Anyone tried the Accusize 5C collet chuck?

[thestelster]

But before you decry that a lathe chuck is unacceptable, whether it's a 3-jaw, 4-jaw, or collet, you better measure your spindle nose first.

Below is what Accusize requires for you to achieve their stated TIR. (those numbers are metric. The equivalent conversion is 0.00016" & 0.0002")

And Bison wants 0.0001" accuracy on those surfaces.

​

The method of measuring 0269-5C Collet Chucks​

1. CHECK THE RUNOUT OF THE LATHE SPINDLE FIRST. ADJUST THE RUNOUT VALUE OF THE NOSE WITHIN 0.004MM; OF CYLINDER OR TAPER WITHIN 0.004MM SHOWN BELOW.​

And, though you might say, well, I'll get a Set-Tru, or Adjust Tru, or just use a 4-jaw and indicate it in. And yes, that will work, but only at that spot where you are indicating. If there is any run out at your spindle nose, no matter what you do, you will have run out further out on your work piece.

But does it really matter? It depends. If you are machining a part on the lathe, and you do all the work while it is in the chuck, it will come out perfectly fine.

But if you have to take that part out, and then you determine you forgot to machine feature on it, and then you go to re-chuck it, it will not run true for the whole length of the part, even if you re-adjust the set screws, it would only have zero run-out at that spot you are indicating.

So before you decide on anything, measure your spindle nose, and then determine whether a Set-True feature is useful.

 

[Susquatch]

But before you decry that a lathe chuck is unacceptable, whether it's a 3-jaw, 4-jaw, or collet, you better measure your spindle nose first.

Below is what Accusize requires for you to achieve their stated TIR. (those numbers are metric. The equivalent conversion is 0.00016" & 0.0002")

And Bison wants 0.0001" accuracy on those surfaces.

​

The method of measuring 0269-5C Collet Chucks​

1. CHECK THE RUNOUT OF THE LATHE SPINDLE FIRST. ADJUST THE RUNOUT VALUE OF THE NOSE WITHIN 0.004MM; OF CYLINDER OR TAPER WITHIN 0.004MM SHOWN BELOW.​

View attachment 29111


And, though you might say, well, I'll get a Set-Tru, or Adjust Tru, or just use a 4-jaw and indicate it in. And yes, that will work, but only at that spot where you are indicating. If there is any run out at your spindle nose, no matter what you do, you will have run out further out on your work piece.

But does it really matter? It depends. If you are machining a part on the lathe, and you do all the work while it is in the chuck, it will come out perfectly fine.

But if you have to take that part out, and then you determine you forgot to machine feature on it, and then you go to re-chuck it, it will not run true for the whole length of the part, even if you re-adjust the set screws, it would only have zero run-out at that spot you are indicating.

So before you decide on anything, measure your spindle nose, and then determine whether a Set-True feature is useful.

Great advice @thestelster, I always forget that not everyone knows that. But you did more than that by also looking up the numbers and referencing them! That should make the whole process a lot easier!

I should add that SOMETIMES as @PeterT has pointed out, you can rotate the chuck to different mounting studs and cancel out or reduce runout. It isn't something you can count on though.

 

[PeterT]

Actually I am somewhat confused by the Accusize D1-4 direct mount 0269-0014 instructions. The video pertains to ER collets & nothing to do with 5C chucks, but set that aside. I understand the requirement to validate the spindle nose because any runout there just takes the chuck along for the ride. Now if one discovers spindle runout/geometry issues I'm guessing its re-grind or re-dress time, but anyway, I agree, its not the chuck's fault on that basis.

https://www.amazon.ca/Accusize-Integral-CAMLOCK-Mounting-0269-0014/dp/B00U31M6EO/ref=sr_1_5?crid=1CFF51WQABGWO&keywords=5c+collet+chuck&qid=1671577418&sprefix=5c+collet+chuc%2Caps%2C636&sr=8-5

their step-1. What is this picture of? A lathe spindle removed from lathe & held in a test devise, or? (It doesn't look like a lathe headstock). The runout values along the lateral cross plane & nose runout & nose seem reasonable. I can understand if the nose is worn, curved, eccentric, deviated taper angle... any combination of dimensional issues where it is not a geometric cone section axial with the spindle shaft & perpendicular to the vertical face, then the chuck will probably not mount correctly & runout issues will ensue. So far so good.



I don't understand this step as it pertains to a direct mount integrated chuck. Screw lock? Cam lock? Multiple strokes crossly? What is he pointing at behind the chuck?
a/p cross section sketch below, if the rear chuck body element has the spindle nose profile factory ground & boss accurately fits the chuck front body recess, what is there remaining to adjust?




I think they are saying now chuck a reference bar & measure runout at both inboard & outboard position relative to collet nose? Makes sense. A cocked chuck may have low runout near the collet nose but be exaggerated at some distance (example 51mm = 2") away? I guess that is where the 0.0006" comes from. Presumably collet must have now entered the picture (no stated runout)? But this assembly is still not in the lathe?

 

[PeterT]

And, though you might say, well, I'll get a Set-Tru, or Adjust Tru, or just use a 4-jaw and indicate it in. And yes, that will work, but only at that spot where you are indicating. If there is any run out at your spindle nose, no matter what you do, you will have run out further out on your work piece.

But if you have to take that part out, and then you determine you forgot to machine feature on it, and then you go to re-chuck it, it will not run true for the whole length of the part, even if you re-adjust the set screws, it would only have zero run-out at that spot you are indicating.

I think I agree but the devil is in the details

Here we have runout of the spindle face. Blue line shows deviation to red zero datum. Could be wear, trapped debris or maybe the whole spindle is bent. If the chuck is engaged to the spindle via cam locks like this it will yaw the chuck nose down. Runout will be worse outboard than close to the collet as the Accusize picture shows,


SetTru principle (blue) spindle nose profile has runout or eccentricity or undersize wear relative to (red) rotation axis. But the vertical face is accurate. A direct mount chuck would either be influenced by the spindle nose or allowed to float up the face depending on actual condition. A direct mount could deviate to runout (green line) whereas SetTru adjustment screws (green arrows) adjust chuck body axis back to rotation axis like a 4J independent chuck. Unlike jaws of a 4J independent chuck, the collet/chuck position is preserved, so if the collet repeats & D-mount repeats, supposedly the part should repeat. Now I'm not sure how realistic or common or plausible this form of deviation is but I think its the adjustment principle they are trying to solve. Recognizing the SetTru backplate is yet another intermediary plate, it cannot correct gross spindle distortion like above sketch. But OTOH there is no bolt solution that can compensate for a bent spindle condition, floating bearings, things like that.

 

[TorontoBuilder]

Actually I am somewhat confused by the Accusize D1-4 direct mount 0269-0014 instructions. The video pertains to ER collets & nothing to do with 5C chucks, but set that aside. I understand the requirement to validate the spindle nose because any runout there just takes the chuck along for the ride. Now if one discovers spindle runout/geometry issues I'm guessing its re-grind or re-dress time, but anyway, I agree, its not the chuck's fault on that basis.

https://www.amazon.ca/Accusize-Integral-CAMLOCK-Mounting-0269-0014/dp/B00U31M6EO/ref=sr_1_5?crid=1CFF51WQABGWO&keywords=5c+collet+chuck&qid=1671577418&sprefix=5c+collet+chuc%2Caps%2C636&sr=8-5

their step-1. What is this picture of? A lathe spindle removed from lathe & held in a test devise, or? (It doesn't look like a lathe headstock). The runout values along the lateral cross plane & nose runout & nose seem reasonable. I can understand if the nose is worn, curved, eccentric, deviated taper angle... any combination of dimensional issues where it is not a geometric cone section axial with the spindle shaft & perpendicular to the vertical face, then the chuck will probably not mount correctly & runout issues will ensue. So far so good.

View attachment 29119

I don't understand this step as it pertains to a direct mount integrated chuck. Screw lock? Cam lock? Multiple strokes crossly? What is he pointing at behind the chuck?
a/p cross section sketch below, if the rear chuck body element has the spindle nose profile factory ground & boss accurately fits the chuck front body recess, what is there remaining to adjust?
View attachment 29121

View attachment 29131

I think they are saying now chuck a reference bar & measure runout at both inboard & outboard position relative to collet nose? Makes sense. A cocked chuck may have low runout near the collet nose but be exaggerated at some distance (example 51mm = 2") away? I guess that is where the 0.0006" comes from. Presumably collet must have now entered the picture (no stated runout)? But this assembly is still not in the lathe?
View attachment 29132

Here is the correct video

 

[Susquatch]

@PeterT - you are giving me a headache. Lol!

I didn't dig into it nearly as much as you did. I just looked at what @thestelster quoted and said "ya, he is right. I should have suggested that too".

But taking a closer look at the photos provided by Amazon, I have no clue what I am looking at either! The only photo that matters is the squareness and concentricity runout. That makes sense. The other photo prolly doesn't belong.

I have often wondered about your question regarding chuck tilt. I don't have enough info about the interface to know how that is handled or how it works. Perhaps @Dabbler or @RobinHood or @Mcgyver knows. I just picked 3 guys who might know - there are others too but I'd have thought you and I would know too so.......

I did get a tiny chip in the flat nose recess of my 4 jaw once that made a tiny impression. So it's a very tight fit, and I have always assumed it's a contact fit. But if so, how do you ensure concentricity? About all I can say (with considerable shame) is that it works and I have never taken the time to look into it enough to understand how it works. It just does! Sad statement to have to make..... LOL

 

[Former Member]

Yes, that's an issue I had intended to address in my post and then forgot. I'm always trying to find ways to shorten my posts cuz they are usually far too long.

I have a prime membership. Living on a farm makes Amazon attractive. The prime membership saves a ton on shipping and also improves the return experience. Amazon has NEVER given me a problem on anything I bought through prime. But this is also why I prefer to buy from Amazon and then call Accusize to get a replacement or deal with an issue. It is important to deal with the real issue, not some baked up reason to return something. In this case, Accusize gives a precision rating on the collet chuck. So if it doesn't meet that claim Accusize actually prefers to deal with you direct. I think it saves them from getting a black mark from Amazon. In those few times when it happened, I have returned the same item to accusize multiple times. Usually, I give them the reason and they will check their inventory to find me a good one. On one occasion they even provided me with copies of their correspondence with their supplier and I pointed out how the supplier was screwing them. Accusize was exceptionally grateful for my help. I got good parts the next time.

I also understand that Amazon has toughened up their return policy just recently. Hence my earlier comment - no bullshite. If you are guilty of screwing up, fess up and bear the consequences. I believe Amazon respects that and it builds your reputation. On the other hand if you are constantly returning everything you buy, I'd fully expect to be building a bad reputation and eventually get the boot as @TorontoBuilder suggests.

I've had the same experience with Accusize. Actually really easy to deal with.

 

[Former Member]

If I want real accuracy i would use a 3 jaw chuck and grind the jaws in place at the desired diameter one a running lathe. Basically it negates run out and such at that specific diameter because the errors get cancelled in doing this setup.

 

[RobinHood]

The thing with D1-X cam lock systems is that both the nose taper (angle AND amount of protrusion from the reference plane) AND the face of the spindle are critical surfaces. Consequently anything mounted to it must also have these three attributes: angle must be correct, taper must be deep enough and the back face concentric and square to the taper.

If the taper is too deep in the chuck, it will mount fine but won’t repeat.

If the taper in the chuck is too shallow, it will mount crooked to the spindle depending in which sequence you tighten the cam locks. It will also not repeat.

A properly machined chuck will bottom out on the taper, with just a little bit of light shining through between the back face and the spindle face. The cams will then pull the faces together. That is why you do not reef on the locking lugs in one go. I usually snug them up in two go arounds and then a third to tighten them all the way. After two cycles, I tap the chuck with a soft blow hammer. Using a torque wrench while tightening is a good idea as well.

A properly fitting D1-X chuck will not normally fall off under its own weight when all cams are loosened. The taper holds it onto the spindle nose. It takes a light tap with a soft blow to get it to come off. A loose fitting one just falls off.

The mounting instructions for the 5C collet chuck are basically saying the same thing - translated poorly from the original language into English.

 

[PeterT]

Here is the correct video

Thanks for this. My guess is the video is a factory test rig. And Step 1,2,3.. are presumably factory rig calibration, not end user lathe setup instructions. The chuck gets mounted to the spin rig, the test bar gets drawn into mate against the collet taper seat, then indicator shows radial runout and/or angular yaw runout. Maybe that's where the 0.0006" TIR within 2" comes from? I just hope they didn't buy their 5C test bar from the same India outfit I bought my MT3 bar from LOL.

 

[PaulL]

and then call Accusize directly if something isn't right

This is the key. I suspect that even with Amazon's take Accusize come out ahead by shipping to Amazon in bulk - individual orders handling is expensive! Worse if you don't move enough goods to keep a full-time shipping person busy.

 

[thestelster]

The Set-Thru design is only meant to adjust for radial discrepancies. If you want a chuck to compensate for angular errors, (like a "bent spindle nose" , large run-out), then you would need a backing plate that is hemispherical and mated with the chuck body. And if you needed to correct for run-out as well, then you would need a Set-Tru feature incorporated into the chuck design. But all these "gizmos" is just treating the symptoms. If you want the cure, you have to address the cause, I.e. "bent spindle nose", run-out error.

Of course if we want our parts held perfect, radially and angularly then, as @Degen mentioned, grinding the jaws, or soft jaws, or spiders.

 

[PeterT]

The thing with D1-X cam lock systems is that both the nose taper (angle AND amount of protrusion from the reference plane) AND the face of the spindle are critical surfaces. Consequently anything mounted to it must also have these three attributes: angle must be correct, taper must be deep enough and the back face concentric and square to the taper.

100% agree.

Here is a question: lets say you acquire a sub-standard backplate or chuck. The nose taper recess is slightly oblong , incorrect taper, vertical offset or whatever. If you mount & dismount to the spindle repeatedly, is there a chance you are slowly but surely distorting your lathe spindle nose? I always assumed the nose surface was hardened for this reason. I think this even came up in discussion before & opinion was that most are not. @Dabbler kept blocking the path to his lathe & changing the subject when I had my file LOL

 

[Susquatch]

The thing with D1-X cam lock systems is that both the nose taper (angle AND amount of protrusion from the reference plane) AND the face of the spindle are critical surfaces. Consequently anything mounted to it must also have these three attributes: angle must be correct, taper must be deep enough and the back face concentric and square to the taper.

If the taper is too deep in the chuck, it will mount fine but won’t repeat.

If the taper in the chuck is too shallow, it will mount crooked to the spindle depending in which sequence you tighten the cam locks. It will also not repeat.

A properly machined chuck will bottom out on the taper, with just a little bit of light shining through between the back face and the spindle face. The cams will then pull the faces together. That is why you do not reef on the locking lugs in one go. I usually snug them up in two go arounds and then a third to tighten them all the way. After two cycles, I tap the chuck with a soft blow hammer. Using a torque wrench while tightening is a good idea as well.

A properly fitting D1-X chuck will not normally fall off under its own weight when all cams are loosened. The taper holds it onto the spindle nose. It takes a light tap with a soft blow to get it to come off. A loose fitting one just falls off.

The mounting instructions for the 5C collet chuck are basically saying the same thing - translated poorly from the original language into English.

Perfect @RobinHood. Even though I've been using a lathe for 40 years and had a D1-5 for 10, I didn't know some of that.

I don't generally like making work for myself, and it's never been a problem (that I know of), but I like the idea of using a torque wrench on the cam locks. Besides, I hate that dumb T-Wrench anyway..... Seems to me that a click style ratcheting torque wrench would never be a problem other than taking a slight set because it wouldnt get relaxed after each use. Once I know what I want, maybe I'd get a torque limiting extention bar instead.

 

[RobinHood]

If you mount & dismount to the spindle repeatedly, is there a chance you are slowly but surely distorting your lathe spindle nose? I always assumed the nose surface was hardened for this reason. I think this even came up in discussion before & opinion was that most are not.

Well, being the owner of three “previously enjoyed” lathes, i have physical evidence that the taper face can be damaged by careless operators as the larger two machines had embedded chips. They were removed and the faces stoned. The dings remain, however. I don’t know how hard the spindle noses are, but they are hardened and ground.

I don’t think the spindle nose gets worn out if one is careful and keeps things clean. Any distortion while tightening chucks should only be in the plastic range and no permanent distortion should result.

I would attempt to replace a worn out (damaged beyond usefulness) taper ring on a spindle if it came to that. One would have nothing to lose as the spindle is toast without a good taper. Other choice would be to replace the whole spindle.

 

[DPittman]

Thanks for this. My guess is the video is a factory test rig. And Step 1,2,3.. are presumably factory rig calibration, not end user lathe setup instructions. The chuck gets mounted to the spin rig, the test bar gets drawn into mate against the collet taper seat, then indicator shows radial runout and/or angular yaw runout. Maybe that's where the 0.0006" TIR within 2" comes from? I just hope they didn't buy their 5C test bar from the same India outfit I bought my MT3 bar from LOL.

India outfit I bought my MT3 bar from LOL

Hee hee I've got one of those too from likely the same outfit. Dang!

 

[PeterT]

I like the idea of using a torque wrench on the cam locks. Besides, I hate that dumb T-Wrench anyway.....

You could make a shorter version of this extension key - square end for camlock, hex end to match a wrench socket.

 

[Mcgyver]

I have often wondered about your question regarding chuck tilt. I don't have enough info about the interface to know how that is handled or how it works. Perhaps @Dabbler or @RobinHood or @Mcgyver knows. I just picked 3 guys who might know - there are others too but I'd have thought you and I would know too so.......

Robinhood gave an excellent account. I recall hearing or reading that with the tapers engaged, you want a thou clearance on the flat. when the cams are engage is pulls it in tight.

I've machined this sort of geometry and it worked out well using feeler gauges. The rotary lap shown below (used mostly for sharpening scrapers), out of a crappy tire grinder, has the cast iron disks mounted using similar geometry. Its needs repeatability (to take the wheels off for charging) very concentric and without wobble for safety and very very concentric without wobble to keep vibration to a minimum. The AL mounts for the wheels where glued to the shaft and turned in situ.

Its challenging work as .001" DOC of the taper moves the faces .008" closer.

The face and taper mating is a challenge, but the next level up on the challenge scale imo is getting two adjacent tapers to mate; the ie. 3 & 45 degree standard instrument/watch makers double taper lathe bearing. Unlike taper and face, there is no visibility or chance to use feeler gauges.and mates to both must be extremely close tolerances.

I haven't done so, but on a cam lock, i would consider grinding the taper and face in situ and using all new backing plates, as part of a reconditioning. Some of taper mounts, victims of the ham fisted, are less pristine than I'd like. Still quite accurate though....there must be some law of averages for the bruises and warts.

And, though you might say, well, I'll get a Set-Tru, or Adjust Tru, or just use a 4-jaw and indicate it in. And yes, that will work, but only at that spot where you are indicating. If there is any run out at your spindle nose, no matter what you do, you will have run out further out on your work piece.

I agreed with all you wrote, but add, that even with the spindle perfect, indicating the 4 jaw or adjust true in only spot only guarantees concentricity at the one point. For really fussy work you have to adjust in two planes and tap it in. Doing so to any degree of accuracy is a tedious miserable very consuming task, but nessary to make sure the work is aligned with the axis.

 

[Susquatch]

Hee hee I've got one of those too from likely the same outfit. Dang!

I think a bunch of us bought one. I was lucky in the sense that I only bought it for the learning experience. So I had no great expectations and wasn't disappointed. Mine was just total junk. So was the shipping and delivery experience.

You could make a shorter version of this extension key - square end for camlock, hex end to match a wrench socket.

I think it's better to just grind the end of a 1/2 extension bar to fit the chuck and skip the socket.

Robinhood gave an excellent account. I recall hearing or reading that with the tapers engaged, you want a thou clearance on the flat. when the cams are engage is pulls it in tight.

What you guys are teaching me is that I have a horse shoe up my butt. All of my chucks and my back plates (I think I have 8 altogether) fit perfectly. Except for my Bison collet chuck they are all Chinese or Taiwanese. Mind you, my spider chucks are machined after mounting and indexing so who really knows about them.

 

[thestelster]

I haven't done so, but on a cam lock, i would consider grinding the taper and face in situ and using all new backing plates, as part of a reconditioning. Some of taper mounts, victims of the ham fisted, are less pristine than I'd like. Still quite accurate though....there must be some law of averages for the bruises and warts.

Here is a Youtube video of that procedure by Mr. Crispin, he's a machinist working for Rolls Royce: