Problem with chuck runout

[bertterbo]

Interestingly one of the reviews for the axminster chuck you have claims runout as 0.04mm, although not clear what they are referring to exactly. Axminster dont appear to give figures, or not that I could see. Before you go throwin more money at the problem i would confirm your existing meaasurements. So make sure the arbour and spindle bore are spotlessly clean and install the arbour in the spindle. Mark the realtive positions of arbour and spindle with a sharpie or similar. Use your dti to take a runout figure at the top of the projecting arbour taper, and repeat at the bottom. Make sure your dti is at right angles to the surface being measured. Ideally the figure should be the same for both readings. If there is a big variation then either the bore in the spindle is at an angle or the arbour is not straight. Mark the high spot on the arbour with a sharpie, ie where it is closest to the dti. Now take the arbour out and re install it 180 degrees from the previous position relative to the spindle. Repeat your measurements, and note where the high point is, relative to where it was before. If the readings are the same but the high point has shifted 180 degrees then your arbour is straight, and the error is in the spindle. If the readings have changed appreciably then there is an issue with the arbour. To check the arbour properly you really need to mount it between centres. If you cant do this then roll it on a flat surface. Even very small errors can be quite apparent. Hope this is helpful.

I'll try this.

I do actually have a lathe, but its MT2 not MT3. How would testing between centers help? as aren't you then introducing error from the tail stock?

 

[Fergie 307]

Re reading my post I may have inadvertently caused you some confusion. If the readings you get from the top and bottom of the arbour taper are different this indicates that one or other bore, or arbour is at an angle. The second part of my descrption assumes that this is not the case but you have the situation where the error is essentially the same along the length, this indicates that the bore and spindle are not concentric.

 

[Fergie 307]

I'll try this.

I do actually have a lathe, but its MT2 not MT3. How would testing between centers help? as aren't you then introducing error from the tail stock?

You need to ensure that your centres are correctly aligned. Your arbour should have a recess either end to accept centres spinning it by hand between centres against a dti on the bed will tell you if its straight. It will have been ground between centres in manufacture, and to be honest should be pretty accurate.

 

[Fergie 307]

I'll try this.

I do actually have a lathe, but its MT2 not MT3. How would testing between centers help? as aren't you then introducing error from the tail stock?

You need to ensure that your centres are correctly aligned. Your arbour should have a recess either end to accept centres spinning it by hand between centres against a dti on the bed will tell you if its straight. It will have been ground between centres in manufacture, and to be honest should be pretty accurate. My concern is simply that before you buy yet another chuck you need to be absolutely sure this is the problem. If you have a situation where top and bottom readings on the arbour chuck taper are, say 0.1 mm with the high point of both in the same orientation, but a reading off a straight bar in the chuck is 0.3, then the error MUST be being caused either by an error in the chuck itself, or in the interface with the arbour. This could be dirt, corrosion or a bruise or other defect in one of the mating surfaces. Since all the parts are new we can hopefully rule out these issues, but you do need to make sure the surfaces are spotlessly clean.

 

[Fergie 307]

If you find there are small errors in concentricity then you may be able to reposition the chuck to get the error to the minimum possible, then just mark the parts in some way so you can always put it back in the sweet spot. All depends on the level of accuracy you feel you need. It is very irritating when a problem like this spoils your enjoyment of your new machine.

 

[bertterbo]

So, back with some results

- When testing the arbor only (no chuck connected), I get 0.005mm for the one orientation, and about 0.01mm for the other

- When testing the chuck with a straight rod inserted and measured about 10mm from the chuck jaws, for 4 orientations, 90 degrees apart, I get 0.2mm, 0.23mm, 0.25mm and 0.32mm ... and this was with the arbor in its best orientation.

 

[Fergie 307]

Ok so what are you referring to by orientation? First important thing is whether the run out at the top and bottom of the arbour taper is the same. If these figures are the same, and it is closest to the dti in the same place, then at least you know that the axes are parallel. So what are those figures?
If you mean that you have the same runout top and bottom of the taper, but it is 0.005 in one position, and 0.01 when repositioned by 180°, this indicates that the arbour may be the problem, assuming, and this is important, that you are turning the arbour from the point at which you got the biggest or smallest error, or when it is closest or furthest way from the dti. Doesn't really matter which as long as you always use the same point of reference for comparison.
If you think about it, if the bore in the spindle was off centre by, let's say for sake of argument 0.1. Then, assuming the arbour was perfect, you would see a run out of 0.1. If you marked the arbour and spindle at the closest point to the dti, then turned the arbour through 180°, you would get the same runout, but the closest point would be 180° from the initial point in the arbour, but in the same place on the spindle. So the error would be following the spindle rotation rather than that of the arbour, indicating a problem with the spindle.

If on the other hand the problem is with the arbour, then the error will follow the position of the arbour, rather than the spindle, assuming in this case that the spindle is perfect.
The most important thing is that you are getting the same run out at the top and bottom of the taper. If you aren't than this indicates that the axis of the arbour is at an angle relative to the spindle axis. This error will then be magnified the further away from the spindle you get. So you have measured the error on the bar 10mm from the chuck, what does it look like say 50mm out? If it is bigger the further away you get then your rod is at an angle.
Hope that all makes sense.

 

[Fergie 307]

If you have a number of arbours from different sources, then worth repeating the measurements on the arbour taper with another one. If all give similar figures then very doubtful its the arbour. Regarding your measurements of the arbour taper, mark the high and low points on the arbour and turn it 90 degrees at a time relative to the spindle and repeat. If the high and low spots remain in the same place on the arbour then the error is following the rotation of the arbour, which tells you the arbour is the issue.

 

[bertterbo]

It's really quite difficult to measure the arbor as the changes are so small, for the most part <0.01mm (less than half a thou), my DTI only has 0.01mm increments, so even if I do try to measure between those increments, I'm not sure if they can be trusted?

If it is bigger the further away you get then your rod is at an angle.

That's always going to be the case though.

To be honest. I'm getting a little tired of messing around with this now. I'm just going to use the chuck with the best results (the cheap scroll chuck) and just call it quits. It's really annoying and disappointing that my new expensive drill has more runout than my last, but I'm just hoping its all down to the chucks and in future I might get a better one.

 

[bertterbo]

If you have a number of arbours from different sources, then worth repeating the measurements on the arbour taper with another one. If all give similar figures then very doubtful its the arbour. Regarding your measurements of the arbour taper, mark the high and low points on the arbour and turn it 90 degrees at a time relative to the spindle and repeat. If the high and low spots remain in the same place on the arbour then the error is following the rotation of the arbour, which tells you the arbour is the issue.

The arbor only has two positions.

The problem is that there is always going to be error as nothing is ever going to be perfect. So sure I can measure as you say, but that doesn't tell me its the root of the problem

 

[Ttrees]

I'd love to know if you guys with accurate machines
notice more error when turning the chuck by hand in reverse,
say at least checked with a large bit lowered into the table bore,
although the test I made demonstrates this better.

If you do notice an oddity, then anyone care to have a go at speculating or answering why?
Just saying as it might hint at something else at play aswell.

Hard to get an answer because the folks in this game, generally have much sturdier machines, or just use their lathes for this kinda thing.
In other words they aint got time for that, so nothings concrete yet from what I can make out
in regards to getting one of these generic machines a whole lot better.
Or they just haven't documented this for whatever reason.

I was working on some techniques regarding using my drills possible theoretic features,
and done some testing of the possible use of the table in regards to utilizing it as a face plate,
As is, there was too much slop and noticeable movement regarding cinching it up,
but if sorted, possibly with some buiscuit tin or whatever and some filing/lapping,
the same thing could be thought about using a lazy susan with a square table.
This may well be the seceret answer in regards to getting it all tuned up,
i.e those MT's using some fancy adhesive backed abrasive like the 3m stuff.
Edit: the face plate feature might not be of use here, but thinking along these lines is likely
a possibility in regards to what I'm sayin.
I needed some hold downs, so considered these could be used largely for this purpose in future, should it make sense.

Just a thought from a non professional optimistic bodger who's got time to think about it.
Hopefully someone might chance answering why my test results are as such.

Cheers

Tom

 

[Fergie 307]

I know where you are coming from, it is very annoying. In your video the lighting isn't great but it does look as though your rod remained parallel with the square, just the distance between the two varied. That does at least suggest that everything is straight, and it's an issue with offset. My money is still on the chuck as well, and your measurements tend to suggest that is at least a significant part of the problem. Maybe someone can lend you one for comparison at some point, before you shell out more. These things can be very time consuming to get to the bottom of and I sympathise with your frustration.

 

[Fergie 307]

I'd love to know if you guys with accurate machines
notice more error when turning the chuck by hand in reverse,
say at least checked with a large bit lowered into the table bore,
although the test I made demonstrates this better.

If you do notice an oddity, then anyone care to have a go at speculating or answering why?
Just saying as it might hint at something else at play aswell.

Hard to get an answer because the folks in this game, generally have much sturdier machines, or just use their lathes for this kinda thing.
In other words they aint got time for that, so nothings concrete yet from what I can make out
in regards to getting one of these generic machines a whole lot better.
Or they just haven't documented this for whatever reason.

I was working on some techniques regarding using my drills possible theoretic features,
and done some testing of the possible use of the table in regards to utilizing it as a face plate,
As is, there was too much slop and noticeable movement regarding cinching it up,
but if sorted, possibly with some buiscuit tin or whatever and some filing/lapping,
the same thing could be thought about using a lazy susan with a square table.
This may well be the seceret answer in regards to getting it all tuned up,
i.e those MT's using some fancy adhesive backed abrasive like the 3m stuff.


Just a thought from a non professional optimistic bodger who's got time to think about it.
Hopefully someone might chance answering why my test results are as such.

Cheers

Tom

I think the answer is probably that a chuck is not that precise a holder, certainly a drill chuck anyway. If you want a greater level of accuracy then collets are the answer. But of course a chuck is much faster, cheaper and more convenient, and where absolute accuracy isn't necessary it is the ideal solution. So for example on my metal working lathe I have a cheap Chinese 16mm chuck. I don't know how far it runs out, doesn't matter because I only use it to drill out undersize, then finish to size using a boring tool. If I need a bore that is too small for the boring tools I have, then I would mount a collet chuck in the tailstock and use an end mill. I use the Chinese one as it is an old fashioned keyed one and much easier to use that the fancy Jacobs high precision job. The high precision ones are great if you want the convenience of a chuck, but with greater accuracy and good repeatability. As you say with most chucks you often have to jiggle about a bit to get them properly seated and the drill bit running truly straight. The really annoying thing for Bertterbo is having an £800 machine with a £100 chuck and it not running very accurately, I think he has every right to expect better. Really not good enough

 

[Ttrees]

Thanks for that @Fergie 307
I will check this out and see if I can notice, or find a way to notice the jaws moving.

Agreed, a machine bought for that much dough, with fancy chuck should have a better tolerance than this.
I hope Betterbo gets this sorted.

Tom

 

[Fergie 307]

For anyone wondering why I would say a keyed chuck was easier to use than a keyless, it's just down to the machine. On my lathe it is not always easy to get sufficient grip on the keyless chuck without sliding the tailstock right back. Positioning the keyed one with the keyhole vertical is just more convenient, for me anyway.

 

[bertterbo]

Axminster have agreed to send me another chuck to try ... third time lucky?

Also, for what its worth, despite me being really disappointed with this drill press, I can't fault Axminster Customer service.

 

[Fergie 307]

Thanks for that @Fergie 307
I will check this out and see if I can notice, or find a way to notice the jaws moving.

Agreed, a machine bought for that much dough, with fancy chuck should have a better tolerance than this.
I hope Betterbo gets this sorted.

Tom

I tend to get the jaws just nipped up, then release quarter of a turn on the key and do it up again fully. Not sure if it really makes a difference, or just my imagination. I do notice from time to time you can do the chuck up and when switching on the machine it looks like the drill is bent. Take it out and roll it on the bench, nothing wrong with it. Scratch head, put it back in, and it runs straight as a die. The back off and tighten again idea does seem to prevent this, so it only happens very occasionally, and then only if I'm being impatient and forget. Can only suggest that on occasions the jaws don't all move precisely together? Can't say I have ever noticed the reverse issue.

 

[Fergie 307]

Axminster have agreed to send me another chuck to try ... third time lucky?

Also, for what its worth, despite me being really disappointed with this drill press, I can't fault Axminster Customer service.

Fingers crossed for you.

 

[Daniel2]

Axminster have agreed to send me another chuck to try ... third time lucky?

Also, for what its worth, despite me being really disappointed with this drill press, I can't fault Axminster Customer service.

If I have understood correctly all the correspondence so far, there appears to be
nothing essentially wrong with the drill press itself, but rather points towards
questionable chucks and/or arbor.

 

[Fergie 307]

Indeed. You would have thought that a £100 odd chuck would be made to a reasonable standard, but the way the readings increase when taken from a bar held in the chuck certainly suggests that this is the problem. Hopefully the replacement will be an improvement. Otherwise I wonder if Axminster would be prepared to send someone out to look at it for him and see what they make of it.