Aggravated with NOVA chucks

[William C Rogers]

Kenneth, I don't think the problem is the chuck if you are only seeing .001-.002" runout on the body. Anytime you remount you will most likely get runout. You can minimize this my marking the position in the jaws. An easy way to check the jaws is ti close them on a steel ruler and measure the runout. I'm not sure why you are remounting. Once I have my wood in the chuck it normally doesn't come out until I'm done and will not need to put it back in the chuck.

I have 4 chucks and I measure as stated above. All of the jaw faces were less than .002". HTC 125 (taper insert) runout was .0035, Nova G3 ( direct thread ) was .003". My Barracuda 2 and 4 both measured less than .001" runout. I've never had any problems with the Barracudas, but I usually use my HTC on my PM 90 and G3 on my midi's.

[Leo Van Der Loo]

Yes the problem shows up when remounting the part or when turning it end for end. I turned a small bowl and the rim on one side was about .25" thick and about .21" thick on the opposite side of the bowl. I thought it was something I did wrong until I came across this board. Also the chuck body has only .001-.002 of run out so the run out is being caused by so-so machining in the slide jaws or the grooves in the chuck body they ride in. Now that I know about the run out I can work around it... I think..

Ken you are really muddling this thread up, you have a chuck that according you has 1 or 2 thou runout, while this thread is about tremendous runout in the chucks, you definitely need some lessons on how to turn your tenon or recess properly before you remount your piece, as there is your problem IMO.

compare.jpg

[John Brown]

Ken, I think your remark about "so-so" machining is a bit inappropriate, if the machining is to close the jaws will lock up because of dust, you must understand we are working with movable wood. You aren't making a Swiss watch.

[Greg Parrish]

I don't see any major issues with the machining on the nova, but that doesn't mean accurate machining was done or that their quality control is there making sure each product is produced with the same tolerance specs. No clue.

Not to get the thread too far off track here, but as the original poster I realize there has to be some give in tolerance on these and it's a matter of how much. However, that tolerance should be a planned, machined in tolerance, and not just the result of sloppy manufacturing. Most places I've read tend to say .001 or .002 is acceptable with .003 being on the higher end of what to expect. But my readings seemed to be much higher at .013 and in one test .035.

The interesting thing is that I can improve it slightly if I remove the spindle adapter and reinstall it, or if I remove the jaws and reseat them according to the method mentioned earlier in this thread, but with none of those processes can I get the chuck measurements down to 3 thousands or less.

Now I might be measuring this totally wrong, but I can visually see the wobble when turning a known round item, like a steel item. The small measurement at the chuck amplifies as I move away in both directions which means an exponentially out of round turn at the edges of a large platter.

The big questions still remain: 1). Am I the issue? 2). Is it the jaws? 3). Is it the spindle adapter? Or 4). Is it the chuck itself?

the answer? No clue. Too many variable to narrow it down yet.

The process? When I get the vicmarc direct thread chuck I'll measure for comparison of a different brand. That will help me rule out a few things, like me and the lathe for example.

[Michael Mills]

According to data which Nova did have on their site they are machined to .004 run out or less. I have only measured on the chuck itself and all except one (of seven) met the guidelines. I may have gotten that one down (.007) if I had fiddled with it a bit but I never tried.

[Kenneth Boyd]

Ken you are really muddling this thread up, you have a chuck that according you has 1 or 2 thou runout, while this thread is about tremendous runout in the chucks, you definitely need some lessons on how to turn your tenon or recess properly before you remount your piece, as there is your problem IMO.

compare.jpg

Lessons would be good..

[John K Jordan]

...Now i I might be measuring this totally wrong, but I can visually see the wobble when turning a known round item, like a steel item. The small measurement at the chuck amplifies as I move away in both directions which means an exponentially out of round turn at the edges of a large platter. ...

Keep in mind that scroll chucks are inherently "sloppy". You may always see wobble holding a round steel object in a self-centering scroll chuck.

Consider precision turning with a metal lathe. When mounting a round metal object in a metal lathe for precision work you really can't use a self-centering chuck. Scroll chucks are usually too "sloppy" due to the necessary tolerances in the complex mechanism. It is very quick to mount a piece in a scroll chuck but it will almost certainly have runout. I use one when I don't really care if the round stock has runout since I'm going to cut it round anyway.

BTW, most self-centering (scroll) chucks for metal lathes are 3-jaw chucks. Due to the geometry three jaws will always grip a round piece of metal securely and evenly with each jaw.

There are 4-jaw scroll chucks available for machining but they fairly rare in machining. The problem is the centering tolerance issues are the same as for the 3-jaw scroll chuck. The 4-jaw self-centering chuck may not grip round metal stock securely since there is no way to get all jaws to apply the same pressure. A self-centering 4-jaw chuck works well with wood since the wood deforms. Metal work cannot rely on the jaws deforming the metal to grip securely.

For precision turning on a round metal object it is often with a 4-jaw chuck with independently movable jaws. A machinist can get right on the money or very close by adjusting the jaws (somewhat) independently and checking with a dial or dial test indicator. Note that this takes some patience. 4-jaw independent chucks are also good for holding square, rectangular, or oddly shaped work. I'll use this 4-jaw chuck on my metal lathe when I need an existing part to be centered without runout, say to drill a perfectly centered hole.

(Disclaimer: I am not a professional machinist. I do have some milling machines and some metal-cutting lathes but I am a hobbyist, learning by reading and doing.)

As for your questions about what is the issue, the process of elimination may be expedited by enlisting outside experience.

JKJ

[Greg Parrish]

Current tracking shows expected delivery of the Vicmarc VM120 on Friday via good ole usps priority. Looking forward to seeing it in person. As always, feel like a kid in a candy store with a new toy in route.

[John Keeton]

Congrats on the new chuck and I hope it arrives soon. That makes all of this discussion moot, but I wanted to add one comment for whatever benefit it may be to others.

You stated - "Now I might be measuring this totally wrong, but I can visually see the wobble when turning a known round item, like a steel item. The small measurement at the chuck amplifies as I move away in both directions which means an exponentially out of round turn at the edges of a large platter."

To do this test it appeared you chucked the round rod in the inside edges of the body of the jaws. Although I will occasionally use that method for temporarily chucking a piece WITH TAILSTOCK SUPPORT, it is not the intended method of chucking. The bearing surface is rather thin and has no serrations or dovetail. It is really easy to clamp a round object canted off center because of this. Combine that with a question of whether those surfaces were ever machined true and the whole measurement method becomes flawed.

[Greg Parrish]

I agree John. This is not an easy subject area and there is not really an easy way to measure these things given the nature of their design, the geometry of how they work, and the lack of a super precise specimen to turn for testing in each chuck. I also wish I had kept the supernova 2 longer to do actual comparative testing side by side as my suspicion is that it was more out of round than my G3, but that is not a scientific measurement of course.

To my surprise, the Vicmarc 120 showed up today via usps, which means 2 days from NV to FL for delivery. Not bad at all as tracking indicated tomorrow.

After giving it a quick spin and comparison during my lunch break I can offer a few observations so far:
> Visually and physically it is huge compared to the tiny G3
> The fitment into the spindle is much better and smoother with the Vicmarc. It will seat completely and easily against the collar at the base of the spindle thread. The nova adapter does not want to easily thread all the way on.
> The vicmarc visually shows less wobble (I realize this is subjective)
> To compare quickly, I took my same 16MM and 17MM sockets and chucked them into each chuck (without the jaws installed) and turned. The vicmarc was able to attain a near perfect visual turn of the socket. I didn't get a chance to measure with the dial indicator.
The nova couldn't get a non wobbly turn with the same method. I am attributing this to the shape of the inside of the jaw slides to give benefit of the doubt to the nova though.

My plan is to put the standard jaw set that came with each chuck back onto the chuck using the factory manual instructions. Then to chuck up a known round object like one of my faceplates (known to be round due to my test results while measuring them on the lathe turning with the dial indicator). Then I can test the barrel of the face plate for run out while it is held in the jaws of the chuck. Unless someone has a better idea for a common item I would have in the shop, this provides a basic round item I can seat fully into my chucks jaws for measurement with the dial indicator.


I'm still willing to entertain the fact that I might be the issue but I'm thinking measurements still point to my copy of the nova being a little more than normal out of tolerance.

Congrats on the new chuck and I hope it arrives soon. That makes all of this discussion moot, but I wanted to add one comment for whatever benefit it may be to others.

You stated - "Now I might be measuring this totally wrong, but I can visually see the wobble when turning a known round item, like a steel item. The small measurement at the chuck amplifies as I move away in both directions which means an exponentially out of round turn at the edges of a large platter."

To do this test it appeared you chucked the round rod in the inside edges of the body of the jaws. Although I will occasionally use that method for temporarily chucking a piece WITH TAILSTOCK SUPPORT, it is not the intended method of chucking. The bearing surface is rather thin and has no serrations or dovetail. It is really easy to clamp a round object canted off center because of this. Combine that with a question of whether those surfaces were ever machined true and the whole measurement method becomes flawed.

[Eugene Dixon]

Reverse chuck a drill bit, i.e., the (smooth) shaft protruding from the chuck jaws.

[Mike Goetzke]

Now I might be measuring this totally wrong, but I can visually see the wobble when turning a known round item, like a steel item. The small measurement at the chuck amplifies as I move away in both directions which means an exponentially out of round turn at the edges of a large platter.

Think this is still linear magnification not exponentially.


Mike

[Greg Parrish]

Think this is still linear magnification not exponentially.


Mike

If you want to get technical with the proper terminology. LOL

I do not see any way for it to increase exponentially and agree that it would be a linear measurement magnified with larger movement at the ends.

[Mike Goetzke]

If you want to get technical with the proper terminology. LOL

I do not see any way for it to increase exponentially and agree that it would be a linear measurement magnified with larger movement at the ends.

Then you agree its linear now? LOL

[Greg Parrish]

Then you agree its linear now? LOL

Well, I think the word "exponential" was used to convey the fact that it is more visible the further away you get. The small measured movement is magnified at distance.

Using the words properly though, I think you are correct in saying it is linear as the movement is not being increased along the length of the distance but rather magnified at distance.


Regardless, the semantics don't change the fact that the initial measurement still appears to show that it isn't in balance or in round. I'll attempt to do some more comparative measurements between the two this evening though just to see. Should be interesting if nothing else.