PM1030V calibration question

[Bill Kahn]

I started a small project that needed me to be able to trust the calibration of the hand wheel dial that moves the apron back and forth.

The dial is labeled as .75" per revolution and has .01" markings.

I started to make measurements of how far the apron actually moved when I turned the hand wheel one revolution. I got all sorts of inconsistent results. I used a ruler, gauge blocks, calipers. I did it over short distances, long distances. And the numbers came back all over the place. Sometimes the apron seemed to move .8" and other times .7" Yeah--all over the place. (Yes, I was always careful about backlash). Any particular measurement though was consistent--it repeated when I set it all up the same way. Sometimes I moved the apron 16 full revolutions (for a nominal of 12") and it would be short by more than .1" (See snapshot--the 12" caliper has not closed yet even though it started at 12" and had at that point done exactly 16 full revolutions.)

Finally I ran a careful calibration run. Set up my 12" caliper and read it every .1" nominal from the hand wheel dial. WOW! It really is way off.

There is really significant drift--on average the apron moves 1.2% less than the hand wheel dial says. This is pretty close to .4/25.4 so perhaps the hand wheel actually moves the apron 25 mm when it says it has moved it 1 inch.

BUT ALSO, there is major periodic error. Period is 2.675". Peak-to-peak of .074". Meaning sometimes the apron has moved .037" more than the continual drift indicates it should have and then 1.3375" later it is reading .037" less than the drift adjustment says it should have.

The very strong periodic error surprised me. The drift error makes the lathe hand wheel dial much less useful, but at least it can be adjusted for. But, keeping track of where in the periodic error I am along the lathe is way too much puzzling for me to ever do.

So, questions:

1. Is this 1.2% error to be expected in the PM1030V? I had thought it was an inch machine. What is the explanation? Can it be fixed?
2. Is the large periodic error to be expected in the PM1030V? Does it happen in other hobby lathes? Is it a bug in my particular lathe? Can it be fixed?

I guess I can think about working to install a DRO. But that feels both expensive and a bit beyond my skill level. If the hand wheel dial were (basically) accurate it would work just fine for me.

Thanks for any guidance you might have.

-Bill

 

[jaek]

That's a beautiful graph. The periodic error has to be an off-center gear or shaft somewhere in the apron. Should be fixable, no idea how much work though.

I bet that all the lead screws are inches, but not the rack for the carriage. That's the way my mini-lathe is.

 

[bill70j]

Bill:

The hand wheel dial on my lathe is off even more than is yours. The chart I have attached shows mine is off by 2.86% vs. your 1.2% But you can see that I don't have those wild swings. In fact my error is essentially linear with an R2 of 0.9962.

On your graph, the period looks to be 4 revolutions of the hand wheel and the amplitudes seem to be cumulative. So what you may be seeing is inaccuracy introduced as your hand wheel gear meshes with other spur gears before finally reaching the pinion shaft. Like Jaek says, one of your larger spur gears may be off center.

A DRO certainly would be the best solution, or you could continue using a DI. But I think a good alternative would be to use a handle and dial on the lead screw. This is a much finer control, plus provides the accuracy of a fixed TPI lead screw for measuring.

One of my lathes came with that kind of a set up, and I am in the midst of modifying a second lathe that didn't have it. Here are some photos.

HTH, Bill

Here is a graph showing how inaccurate my hand wheel is.

Here is the handle and dial on the tailstock end of the lead screw on one of my lathes:

I have added a handle and am now making a dial for this lathe

 

[7milesup]

Interesting with the handle on the lead screw. I have never seen that before, but of course I'm a newbie also.

 

[higgite]

FWIW, I use one of these on my bench lathe and don't pay any attention to the carriage handwheel dial. Very simple to install.

**Gold-N-DRO** 12" "X" Axis Modified Igaging Magnetic DRO - Universal Kit | eBay

TheGOLD-N-DRO was designed using the Igaging DRO. Instead of permanently mounting the scale and readout to the lathe, I have designed a kit that would allow the scale to be magnetically mounted to the ways & carriage and Readout to the carriage.as you can see by thevideos.

www.ebay.com

Tom

 

[Blondihacks]

I believe others have covered everything that is likely happening here, but I'll throw my 2 cents in. I believe that there are two issues here:

1) The gear train and the rack that move the carriage are still metric in these machines. The main leadscrew and cross-slide screw are imperial, but nothing else is. That trick where they change the markings on the carriage hand wheel to make it look imperial only works up to one rotation. As soon as you crank the hand wheel around more than once, you are introducing error. One “lap” is not exactly 750 thou. It’s like 749.6789 or something (I made up that number, but you get the idea). Every subsequent rotation is going to accumulate more and more error.

2) The periodic error is likely due to imprecision in the gear train and meshing of the rack (or something bent, as has been said). I have not pulled apart the carriage on my PM1022, but I imagine things in there are made to a price point, as suggested by the fit and finish of the change gears. That's not a criticism of PM by any stretch- the machine is a great value. It ain't no Emco Super 11, but it also costs half as much.

I haven't done this test on my carriage, but I honestly wouldn't expect it to hold distances well other than when threading (which is why I've never bothered to check, I guess). The markings on the carriage hand wheel are so coarse that I don't trust them much. When I need a precise length on a part, I always use an indicator anyway.

I hope something in there helps!

-Quinn

 

[DavidR8]

Welcome Quinn! (@Blondihacks)


Sent from my iPhone using Tapatalk

 

[Dhal22]

Welcome Quinn! (@Blondihacks)


Sent from my iPhone using Tapatalk

Indeed.

 

[shooter123456]

2) The periodic error is likely due to imprecision in the gear train and meshing of the rack (or something bent, as has been said). I have not pulled apart the carriage on my PM1022, but I imagine things in there are made to a price point, as suggested by the fit and finish of the change gears. That's not a criticism of PM by any stretch- the machine is a great value. It ain't no Emco Super 11, but it also costs half as much.

-Quinn

As a note, I have pulled apart my carriage and everything inside was leaps and bounds ahead of the fit and finish for the change gears. Everything was much held much tighter and they fit together very well. Several parts looked hardened and ground and looked brand new despite 4 years of use.

 

[Blondihacks]

As a note, I have pulled apart my carriage and everything inside was leaps and bounds ahead of the fit and finish for the change gears. Everything was much held much tighter and they fit together very well. Several parts looked hardened and ground and looked brand new despite 4 years of use.

Hey, that's great to know, thank you! That's reassuring and suggests Bill's situation may be an outlier. I may pull my carriage apart at some point to see if I can improve the feel of the half-nut. The engagement on that is a bit stiff and not very confidence-inspiring on my machine. Probably something that could be fixed with some deburring and polishing of something in there.

And thanks for the welcomes, all! Long time lurker here, but Bill's question inspired me to jump in.