PDA

View Full Version : OOPS so close but the cigar is toast



darryl
02-11-2008, 02:50 AM
Some of you may recall the indexing table I was making out of pvc. Well, today I finally got around to marking it out in pencil as a double check before doing the actual scratching of degree marks. Counted off five degrees, mark- another five, mark- another five, mark- Coming around, 340 degrees, 345, 350, 355, 359/0 - oops

Turns out all the double checking I did was for naught- I have only 359 divisions instead of 360.

I'm trying to remain calm. I don't know if I should try to raise my blood pressure or lower it. Should I go eat another big meal, or try to bring up what I last ate. Put a brick through the tv screen- maybe now's the time to clean the carbon out of the van's engine- maybe I should just sit back and ponder the pleasures of needlepoint-

$%$^%#^p()(&(_*&%$^$%w$)(*%o%*&@^w^%%!

Well, that didn't help much. Sigh.

It will be quite a bit of work to fix it, but it can be done. I'm going to see how well epoxy sticks to pvc, since I'll have to rip out the beautifully made threaded rod loop, and the indexing section, since that will no longer fit perfectly to the new loop. I will have to make a new loop since I'm not going to be able to get the epoxy out of all the threads so I can re-glue it with one more space.

This is not what I wanted to find out- it is an error and it is a big one for this project. What probably happened was that I counted the threads starting from 1, and ended with the 360 in the same thread, when it should have been 360 ending in the 0 position.

Obviously I'm human. I need a drink.

Oldguy
02-11-2008, 04:10 AM
Know how you feel Darryl. Last night I finished putting together the last of the 14 drawers for the tool cabinet in the garage. Built the cabinet with a workbench top to replace the Craftsman roll around that I never roll around. Anyway, mounted the full extension drawer slides to a drawer and go to slide it into place, ARRGH, it a tight fit, too tight. After all of my careful measuring and calculating the inch plus of room is about an eighth of an inch short of what is needed. To say the least, I was not a happy camper. At least, it's woodwork and it can be salvaged, but I'm still p****d at myself for not doing a test fit after the first drawer was built.

I was watching your rotary table project and I've got the allthread idea stored away for future use. Hope you can salvage the nice work you've already put into it.

Glenn

oldtiffie
02-11-2008, 04:48 AM
That really is a shame as it was/is a superb job Darryl.

Yep - always start at zero.

It is surprising how prevalent that problem is.

The end of a rule is at zero - the 1" mark is 1" further on.

Try convincing some people that a baby less that 1 year old is in its first year - which is true. The clock starts ticking at "zero hour" - when the baby is born.

Its first birthday is the first anniversary of its birth.

On my 21st. birthday, I had been alive for 21 years and was just commencing my 22nd. year.

On my 60th. birthday I had been alive (from when I was born 60 years ago on zero time) for 60 years, was entering my 61st.year and my 7th. decade.

See how many people will get it wrong if you ask them to count the teeth on a gear with a large number of teeth.

Anyway, I really do hope that you are able to recover and restore it.

rantbot
02-11-2008, 05:21 AM
[never mind, probably not relevant]

Your Old Dog
02-11-2008, 05:33 AM
Darryl, while upsetting, these are growing and learning experiances. That notion, along with "need is the mother of invention" will likely propel you to a better answer. My guess is you and your project will come out of this better than you went into it. It's all in what you choose to make of it. Anyone to pull off a project as slick as yours is sure to finish up front in the pack.

tony ennis
02-11-2008, 08:37 AM
Yep - always start at zero.

It is surprising how prevalent that problem is.

We computer geeks call that an 'off-by-one' error and it's probably the biggest re-occurring programming error.

Ken_Shea
02-11-2008, 08:52 AM
Darryl,
Sorry fot that , wish saying we all have done similiar and likeley will do again would help but I know it wont, wouldn't help me either!

Any way, maybe some forum sympthoy post and the time it takes to read them may help calm down that temptation to kill your TV.

Kenny

pntrbl
02-11-2008, 09:53 AM
I'm absolutely heartbroken at your loss Darryl, but I sure enjoyed the way you wrote it up! That was good for a big laugh. I'm sure we've all felt those emotions at one time or another and you described them very well.

Hard to lose a piece you've invested a lot of time in tho. Now that I've stopped laughing you have my condolences and a sincere wish for success on the rework.

SP

Evan
02-11-2008, 10:13 AM
We computer geeks call that an 'off-by-one' error and it's probably the biggest re-occurring programming error.

Yep, except I have always heard it called a "fence post error". Example, how many fence posts does it take to make a fence with 10 spans of wire?

garyphansen
02-11-2008, 11:13 AM
HEY, I am surprized no one has pointed this out! Leave it alone, do not try to fix it, it is not worth the effort. You are only off one degree in 360 degrees. that is only an error of 10 seconds per degree. You can not read the protractor that close. On most things it is just not going to matter. Or am I wrong? Gary P. Hansen

Alistair Hosie
02-11-2008, 02:59 PM
Sorry to hjear of your dissapointment son keep up the otherwise very good work kindest regards Alistair

darryl
02-11-2008, 03:54 PM
Uh, what is it- nine fence posts- yeah that's it- (or is it twelve) :)

Thanks for the encouraging words. I will be able to salvage it- I stayed up a bit last nite to see if I could pry the epoxied rod out of the groove without damaging the table, and it wasn't that hard. Pvc doesn't grab onto epoxy well enough to be a problem- unless you need it to hold something- . In my case it's mostly a filler in a groove, so it does the job well enough, and I was able to remove it without problem. I curved up a new rod and cut it to length before I retired last night (this morning) and it will work out fine.

I had a similar problem with some drawer slides I installed- final dimensions were off by about 1/16, and I'm using them but it's tight. It eats at me a bit each time I open a drawer. Talking about that 1/16- I bought a tape measure the other day and remembered the experience I had with one being short by that sixteenth. I went through about 6 or 8 tapes before I found one that read properly. It was the cheapest one there too, so now I'm the proud owner of a piece of crap that measures correctly.

It's testy being a mortal sometimes. But therein lies the gist of this 'test'- how well can you cope with the things that life presents you- that is one of the rocks in the jar as opposed to the sand. This incident of mine, and the errors some of you have mentioned are but specks, though it sure didn't seem so last night. I was starting to envision things like cam gears, where one has 40 teeth, and the other 'happens' to have 79- or other more realistic things like threaded rod measuring up to have 19.7 threads per inch instead of 20, etc.

Anyway, I have the new rod circle made and fitted, not epoxied yet, and there will be no loss of concentricity, etc, and the project and I will survive. Thanks again for the support.

oldtiffie
02-11-2008, 03:58 PM
We computer geeks call that an 'off-by-one' error and it's probably the biggest re-occurring programming error.



Yep, except I have always heard it called a "fence post error". Example, how many fence posts does it take to make a fence with 10 spans of wire?

Yes, Evan,
that is the corollary of it.

The error in the first instance was counting zero as one - as in measuring a distance.

Your example was an error in numbering - ie in 10 spaces - has two variations:
1.
assuming that the fence started with a post there is a post at zero, hence there are another 10 posts for 10 spaces hence 10 +1 = 11 posts; and

2.
assuming that the start of the fence was not a post but say was fixed to a wall (or a pre-existing post/fence) there are only 10 posts required for the 10 spaces - ie the first post was at "1" and not at zero.

But back to the unfortunate events regarding the rotary table, I would have thought that the increase in diameter to accommodate the 1 extra space in 360 would not be very big at all.

Even if that were not the case, would it make any functional difference if a new 360 "space/tooth" ring of all-thread were used with a little of the outer "crests" of the "teeth/threads" removed to get the same outside diameter as the previous deficient "ring"?

2ManyHobbies
02-11-2008, 04:59 PM
I had a similar problem with some drawer slides I installed- final dimensions were off by about 1/16, and I'm using them but it's tight. It eats at me a bit each time I open a drawer. Talking about that 1/16- I bought a tape measure the other day and remembered the experience I had with one being short by that sixteenth. I went through about 6 or 8 tapes before I found one that read properly. It was the cheapest one there too, so now I'm the proud owner of a piece of crap that measures correctly.


S.O.B.
Well, that accurately describes my current project. I've got a cabinet that is 24 and 3/4 wide with a pair of doors that are 12 and 3/8 wide. Check the doors just before doing the glue-up and walking away and I'm about 1/16th of an inch wider than the cabinet. Of course I'm using the tape to measure, because 24.75" is just outside of the range of all of the solid measuring products I own...

The kicker is that I know better than to use the end of the tape for anything that needs better than 1/4" measurements. I am beginning to think that the measuring tape is more dangerous at 2a than the power tools are!

darryl
02-11-2008, 08:08 PM
When I was a kid I 'fixed' my dads tape measure by tightening the rivets on the end. I thought that it was loose- and I guess I shouldn't be surprised that many people don't know the use of that looseness. It's there because the tape can be used for both inside and outside measurements. The amount of looseness is equivalent to the thickness of the hook on the end. If you can measure by hooking over an edge and making a mark, then clamp a block to that edge and remeasure by pushing the hook up to that block, the reading should be the same. I haven't found any significant error on that account, just error in the actual length reading because the rivets are in the wrong spot. You can't make either reading (inside or out) properly in that case.

My dads response to my fixing was 'well, that tape's no good anymore'. I was perturbed until he showed me how they are supposed to work.

By the way, I checked a Stanley (good solid brand name) and it was out- why pay $12 plus tax for that when you can get the same error in manufacture for $2- I got my good one for less than that, but I had to test several before I found a good one. I would recommend for anybody buying a new tape to check it in store first.

I finally took my new baby out of the package and used it. It measures accurately, so fine. I put the batteries in it for the light, and got my moneys worth just in laughs. Even in the dark, I couldn't read the tape with that pitiful weak red led focussed off into space somewhere. Replaced it with a bright orange led I had, and now it's a mercedes. Whoooohooo!

Evan
02-11-2008, 09:16 PM
But back to the unfortunate events regarding the rotary table, I would have thought that the increase in diameter to accommodate the 1 extra space in 360 would not be very big at all.

Consider this: Imagine the earth as a perfect sphere, smooth as a cue ball. A string that unaccountably has exactly zero elasticity is stretched tight around the entire planet in a great circle.

Now the question: If you add just one meter of additional length to the string and evenly distribute it all the way around, how far will it stand off from the surface assuming that the standoff is made exactly equal all the way around?

wierdscience
02-11-2008, 09:37 PM
Don't feel bad bud,I bent 28 pieces of 1-1/2 channel iron exaclty the wrong way last week.At least you can fix yours,once channel is bent,it doesn't go back the otherway well.

On the bright side I now have a fresh supply of channel drops to play with:)

andy_b
02-11-2008, 10:54 PM
I had a similar problem with some drawer slides I installed- final dimensions were off by about 1/16, and I'm using them but it's tight. It eats at me a bit each time I open a drawer. Talking about that 1/16- I bought a tape measure the other day and remembered the experience I had with one being short by that sixteenth. I went through about 6 or 8 tapes before I found one that read properly. It was the cheapest one there too, so now I'm the proud owner of a piece of crap that measures correctly.





My dads response to my fixing was 'well, that tape's no good anymore'. I was perturbed until he showed me how they are supposed to work.



I finally took my new baby out of the package and used it. It measures accurately, so fine. I put the batteries in it for the light, and got my moneys worth just in laughs. Even in the dark, I couldn't read the tape with that pitiful weak red led focussed off into space somewhere. Replaced it with a bright orange led I had, and now it's a mercedes. Whoooohooo!


dude, you made water come out of my nose twice in the same thread. you're cracking me up.:)

glad to hear the PVC rotary table is going to work out. i thought it was a cool project.

andy b.

2ManyHobbies
02-11-2008, 10:57 PM
Consider this: Imagine the earth as a perfect sphere, smooth as a cue ball. A string that unaccountably has exactly zero elasticity is stretched tight around the entire planet in a great circle.

Now the question: If you add just one meter of additional length to the string and evenly distribute it all the way around, how far will it stand off from the surface assuming that the standoff is made exactly equal all the way around?

Meter of two pi, though only if we are allowed to approximate the horse as a cube.

torker
02-11-2008, 11:55 PM
Darryl... glad to hear you sucked it up and went on with the repair. I hate when people give up on such a project. I'm actually a bit guilty there. Had to give up on my CNC rotab due to time constraints but i'll get back at it soon.
LOL! Funny story about the tape measure. I did the very same thing once. Tightened up them stoopid floppy rivets with a hammer.
Russ

darryl
02-12-2008, 01:04 AM
To Evan's question, I can't calculate that to show a difference with only 8 places on my calculator.

In my case, the error required me to open up the groove by 12 thou of diameter to accomodate the extra .050 on the length of the rod. I actually had to dial in .004 to get .008 extra diameter, since I had that small amount of play initially. Now the rod is tight to the outer diameter of the groove, and I have exactly that one extra spacing made up for. The .004 difference in the radius is just barely too much to allow me to use the short arc section of rod as the indexing piece. I will have to move it if I want it to click in solidly. No matter- I can just remove it, grind some off the outside of the threads, then epoxy it back in place. I'm sure glad in this case that the epoxy doesn't stick to the pvc all that well. Where I have to grind is not on the part of the threads where the indexing takes place, so there's no compromise there.

My new circle is epoxied, and I'm going to wait til tomorrow to unclamp it so I can glue the arc back in place a few thou further out.

One thing that's good about this is that there will be no traces that this repair was even required- no extra holes, no visible re-machining, etc.

I only hope I made the full 360 spaces this time :)

To Garys question- if I left the thing as it was, I'd have an error of 1/2 a degree at the 180 degree point- far too much since at least one of my applications will show an error of 1/4 inch over one meter- I can skilsaw closer than that by eye, so yes, it does matter.

winchman
02-12-2008, 02:57 AM
That's great that you were able to fix it and save all that nice work.

Too bad the error was 359 and not 361. You could have made a super accurate vernier scale before you fixed it. http://bbs.homeshopmachinist.net/images/icons/icon12.gif

Roger

dp
02-12-2008, 03:08 AM
Yep, except I have always heard it called a "fence post error". Example, how many fence posts does it take to make a fence with 10 spans of wire?

10 if the fence is an enclosed polygon. N+1 otherwise (where N=number of spans).

Oldguy
02-12-2008, 03:45 AM
On the tape measure issue - make sure that all of your tapes are accurate AND agree with each other. I went through mine, kept the accurate ones, tossed the others, then went and bought a couple more that agreed with the good ones. Now I don't have to remember which tape I was using when I started a project. Of course, if I would finish a project in a reasonable time frame that wouldn't be a problem.

Glad you were able to salvage your rotab. I'm heading to the store for some more plywood.

Glenn

Evan
02-12-2008, 04:14 AM
To Evan's question, I can't calculate that to show a difference with only 8 places on my calculator.
Is joke, yes?

No need to calculate the planet at all, just the addition to the radius.

Radius = 1 metre circumference / pi / 2

Added Radius = 15.91549430918953357688837633725143620344596457404 ... centimetres

Which equates to approximately 16 centimetres additional standoff. About 6 inches. All the way around.

John Stevenson
02-12-2008, 04:24 AM
Yup,
Error of one.
More often done with holes on a dividing plate.
I can remember doing 42 1/2 splines on an Alpfa Romeo half shaft once.

I remember it well because when the guy came to collect it I learnt a whole bunch of new words :D

What made it worse was that after being told I was incompetent or was it incontinent ? or both ? he took a new one to some expensive high flying gear company and they did the same :D


.

Bguns
02-12-2008, 04:41 AM
Will not Lift ring off earth 6 inches.....

"Is joke, yes?

No need to calculate the planet at all, just the addition to the radius.

Radius = 1 metre circumference / pi / 2

Added Radius = 15.91549430918953357688837633725143620344596457404 ... centimetres

Which equates to approximately 16 centimetres additional standoff. About 6 inches. All the way around."

Evan your math is bit off,

Earth is ~ 40,075.15 Km in circumference, 1 meter extra will not quite do that...
Just off top of head I see a problem

Evan
02-12-2008, 04:46 AM
No mistake. The increase in the radius is the same regardless of the size of the original circle.

Bguns
02-12-2008, 05:29 AM
Late night cobwebs :)

Even my little solar calc could handle that... :)

oldtiffie
02-12-2008, 07:45 AM
Consider this: Imagine the earth as a perfect sphere, smooth as a cue ball. A string that unaccountably has exactly zero elasticity is stretched tight around the entire planet in a great circle.

Now the question: If you add just one meter of additional length to the string and evenly distribute it all the way around, how far will it stand off from the surface assuming that the standoff is made exactly equal all the way around?


Is joke, yes?

No need to calculate the planet at all, just the addition to the radius.

Radius = 1 metre circumference / pi / 2

Added Radius = 15.91549430918953357688837633725143620344596457404 ... centimetres

Which equates to approximately 16 centimetres additional standoff. About 6 inches. All the way around.

Evan is correct.

This is the solution for the general case:
http://i200.photobucket.com/albums/aa294/oldtiffie/Circle-rad1.jpg

Co is the commencing Circumference.
C1 is the circumference with 1 unit added.

C = 2.pi.R in every case

C1 = Co + 1

R1 = Ro + 1/2.pi

= Ro + 0.1592

So in every case, no matter what the starting radius (Ro) is, the radius resulting from an increase of Ro by 1 unit will always be R0 + 0.1592

Just put the figures in an run the math.

darryl
02-12-2008, 07:02 PM
Was not joke, no. Brain not do formula, blame calculator. :)

Rotab fiisked, 360 disivion exacry now, sepell check bad as calcuator. :)

Well, that's back on track, but must put aside for now and fix the van. I wish it was like the old days where there was enough room for a house under the hood. Oh well.