In response to the OP When I got my PM727 I had to keep telling myself I didn't buy a K&T or such and that for dollars invested am very happy. It does seem odd to have error like stated but it could be the graduations on the collar too, in respect to spacing? Of course I can argue that out in my head and get the answer no if you always stop on the same mark because a revolution is a revolution. Sorry stupid post.
-
Welcome back Guest! Did you know you can mentor other members here at H-M? If not, please check out our Relaunch of Hobby Machinist Mentoring Program!
PM25 DRO and Hand wheel off by 1 part in 1000?
- Thread starter Bill Kahn
- Start date
- Joined
- Oct 13, 2014
- Messages
- 9,685
This thread piqued my curiosity as to the accuracy of my own mill. I have high quality Taiwanese made Bridgeport-esque clone fitted with an Easson DRO with 1um glass scales. I tested the DRO against the dial at 1, 5, 10 and 20 inches, indeed there was a consistently progressive error at around .00015 at one inch, .00075 at five inches, .00145 at 10 inches and .00295 at 20 inches. so it looked like a lead screw error. However, when I tested against a 4" precision gauge block with a DTI I get about 00055 error, which is consistent with the above error, so it does indeed look like the DRO may be off! I will try further testing and see if I can find and confirm the exact source of the error...
- Joined
- Sep 29, 2014
- Messages
- 2,102
.00015" in 1", you hit the jackpot there my friend!
- Joined
- Mar 3, 2013
- Messages
- 1,448
I'm not sure who first said it but I remember this quote from high school shop class "If you have one ruler you know how long something is, if you have two rulers you never know".
That sounds great to me at .00015" per inch systematic error. So, a 20" part might be wrong by 3 thousandths. -Bill
I missed a key idea--what is "the dial" against which your tested the DRO? I am guessing it is a known 20". What do you have that is a known 20" to an accuracy of like .001" or such? Thanks.
OK, Two weeks into my PM25MV ownership and I am now looking at the Z axis.
First, One full 360 degree turn of the Z hand wheel moves the head by .083". Now, I ask, .083? Where ever did that come from? I have considered maybe it is actually 1/12th of an inch and it is really .08333. But no. The marking shows exactly .083. The extra .3 thou would be quite obvious and it is not there.
Second, the Z accuracy is somewhat worse than the X and Y accuracy. I moved the Z exactly 5.000" as established with my 1-2-3 blocks and as my caliper measures as 5.000". The Z hand wheel showed up as 60 turns (of .083" each) plus .0115" for a total of 4.9915". So an error of .0085". Or .0017" per inch. 50% more than the X-Y error. If I guess that one turn really is 1/12th of an inch, not .083, then the error would be even greater--.0023" per inch. The other direction. But as the dial shows .083 I do not think thinking it is really 1/12th makes sense.
The DRO Z accuracy is better, but unlike the X-Y where the DRO is basically correct, the vertical DRO showed 4.99764". For an error of .0024" in 5" or .0005" per inch.
I am most happy with the PM25MV. I have nothing to compare it to, but whatever expectations I did have are being well met. The DROs seem fully adequate to meet any need I can imagine having--if some year (decade) I decide I want to start working to 1 thou accuracy I will 1) need a totally new set of skills 2) need all new ancillary equipment and 3) need a higher-end mill. All three seem to be well balanced now and more than adequate for the next dozen projects I am planning out.
But, .083" per turn? Any thoughts? (For all I know all mills have that standard? Maybe some throw back to ye merry old England strange length unit? I have no idea.)
-Bill
I have read this thread and am sucked in to the issue. I agree that it would be great to have dials on our machines that are "correct". I have no idea how accurate professional machines are, an I have no doubt our hobby machines are less accurate. I also know professional machinist can produce very accurate parts on less than accurate machines. My thoughts are they use the dials on the machines to get dimensions in the ball park, then measure with accurate tools to know how much difference they have between the part and the needed dimension. The resulting cut will then be a small cut, so the inaccuracy in either dial or dro will result in a tiny error.
I know I use this technique when making furniture (which I am much more knowledgeable about) on woodworking machines which are not accurate at all, an easily work with 0,1mm tolerances. ( My "accurate measuring tools" being a tape measure and calipers
)
When working metal I am not quite so knowledgeable, so I haven't ventured into the problems described in this thread, but I am almost sure proper technique is an important issue here. And I agree with the notion that you need known accurate measuring tools to do accurate work in any case.
Peter
I know I use this technique when making furniture (which I am much more knowledgeable about) on woodworking machines which are not accurate at all, an easily work with 0,1mm tolerances. ( My "accurate measuring tools" being a tape measure and calipers
)When working metal I am not quite so knowledgeable, so I haven't ventured into the problems described in this thread, but I am almost sure proper technique is an important issue here. And I agree with the notion that you need known accurate measuring tools to do accurate work in any case.
Peter
- Joined
- Feb 1, 2015
- Messages
- 9,985
If your PM25MV head is anything like the head on my old mill drill, my dial reads .085"/revolution. The drive for the spindle quill is essentially a rack and pinion, the distance per revolution of the mail crank being determined by the pitch diameter of the pinion gear.
I measured the distance separating 13 teeth as 104.3mm or 8.023mm/pitch. Assuming that the manufacturer cut a nice even pitch length, that would give the rack on my mill a pitch of 8mm. One full turn of the main crank moved approximately 3.14" or 80mm so my pinion gear would have 10 teeth.
The fine adjust wheel is a worm gear drive. Each crank of the wheel moves the engaging gear one tooth. It took 37 turns of the fine wheel to travel 79.7mm by the DRO so the gear on my worm drive must have 37 teeth. This calculates to a distance/rev. of .08482".
The nice thing about gears is that all of the numbers are ratios of relatively low whole numbers. The only used assumption here is that the rack was cut to "exactly" 8mm pitch. Forty turns of the fine wheel traveled 86.21mm for 2.155mm/rev or .08485"/rev. This agrees fairly well with the predicted value of .08482"/rev.
Note that the manufacturer would not be likely to put 84.8 divisions on the dial but would round it to the nearest whole number of 85 divisions.
Finally, I did notice differences in DRO readings depending on whether the quill was locked or not. This is not surprising to me as for years, I have tightened the quill lock slightly when drilling to precise depths. I did a quick check of the calibration of the DRO using calibrations standards for my micrometer set and the was some deviation but generally in the range of a half a thousandth.
I measured the distance separating 13 teeth as 104.3mm or 8.023mm/pitch. Assuming that the manufacturer cut a nice even pitch length, that would give the rack on my mill a pitch of 8mm. One full turn of the main crank moved approximately 3.14" or 80mm so my pinion gear would have 10 teeth.
The fine adjust wheel is a worm gear drive. Each crank of the wheel moves the engaging gear one tooth. It took 37 turns of the fine wheel to travel 79.7mm by the DRO so the gear on my worm drive must have 37 teeth. This calculates to a distance/rev. of .08482".
The nice thing about gears is that all of the numbers are ratios of relatively low whole numbers. The only used assumption here is that the rack was cut to "exactly" 8mm pitch. Forty turns of the fine wheel traveled 86.21mm for 2.155mm/rev or .08485"/rev. This agrees fairly well with the predicted value of .08482"/rev.
Note that the manufacturer would not be likely to put 84.8 divisions on the dial but would round it to the nearest whole number of 85 divisions.
Finally, I did notice differences in DRO readings depending on whether the quill was locked or not. This is not surprising to me as for years, I have tightened the quill lock slightly when drilling to precise depths. I did a quick check of the calibration of the DRO using calibrations standards for my micrometer set and the was some deviation but generally in the range of a half a thousandth.