Questions regarding drilling small holes, and runout.

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Hi, Just joined this (late) so may be off beam with my comments from what I am reading (from the start).
Re: motor vibration. I like Baron's comments: stuff I had not thought about, but:
"Single phase motors do vibrate and this can be an issue, three phase motors less so. DC motors don't vibrate but other parts of the machine may." - "The motor appears to be fine. Running just the motor with no pulley has vibration of course, but a very low acceptable amount I would say."
Single phase motors do have MORE torsional vibration than 3 phase motors, as well, which causes more-pull, less pull, vibration in the belt to the next pulley. DC motors will also have a tiny amount of torsional vibration, as well as a tiny amount of mechanical imbalance (Nothing is perfect in this world)...

"Out of balance pulleys, belts not quite as tight as they should be." Belts are designed so the resonant vibration (from Motor torsional vibration) does not happen at the correct belt tension. But "Resonance" can happen when NOT at the correct tension. Also, pulley sizes are usually arranged so the belt "hunts" or creeps along the pulleys to even out the wear. But even new belts have some (Small) imperfections that will generate some torsional vibration in the spindle. Hopefully insignificant in your "Quality" machine?
Bearings will also generate vibration, but the machine designers will ensure this ia all minimised in their design.
Anyhow, all this is irrelevant:
A runout of 0.1mm is unacceptable, and I should look at bearings FIRST, possible a bent /damaged quill as well? (your 0.1mm runout is half the diameter of a 0.2mm micro-drill! Like a 1/4" runout on a 1/2" drill! = HUGE!)

I often drill holes at 0.3mm, 0.25mm - but that's as small as I go! Your runout (0.1mm) WILL BREAK DRILLS. Been there and done it many times. You need to be as close to perfect as you can get as these drills are more fragile than dry hair.... They look like the tiny bend will be OK, then "puff" the drill has disappeared and there is a bit of very hard drill material blocking the hole...
You need "perfect" drill alignment in the chuck, perfect alignment with the tiny centre spot where you are drilling (Use a magnifying glass to get close and really see how well aligned you are!), then "perfect" straight insertion of the drill along its axis. ANY Misalignment anywhere will simple break drills below 0.5mm, and often larger ones. The materials for drills are not meant to bend under any circumstances. - Ever. The only true cutting edges are the end of the drill bit. Because of fluting for swarf clearance, the drill is just a twisted bit of thin flat - but very hard and brittle - metal. So they flex easily and break even easier!

Centre drills are OK... but not perfect. What I use is the smallest regular drill (maybe 2mm? or 3mm?) to very carefully touch the component (rotating at speed) so it makes just a tiny spot for the sub-millimetre drill to align on the component. The centre of a 3mm drill is smaller than the centre on my centre drills, except the very small sizes. (1/8"). Which are so fragile they are kept for special applications! The trick by using (say) a 2mm drill is that it it so flexible, you can see it wander if the component surface is too rough, not square to the quill, or whatever. But when it makes a tiny spot for starting the micro-drill, you know that is good with machine alignment. I have had small centre drills that will drill off centre and then the micro-drill is the first to show wiggle when it touches the surface... so I don't trust any unless proven OK on a test piece first.

I'll read on now...
K2
 
Very good that you found one potential source for the vibration. It seems to me that the bar that is driven by the handle / ratchet system runs in a bushing. Now I'm in horology, and the name of the game there is to replace worn bushings to make grandfather's clock run smoothly again. Maybe you can do the same here. Steps 1) remove the bar and bushing 2) measure bar and if too much out of cilindricity grind it or have it ground to a nice cilinder again 3) Make new bushing to fit with a running tolerance 4) re-assemble.
 
He is right.
I missed carbide drills. I suppose if you looked at engagement slow motion you might see one drill flute in gage before the other thus causing off set twist and fracture if metal is soft it might not matter as the flute would begin cutting. I’ve seen the vibration in high speed drills so center drills need to be very good . I’ve broken very small carbide drills just redrilling small carb jets . Brass material . Any runout in the spindle all hurt as will off center location. Drill press is probably not the best machine tonuse. Very precise mill would be better choice I YHINK . Ou little grizzly lathe has hard time with 1/8” high speed steel small drills . Some times I YHINK hand held dremel works better . LOL
 
Whatever videos you watched that show marking and then FILING collets due to excessive run out are being produced by literal know nothing morons Gavin. Well made collets get internally and externally ground to a few thousandths of a mm or a few 10ths of an inch after heat treatment, or even less for the much better ones. When the collet through holes get small enough, I suspect there instead lapped for straightness, specific size, roundness and surface finish. That's were the guaranteed maximum allowable run out numbers come into play and the costs start to go up. Exactly how is that hand filing ever going to approach low 10ths or well below level of precision that internal grinding and/or lapping produces never mind attempting to correct inaccuracy that shouldn't even be there if your not buying the cheapest possible collets. Buy from a dealer that specifies the maximum run out collets or a chuck will have and what will also have an acceptable amount of run out for what your doing. Then check them youself to be 100% sure you did get exactly what you were paying for.



Center punching and obtaining anything even close to precision? It's never ever going to happen. And the less experience anyone has, the more inaccurate those layouts and center punching is going to be. Then there's further inacuracy using any drill press no matter what it costs or how good it was or is without also having a good X,Y table under the spindle. Your first post that started this thread and the details you gave. Your current run out is between .08-.10 mm or .003"- almost .004". Obviously that was something you considered worth checking or you wouldn't have mentioned it. I don't care how good anyone thinks they are about doing a hand layout, prick and then center punching the hole location. Once that hole is drilled it's going to be at the bare minimum a few multiples out of position of what you already have for run out unless you happened to be extremely lucky on one single hole. And they would average much worse with a line or circle of holes on the same part. For the small parts most of us are generally doing, that center punching should have died out even the hobby level at least 50 years ago. It vastly increases inaccuracy and that statement comes from my own experience and known inarguable facts. Adding a rear fence to a drill press table, edge finding the part and then moving it against the tip of an indicator would definately help. In that case your using better techniques and methods to increase the accuracy of where your exact hole location will be. Yeah there were and maybe still are a few real machinists that laid out and even center punched, and could still produce high precision parts. They also had years or decades of doing exactly that. Most of us couldn't even get close to what they were doing everyday.

And since spotting drills were also mentioned in this thread, don't bother. They can't and will not provide any better location accuracy with the equipment and methods your using now. On high end CNC's they have there place and some benifits. On manual machines and especially with the smaller one's they don't.
 
Hi Guys,

Just to illustrate some of the issues with collets, not just cheap Chinese ones either !
Non of the collets in these pictures provided a run out anywhere near the specs...
I think that at best they were barely usable even after being cleaned up !

Some of the problems are obvious and were easily cleared using a Stanley knife blade. These came as part of a set.
 

Attachments

  • 26-06-2019-Collets-01.JPG
    26-06-2019-Collets-01.JPG
    221.4 KB
  • 26-06-2019-Collets-02.JPG
    26-06-2019-Collets-02.JPG
    164.7 KB
  • 26-06-2019-Collets-03.JPG
    26-06-2019-Collets-03.JPG
    196.3 KB
  • 26-06-2019-Collets-04.JPG
    26-06-2019-Collets-04.JPG
    484.3 KB
A single typical good German made TOPAC (Top Accuracy) collet with runout within a couple of microns will cost about twice the price of a complete Chinese collet set including colletholder. The difference is that you don't have to throw away the German collet after the first use in total disgust.
 
A single typical good German made TOPAC (Top Accuracy) collet with runout within a couple of microns will cost about twice the price of a complete Chinese collet set including colletholder. The difference is that you don't have to throw away the German collet after the first use in total disgust.


OK - - - - I'll ask the almost stupid question.

I'd love to buy top quality tools - - - that would mean that I just wouldn't be buying.
Just don't have the funds to buy things at that level.
So how about recommendations of collets that are reasonable AND don't break the bank.
There's got to be some someplace (I hope!).
 
OK - - - - I'll ask the almost stupid question.

I'd love to buy top quality tools - - - that would mean that I just wouldn't be buying.
Just don't have the funds to buy things at that level.
So how about recommendations of collets that are reasonable AND don't break the bank.
There's got to be some someplace (I hope!).
A lot of CNC users tend to use a limited variation of collet sizes, instead they use drills or other tools that have oversized shafts.
drillb.jpg

This maker calls the drills "endmill shank drills".
Much more expensive than HSS, but usually high quality. Occasionally I find them 2nd hand.

For very small holes I would get one 1/8 inch collet from a reliable source and use drills with 1/8 shank.
(from certain vendors the 1/8 variant is always much cheaper than the 3 mm type :) for otherwise identical product)

Some collet types and brands are nasty expensive and a complete set would cost a few thousand dollars.
So I only bought the "needed" ones and they were still broke the piggy bank.

Collet nuts can also be the cause of bad performance. The most expensive collet cannot fix "bad everything else".
 
Last edited:
If those collets weren't even close to matching what's in writing for there run out specifications Baron, then they should have been immediately returned to the vendor. I don't try to re-machine, repair or modify what the manufacturer should have already been producing if that's what there specifying them to. And I fully expected not to do when I chose what I did from there own written specifications and what I was then paying for. Ask yourself this, if you paid with Monopoly money, would the tool dealer be fine with that or instead expect to get real money? Then what's the difference if your not getting what they guaranteed in writing for what you were buying.

Sure those very high accuracy collets are available with a price to match. But then what are the run out numbers on the chuck, it's nut, spindle taper & bearings? Then just how rigid is the whole machine there being used with? To get everything your paying for with those high accuracy collets, then all the rest of the components have to be capable of as good or better numbers as well just like Timo already mentioned.

Yeah you could buy the more specialized cutting tools with a few standardized shank diameters to save buying more of those high cost collets. But think about that from a logical perspective. Cutting tools have a limited life span even if there being resharpened. So your going to have to replace any cutting tool at some point. There a consumable item. All your accomplishing by using that limited number of tool shank sizes is saving on the cost of some collets. And from then on and it won't take long, your then spending more for your cutting tools than those extra collets would have cost. Your literally trading one price for a work around that will still end up costing more. It's just spread out over a longer period of time. Plus you still can't use a lot of off the shelf but less costly industrial quality and performance cutting tools that might have shank sizes outide of what your collets can grip properly. Maybe that idea might work in a production situation, but I honestly can't visualize any scenario where it might since there's more than a few other options. In a home shop? Collet chucks, nuts and collets in those industrial situations and when there not using shrink fit tool holders or other methods will still get replaced after enough use or after a serious crash. In a home shop and with a decent set of collets and chuck, it's highly doubtful almost anyone here would ever need to do so for at least multiple decades or possibly in there lifetime. Even if you had to replace collets, then it should still only be a few of the most used sizes that a large number of your tool shanks were used with. Cnc and machines with tool changers, there using at least multiples of each tool shank collet size, multiple chucks, and the tools get torqued into the collet. Then the tools Z axis height and diameter are entered into the tool progam parameters. At that point, the tool itself would rarely be removed from that collet until it's useable life is gone. How we use collets and without those tool changers is a whole lot different than they do with todays high volume part production.

I'd agree, those extreme accuracy brand name collets, chucks and nuts might well cost a few thousand or even more. Is the rest of what you have for machines and tooling capable of justifying the price, or would most here even gain anything with them? How many members here spotlessly clean the spindle taper, chuck shank, male and female nut threads, ID, OD and slits in each collet, chuck interior, collet and nut extraction rings, and at least lightly lube those nut threads. Anyone think that isn't going to be absolutely necessary with those low miron collets? There's multiple ER collet manufacturers. Rego-Fix is probably one of the better known since they invented the ER collet design, and there also known for producing unquestionable quality that will meet or exceed there stated specifications. When I bought my Bridgeport clone I wanted to use ER collets. The ER 32 series would have held any tool shank size a BP and it's R8 taper could be expected to drive. But I also wanted dual duty out of what I was buying and also use them on my lathe. So I spent a fair bit more and bought ER 40's. I also knew I didn't want to pay the extra for those Rego-Fix or Schaublin ER colets.

And with all the videos I'd watched and dozens of forum posts about just how bad those much cheaper off shore collets, chucks and nuts usually are, I refused to buy those at any price. Yes you might even get lucky sometimes and at least a few collets in a set might or might not turn out to be very accurate. I can't afford to trust that would happen, I buy due to the stated specifications and fully expect to get what that higher price had better provide. I also don't blindly trust anything no matter who's brand name is on it. So whenever I can, I personally verify that I did in fact get what the manufacturer stated. So I did my own research, and after enough of that, decided on Bison as the manufacturer and bought a complete set of there 40 series, collet chuck with an integral R8 shank and the nut came with the chuck. All of mine are under or well under .0005" for run out checked with gauge pins and a very good 10ths dti. Mine cost me a bit under $1,000 in Canadian dollars, and likely anywhere in the U.S., Europe or the UK, prices would be even less than where I am.

Bison's chuck and collets are quite close to what a set of Schaublin ESX 25 collets I also own have for run out numbers, and that set of much smaller collets cost almost as much as those 40 sized collets did. In general and other than maybe those Schaublin collets, as the collet series get smaller, the costs drop much lower as well. The only real deficit that 40 series sometimes has for me is the diameter of the chuck and nut. So one further addition I don't yet have would be a set of either ER 11 or 16 just for the much smaller diameter nut and collet chuck that at times would be useful driving the smaller diameter and shorter tools since it would also allow easier access in or around more confined areas. And fwiw, those 40 sized collets can properly grip either imperial or metric tool shanks, and anything from the .125" Dremel tools on up to 1" tool shanks or any part sizes within that range on the lathe. But if your using just about any mini or bench top sized mills, then you don't need nor can you even use either the 32 or 40 series collets. The 25's should be just about all you'd need. And even most of the better known brand name industrial level chucks and collets really aren't that expensive. For any tooling, I've learned to stick with U.S, Eastern Europe, the UK or Japanese manufacturer's. There's also lots from western Europe, but always at very high prices.
 
I've put a ER11 collet chuck on it now and my static TIR with a cheap collet (around $15) seems to be 0.04mm now. Down from 0.1mm. Not too bad. Guess I'll try my operation again now and see how it goes.

Edit: Techniks 04211-1/8.
Not sure how other brands compare.
 
I've put a ER11 collet chuck on it now and my static TIR with a cheap collet (around $15) seems to be 0.04mm now. Down from 0.1mm. Not too bad. Guess I'll try my operation again now and see how it goes.

Edit: Techniks 04211-1/8.
Not sure how other brands compare.
Did you try to rotate the collet a little? Tighten and measure runout again, sometimes it helps.
Runout combines, seat in the nut and taper. The best collet will not make up for a horrible nut, or horrible thread on the chuck.
And I saw some people tighten the nut moderately, and bumping the nut or tool shank with a drift. They claim it shifts the tool a tiny bit. (do not break drills :) )


Good luck.
 
Still not good enough for a 0,25mm drill IMHO. I buy those in sets of 10 x 3mm shanks, and make around 10 ~dozen holes. Each set-up is fraught with fiddling to off-set all the imperfections against each other to find the best compromise so the drill looks good. A really good set-up will drill a few holes, a bad set-up won't drill any. Also with those drills, the higher speeds of a Dremel can cause the cutting end to fly off with centrifugal force if not perfectly centred and true. So fit a drill, run at a lower speed and look carefully. Rotate the collet a 1/4 turn and see if better or worse. and after 6 iterations you'll see the best position for all the bits combined.
K2
 
Still not good enough for a 0,25mm drill IMHO. I buy those in sets of 10 x 3mm shanks, and make around 10 ~dozen holes. Each set-up is fraught with fiddling to off-set all the imperfections against each other to find the best compromise so the drill looks good. A really good set-up will drill a few holes, a bad set-up won't drill any. Also with those drills, the higher speeds of a Dremel can cause the cutting end to fly off with centrifugal force if not perfectly centred and true. So fit a drill, run at a lower speed and look carefully. Rotate the collet a 1/4 turn and see if better or worse. and after 6 iterations you'll see the best position for all the bits combined.
K2
Well the taper on the spindle is at 0.01 runout.
I took off the chuck and cleaned everything well with acetone and TIR appears to now be at 0.03mm. (previously .04)
So either cleaning helped, or maybe the way I fit it counteracted the taper runout?
Or heck, could be that I just rotated the collet a bit when re-inserting. I'll count it as a win for this setup. I at least know that .03 is achievable now with what I have.

Either way, I digress, what runout do you think is required to achieve 0.25mm well? Or essentially zero?
Whatever videos you watched that show marking and then FILING collets due to excessive run out are being produced by literal know nothing morons Gavin.
Noted. I admit it sounds bad at first, but so do many other things that end up being acceptable. It is for this reason that I generally use multiple sources.

Then there's further inacuracy using any drill press no matter what it costs or how good it was or is without also having a good X,Y table under the spindle.
Yes, this is unfortunate. I did purchase and attach a Proxxon KT70 table to it. But this isn't exactly something that I would consider to be high quality. I had a big problem finding xy tables that would fit onto the drill press table. The KT70 is essentially the maximum perfect size to fit.
I did find one Demanders table (very small & swiss). But there's no telling how much they cost.
Probably way more than what I can fork over at the moment.

link to demanders here. Smallest xy I've seen. model FB-H

edit: I see on a distribution site the FB-H selling new for $900 USD. That's gonna be a no😅
At least for now.
 
Last edited:
Just now I had to make a new spindle-chuck connector with two opposing tapers for a 50 year old heavy industrial French Précis tabletop drill. It downed on me that to achieve a 0,01 mm runout at drill point the runout of the connector 120 mm away should be no more then a couple of microns. It took me hours of set-up time to achieve this between centers on my Schaublin 102 VM lathe. Runout at drillpoint is now 0,02 mm. So a runout on the taper of a spindle of 0,01 mm may just not be good enough.
 
I need to make a new pricker for a blow-lamp:
The pricker screws into the end of the valve rod for the blowlamp. So every time the valve is closed, it pricks the jet clean. A previous owner "lost" the pricker.
2 Options for the pricker:
Turn a 0.25mm wire end (5mm long) onto a 9BA screw (1.8mm long!), or
Drill a hole and insert a steel wire into a 9BA screw. Like clock-making. A fiddle!
So: I have been trying to drill a 0.25mm hole in a mild steel rod. Failed totally 4 times. The drill penetrates ~1mm then simply "explodes" - so I have to re-face and centre-drill and try again. I guess the torque for Mild steel is more than the drill can handle? - I use a slight pecking action (I have seen the drill buckle but not break doing this on brass - I know then I am too heavy handed!). I was making about 0.002~0.004in. progression at each peck... The steel was a bit of a bitch, as trying to machine the 0.25mm pricker for a blowlamp jet, I only managed to get down to about 0.7mm dia before the end of the prick I was turning simply broke into small pieces... I had expected it may bend and break off, but it was brittle and broke in my fingers... Wierd steel? - I'll try again using some decent steel or brass.
Needs another couple of hours to manage this one! - My good eye can only manage to see the 0.25 mm drills for less than an hour or so before I can't focus properly. (using 3 x magnification lens over my prescription glasses). There is no "feel" when drilling this small. - Like threading a needle!
Possibly I shall admit defeat later?
K2
 
Back
Top