. Turned out it was very poorly ground bit that was jobber length and a very light spot drilling. Some of the holes were off by almost half a diameter. The lesson learned was to always examine the bits before they go in the "good" drawers.

beanbag
10-05-2011, 11:52 PM
Did u check the DRO against an indicator mounted to the spindle?

Also, wouldn't it be more fair to drill undersize and then plunge with an end mill?

Davo J
10-06-2011, 01:25 AM
My mill came with Sino SDS6 and I just looked through the manual and I can't see anything about adjusting the read out if it is reading wrong, which i find strange.

I have had 2 Meister's for around 5 years before this and they have a section in the menu where you can check the DRO against gauge blocks, etc and adjust it if necessary by putting in the PPM (parts per meter) that it is out by.

I would be checking the scales and read out against some gauge blocks to see if it's reading true, but by the look of it there is no adjustment.
I would say there is a way, but it is probably a factory set and you would either have to know how to get into the program, or it may not be adjustable on these.

If you interested in this PPM part of the Meister, PM me you email and I will scan it and send it through to you.

Dave

EVguru
10-06-2011, 04:35 AM
My mill came with Sino SDS6 and I just looked through the manual and I can't see anything about adjusting the read out if it is reading wrong, which i find strange.

I'm pretty certain that the Sino DROs come with not only linear compensation, but segmented compensation. The Sino scales come with a certificate showing the deviation.

My Sinpo DRO has linear compensation and it looks like it has segmented compensation as well. I've not been through the manual yet.

Davo J
10-06-2011, 05:24 AM
Your right there, I just had a good look through and it is their.
Earlier on I just looked at the index and had a quick skip though the manual.

So it can be adjusted for Johns POS, even if it is out. LOL

Dave

Forrest Addy
10-06-2011, 05:50 AM
Well, a turret mill isn't a jig borer and 30 years of use and hit and miss TLC doesn't help. They have limitations but if in good condition and accurate tram and equipped with a good measuring system, a turret mill should bore an array of holes within 0.001" (25 microns) of true location referenced to an existing coordinate grid. With care and a full machine tune up including gib adjustments you might halve the the true position readimgs to 0.0005" (12+ micron).

Getting consistant results on a turret mill to microns or tenths is a crap shoot.even with a super high resolution DRO. There's simply too much error, lost motion, and elasticity in the machine.

Notice I said "consistant". Lots of people me included have worked to closer tols than 0.0005" on a BP for feature size and location but it requires excrutiating, time consuming care. You can do the same job in a real jig borer and hit 0.0001" (2+ micron) positioning in a third the time.

Weston Bye
10-06-2011, 06:27 AM
Sir John,
I can see by your own admission that you're no George Wilson.:D

Clumsy bastard.*

*a term of endearment

John Stevenson
10-06-2011, 07:12 AM
The point I was trying to get over isn't so much the inaccuracy as knowing about it.

There are three parts to this assembly and all three mate up fine so REPEATABILITY is OK it's the POSITIONING that has errors.

Without a CMM to check how far out is just how good are the average home shop machines given they have probably had a hard life in industry prior to us getting them ?

These are drilled holes and a drill is probably the most inaccurate tool invented, I can't bore as they are too small and the errors are within tolerance anyway, the customer isn't moaning, they just sent a CMM sheet out as it had additional info on it that needs adding to the drawing.

Now I have free access to one of these machines out of interest I'm going to send some of the small converted CNC gear in for a check, just out of interest.

David Powell
10-06-2011, 07:48 AM
Firstly the universe of unobtainable perfection, sought after relentlessly by some, never found by any, existing only on paper, not in finished products, frequently written and argued about. Secondly the universe of admired by all, that is work of consistently very high quality, aimed at by most, achieved quickly and easily by some, slowly and laboriously by some. never quite reached by the majority., but none the less often aimed for by most. Then thirdly the universe of good, that is plenty good enough for the job it is to do, made without fanfare, argument, or even a thought that it is worthy of more than a passing comment. I am sure these universes connect and sometimes their inhabitants cross from one to another. Just some early morning thoughts regards David Powell.

Weston Bye
10-06-2011, 08:04 AM
Wow, David, heady stuff for so early in the morning. Do your thoughts get better as the day progresses, or have you now used up your ration of deep thoughts for today?

Good stuff - worth writing down.

J Tiers
10-06-2011, 08:44 AM
Thinking about a BP or any other similar machine.........

The largest point of irreducible error I would think is the quill.....

Sure, there is flex, movement of slides, etc, but the quill is the lightest sliding part with the least bearing area and likely the most wear potential.

The quill must move, fairly easily, so it needs to be SMALLER than the hole it goes in. That slop range is unavoidable.

Regardless of lube, there is a point of imprecision (error) right there.... you cannot KNOW the quill is at or moving down the center of the hole, or even that it is located where it was a short time before when the last hole was drilled or cut taken.

Whatever the clearance between quill and bore, less the minimum lube space, IS the potential error, And I don't see how anyone can get it to be better than that consistently. So since that space is going to be at least around 0.001 total, it seems unlikely that one can do ANY better than that as far as location, even with a perfect machine.

A "real" machine will have added wear, etc, making the error larger.

All that oil, grease, etc can do about it is to hold against transient forces, any consistent force in one direction, ESPECIALLY if associated with vibration , as when milling or drilling, should force lube out and move the quill.

Quill adjustments for tightness are usually pretty minimal and ineffective, things like partly split castings with nipping bolts, and the like. They don't keep a round hole, and they often only affect the lower end of the quill, with no provision for the top end.

Obviously the longer the quill, the less the angular error from a given slop, but the absolute error amount isn't affected unless the length gives more bearing area.

If anyone doubts the idea of wear, just consider the lathe tailstock, which is used LESS than the quill of a mill, but commonly wears sloppy, and often has NO means of adjustment whatsoever.

The Fixer
10-06-2011, 09:07 AM
With all due respect John, this is quite obviously a case of operator error being blamed on the machine, The POB (poor old bridgeport) was only doing what the operator set it to do.... and yet once again it's subjected to the name calling..POS.
I think perhaps it's time you gave that POB away to a more caring home, surely some one nearby would be willing to take it into their shop and treat it with respect!

al

aboard_epsilon
10-06-2011, 09:15 AM
it's the table , and knee locks ...that's were the inadequacies come from.

You can tram it with either all locked up or not.

Then you can decide to use it with it either all locked up or not .

in practice, works out like this ..

Tram it with it locked up ..then try and do accurate work on it, with it not locked up ..then you will get the inaccuracies

vice versa ...tram it up without it locked up .....then try and do accurate work on it with it locked up.you fail again ..even with one lock on.

and the further you have to pull the quill down whilst it's in any of the situations above ..the worse the inaccuracy will be....especially for the knee lock.

Also, the gibs must be adjusted as tight as you can get them without effecting how hard it is to turn the handles at each end of its travel.....the gib free play will directly effect how much inaccuracy there is ..when you operate the locks.

indicate a hole ....then put your locks on ..any of them ...watch the dial move .

wind the knee down ..see what the dial is showing in the hole , when you have to pull the quill down.

more

locking and unlocking between holes no matter how it was trammed can directly effect the machines accuracy..you must think about what you are doing.

not directly aimed at you john..for everyone ...but me knowing this ..i still forget about this phenomenon and screw up sometimes.umm well not completely screw up ..im just sad that i could have done a little bit better.


all the best.markj

Forrest Addy
10-06-2011, 10:41 AM
I was curious so last night I went out and roughed in a test piece duplicationg Stevenson's disaster.

I got up extra early and did the hole locations. I tracked with a 0.0001" indicator the exact place the DRO clicked over (0.0005" resolution), set the table and saddle clamps to drag a little, allowed lots of time for the heat of machining to bleed out of the part, etc in other words used all my tricks. I drilled 0.010" U/S, bored to locate with an U/S endmill, and reamed to size the bore using two reamers the semi-finish was 0.0005" smaller than the finish. Then I milled a side and end ensd to make reference planes.

I don't kave a CMM so it had to do it the old fashioned way, Jo blocks, granite flat, and transfer stand with a Federal lever head and a gage amp. The best I can do on my 1981 BP clone with about 200 hours on a light re-scrape is about 0.015 mm of position on the XY grid, And the holes were parallel in Z to the skimmed edge to 0.008 mm in 20mm on the average with almost unmeasurable splay. Not bad for a 0.0005" resolution DRO.

I didn't bother running a full blown inspection. I just went over it enough to compare Sir John's POS to my own not-bad clone. And maybe rub it in a little.

Conclusion: Sir John does indeed own a POS BP but not that much worse than mine. If he got 15 micron feature location on an array of holes merely drilled not precision bored, my hat's off to him.

mike os
10-06-2011, 10:49 AM
hang on.. if its a repeatable error it cannot be blamed solely on the machine if you are using a dro surely?

if it were all over the place or you were using "number of turns on handwheel".... machine fault/errors or holes not perpendicular (assuming z is vertical)

if turn to 28.664 x & 13.443 y on dro & not in right place surely thats a dro error?

if it needs a new home......

MetalMunger
10-06-2011, 03:40 PM
John-
Now you know why Jig Borers were invented and why they are kept in a climate controlled area.
If you pay the shipping FOB Anchorage and include that slotter I think I might give you $100 for that POS Bridgeport. :D

Forrest Addy
10-06-2011, 03:47 PM
I don't pretend to savvy the guts of a DRO but as I understand it the scale has its own resolution and there's a - um - math processor (?) that takes the scale pulses and calculates updates for the display according to the units set, the resolution parameter, plus it cranks in whatever corrections are called for on the segment by segment calibration table.

So there's several hurdles between the scale and the display that make resolution sized move chancy on a DRO. I know that some 0.0005" moves on my DRO display may actually be 0.0004 to 0.0007. They only average 0.0005" because the scale is metric and Imperial units display updates vary with the 10 micron resolution of the glass scale. The rounding error is never really significant. If I want to work to the best accuracy of the DRO I'd improve my chances an unimpotant trifle making math conversion from Imperial to Metric to 5 digits and working in Metric units rather than Imperial units.

It's one of those hair splitters I think about waiting for cuts to run to completion. Gotta do something or I'll go batty -er..

Timleech
10-06-2011, 03:51 PM
I don't pretend t understand the guts of a DRO but as I understand it the scale has it's own resolution and there a um math processor that take the scale increments and clicks the display according to the unts and the resolution parameter set plust it cranks in corrections on the segment by segment calibration table if set.

SO there's several hurdles between the scale and the display that make resolution sized move chancy on a DRO. I know that some 0.005" moves on my DRO display are actually 0.0004 to 0.0007. They only averave 0.0005" because the scale is metric and the display updates vary with the 10 micron resolution of the glass scale incrementsd. The error is never really significant. If I want to work to the best accracy of the DRO I'd improve my chances making math conversion from Imperial to Metric to 5 digits and rounding rather than operating the DRO in imperial units.

It's one of those hair splitters I do waiting for cuts to run to completeion. Gotta do something.

I have a Sino DRO, probably much the same as John's, it's metric based & the imperial conversion is poor & pretends to be much more precise than it actually is. John was working in metric dimensions, though.

Tim

J Tiers
10-06-2011, 10:17 PM
The conversion should be perfect.......or nearly so.

it depends on how they do the math. if they truncate, or decide to do math in 16 bit integer instead of "float32', the conversion might indeed stink.

There are lots of ways to get an error by choosing a faster simpler way to do conversions.

There isn't a good excuse for acually DIONG that, though. Not on a device that is supposed to be accurate.

becksmachine
10-06-2011, 11:33 PM
Results off the CCM were as follows :-

X Y
Hole A1 -62.8875 -18.1487
Hole A2 -62.9432 -42.8534
Hole A3 62.9436 -43.1674
Hole A4 62.6783 -17.8952



There should be a clue here in that two dimensions are within .0004 mm of each other but in different directions from zero.

Granted, they aren't really that close to the desired dimension, but the consistency seems to be out of place among the rest of the wildly dispersed inaccuracies.

Dave