-
Scam Alert. Members are reminded to NOT send money to buy anything. Don't buy things remote and have it shipped - go get it yourself, pay in person, and take your equipment with you. Scammers have burned people on this forum. Urgency, secrecy, excuses, selling for friend, newish members, FUD, are RED FLAGS. A video conference call is not adequate assurance. Face to face interactions are required. Please report suspicions to the forum admins. Stay Safe - anyone can get scammed.
CX701 gearbox questions
- Thread starter PaulL
- Start date
I was thinking about that, but I worry that my dog might lose contact with the drive pin when I stop the spindle, which I've seen happen the my first (and only) try at turning between centers (to true up my tail stock). This seems like another way to spoil the work, though a bit of discipline I could make sure the dog was in contact with the drive pin before each threading pass.
You could wire the dog to the drive pin with tie wire, or tack it, if you have a welder
It is clear to me that you think of yourself as a beginner. But you are really doing quite well. My read of your excellent desciption of what is happening goes like this.
1. Your shaft is plenty long enough to be able to chuck your part in the 3 jaw. Any run-out on the chuck is unlikely to be bad enough to cause a problem on what is probably a 1/2 or 5/8 shaft (I don't recall what you said it was). In any event, you can keep re-clocking the shaft to get the best runout possible when you swap ends. Also, if you cut the big end first, the runout on the small end will get partly consumed by the size change. Whatever runout there is left won't matter when the part is installed.
The thread on your metric end is short enough to be able to chuck the part on the shaft and not damage the thread on either end. Just push the thread into the spindle further than the Jaws reach. If I am misunderstanding, you can always chuck right onto the threads by making a sleeve out of a pop can to protect them.
ALWAYS use a thread gauge on a very very light threading cut to check the thread pitch after that first light pass. If it's wrong, you change the settings, put dykem on the part and try again. Doesn't matter how sure you are. Check it anyway until it's right.
I trust you are only using the cross slide to back the threading tool out and return it to the same place. Threads can be cut with a plunge cut using the cross slide, but it's not good practice. The depth of cut is set using the compound which is set at 29.5 degrees to cut on the leading edge of the threading tool. You must use geometry to calculate the advance required to cut to the right diameter. (or use trial and error)
I hope that helps a bit.
If not, send me a pm with your phone number and I'll call you and walk you through it.
Once you get your head around the process, I promise it will become easy. I enjoy threading. Perhaps more than any other standard lathe activity.
That's the kind of thinking I come here for. Yeah, tie wire is the perfect solution to keeping these in sync. I'll try it.
I'll be honest, I really don't like this approach. As the thread cut gets deeper the load on the wire and dog will increase which could allow the part to rotate and skew the thread. As I said above, I really doubt your three jaw is out enough to screw up your threading and what little there is can be accomodated by cutting the bigger shaft end and thread first.
That's not to say that a dog won't work, just that it's something else to go wrong and it is not needed to do good work for your application.
Last edited:
I've gotten to about 4 thou on a part I put back in. It's probably good enough for this project, but given I get to less than 5 tenths when I set it up in the 4-jaw... If only it was less painful to switch the bolt-on chucks. For some obnoxious (cough *cost*) reason they included only 3 studs to hold the chuck base plate to the spindle, so I also have to move the studs across every time I change over. Getting faster at it, but also more frequently just chucking straight up in the four-jaw.
Thanks.
That's one I've learned now. I think I'm still too heavy-handed on the touching-off so my test passes feel like they are leaving too much mess. But I think that's explained by your next point:
I was trying to "simplify" my process and drop the compound out of the equation, and I see now from your description how I can return the cross-slide reliably to the same spot, separating depth of cut from position of slide by using the compound. That makes sense of the more complex setup that I screwed up early on. Separating the actions will reduce the error potential. Thanks.
But the load would be on the pin not the wire, the wire would just keep the pin in contact with the dog if he were to reverse the spindle.
I was thinking about this more Im not sure that it would even matter, someone correct me if I am wrong here, but the spindle is directly connect via gear reduction to the lead screw, say you reverse the spindle to ok go backwards, the dog slams to the other side the the pin, no biggy your not cutting anything yet, you run the lathe forward to continue cutting your threads, the dog slams back to the other side, everything should still be in time, you just have ~355 degrees of backlash to take up, I don't believe threading between centers is some Voodoo magic...I'm sure that's how they did it before Chuck's were a thing
But everyone is right Realistically the run out of his 3 jaw isn't a big deal for this application, it was just an idea and maybe a learning project
Yes, thinking of it as 355 degrees of backlash makes sense. My concern was that as I start turning, I often wind up "turning" but without the dog picking up yet - I had enough friction in the spindle-end center to drive the work without the dog, though not enough to drive while cutting, so I can see the cut starting with a judder and a slip as the dog engages.
I agree on the learning opporutnity - there's a lot to learn here!
Holding a threading die in the tailstock may not be as accurate as one might think either. There a few sources of potential error creep & if they collectively stack in one direction, make for less than ideal threads. You may not notice it as much on larger threads but could on smaller/finer pitch:
- tailstock position relative to spindle axis. Hopefully you check this often but they can drift, which then moves the die relative to part rotating axis & cuts the threads unequally on the shaft. Most commonly TS can drift in/out over time. But someone posted a review video of brand new BB hobby lathe where quill center was high by 6 thou. He was under the impression this was to account for future bed wear. Nope, sorry, that 100% QC problem. He rectified by resurfacing base & got it to within 1/2-thou, much better.
- die in die holder. Most dies have some sort of adjustability so they can cut slightly tighter/looser threads. Die holders account for this by slightly looser ID. But that means the OD is no longer circular so when die holder retention set screws are engaged, another source of off-axis. This isn't a problem when hand threading because the die self-centers. But on a lathe setup in TS you have now locked this axis.
- die holder. Some of the aftermarket barrels are so-so quality. There's the barrel, which slides on a shaft, which may be held in a chuck or integrated into MT taper. I think runout would be pretty low, but you get what I mean, a thou here, a thou there....
As the shaft OD increases & material gets tougher, dies start requiring a lot of gronk because its a lot of material to cut even though that's what dies are designed to do. You may find the shaft starts slipping in the chuck & other issues. If you're making a batch of parts, best thing you can do is set up for threading, then rinse & repeat swapping the parts in, thus preserving your compound angle setup & depth of cut etc.
- tailstock position relative to spindle axis. Hopefully you check this often but they can drift, which then moves the die relative to part rotating axis & cuts the threads unequally on the shaft. Most commonly TS can drift in/out over time. But someone posted a review video of brand new BB hobby lathe where quill center was high by 6 thou. He was under the impression this was to account for future bed wear. Nope, sorry, that 100% QC problem. He rectified by resurfacing base & got it to within 1/2-thou, much better.
- die in die holder. Most dies have some sort of adjustability so they can cut slightly tighter/looser threads. Die holders account for this by slightly looser ID. But that means the OD is no longer circular so when die holder retention set screws are engaged, another source of off-axis. This isn't a problem when hand threading because the die self-centers. But on a lathe setup in TS you have now locked this axis.
- die holder. Some of the aftermarket barrels are so-so quality. There's the barrel, which slides on a shaft, which may be held in a chuck or integrated into MT taper. I think runout would be pretty low, but you get what I mean, a thou here, a thou there....
As the shaft OD increases & material gets tougher, dies start requiring a lot of gronk because its a lot of material to cut even though that's what dies are designed to do. You may find the shaft starts slipping in the chuck & other issues. If you're making a batch of parts, best thing you can do is set up for threading, then rinse & repeat swapping the parts in, thus preserving your compound angle setup & depth of cut etc.
Last edited:
Jezz, I sure hope I didn't come across as dogs = voodoo magic! LOL!
I only wanted to point out that there is a reason everyone uses chucks nowadays for most operations instead of dogs. Dogs are not fool proof especially when things are already complicated. Anyway, that decision is just a choice, all the methods discussed will work.
Perhaps I am also biased by the fact that I just finished making a tool post bolt the same as the OP is making. My choice was to use a collet chuck for the job but if I didn't have that, I would have used my 3 jaw and the end result would have been the same.
Glad you figured out the tool geometry, that should be a huge help making a good thread. For this job, you could set the compound for 30 degrees instead of 29.5 and it would work fine. The purpose of using 29.5 is just to keep the secondary surface (the one the tool slides along) clean with a slight 1/2 degree cleanup on the right side of the tool tip while the left side does all the work.
I find that most people cutting threads for the first time get confused by all this 29.5 degree complexity. But it sounds like you noodled it out. But just in case, just think about cutting a 60 degree thread by coming in at a 30 degree angle (half of 60) and using the left side of the tool to cut progressively further into the thread. Then twist the compound another half a degree to keep cleaning up the right side a teeny tiny wee bit too. It becomes second nature in no time.
Lastly, don't worry about how messy those test cuts look on the surface. Just keep them shallow. Use dykem on previous cuts just to hide them until you are happy. You don't even need to change the compound one iota while testing. The new threads will show up in the dykem.
I am also a bit surprised you are not screaming like a wild banshee about cutting metric threads. They are not so easy. If you are cutting metric threads as easily as imperial ones. I AM IMPRESSED!
If not, then search this forum for metric threads using a thread dial. It's way easier that stopping the lathe perfectly each time. Or we can set up a call.
I just took an hour in the shop out of my workday to give this another go. My test nut goes on the 18tpi end, and in a couple of hours I'll get to go back out and turn it around and get the metric end sorted. Lots of space to hold onto the shaft without damaging the threads. Used the cross-slide only to engage/disengage, which meant I could keep bringing it back to my dialed-in zero, and the compound to from the thread, letting me inch forward by a couple of thou at a time, reliably. I'm not convinced my thead forms are perfect (they seem a little asymmetric), and the gutter at the end of the threaded section is a bit rough, but I'll take it!
Thanks for all the feedback!
Thanks for all the feedback!
If I understand the thread dial correctly, doing metric with it should come down to just always starting on the "1" rather than trusting any of the other partitions of the screw, right? Mine is numbered 1-4, so I assume each numbered position corresponds to engaging the 0th, 2nd, 4th, and 6th "position" of the 8tpi screw within each inch modulus. Inch modulus is meaningless to metric, so just be consistent and start on the same thread modulus each time, from within 1" of travel of the point where the thread starts.
Or will that lead me grossly astray? I'm going to try it on a piece of scrap and see where I wind up.
The hard part with this end is that I don't have a test nut - the nut is under my tool post, and I'm not liking the idea of breaking down my setup to test the threads. I guess I'll have to do the actual math on it.
I had tail support in that I pulled out so I could get the nut over - you can see the center in the end.
And now you get my confession - yes, I had a bad moment part way through the threading. I managed to dial my cross-slide in one turn too far during one cut, and stalled the lathe and budged my toolpost. I think my tail support gave out, I may have not locked the tailstock correctly. I didn't notice that the work had turned in the chuck, about 1/6th of a turn:
There was a bunch of tooth-gnashing and careful re-alignment of the tool in the thread on the right mark of the threading dial. If I'd noticed it had rotated in the chuck I'm not sure I'd have tried to fix it!
It might very well.
I presume your lathe is Imperial. For Metric threads, engage the half nut lever and leave it engaged throughout the whole metric threading operation. Use spindle reverse to move the carriage back to the starting point. You can not use the threading dial when threading in the “off units” system to which the lathe lead screw was made.
It happens. Perhaps a little above what you can do at this time, but there are ways to “chase” an existing thread and salvage the part.
Look at this video from Joe Pie (there are others as well). He explains how to chase a thread.
@plalonde - please allow me to explain that again.
I deleted my previous post after redoing it here.
About metric threading. No, it isn't as simple as choosing only the number one. It actually doesn't matter what number you choose as long as you never change it. However, for metric threads it's a bit more complicated.
Most references will tell you that you should never disengage the half nut at all when threading in metric. That way it's always in the same place so thread indexing is never lost. It's not about always using the same number, it's about never disengaging it! It must stay exactly where it was for all cutting passes. But I find this too difficult to do. It also requires a very large thread relief.
I accomplish the same thing by disengaging it when I reach the relief, letting the thread dial move however much it moves (but never more than one turn) and then putting it back where it belongs when that setting comes back when I reverse.
The process goes like this:
Take your first pass with the half nut engaged at the location of your choosing. For discussion purposes, let's say we choose position 2.
When the relief is reached, disengage the half nut. The cutting tool will stop moving, but the threading dial will start moving until you shut off the lathe. Shut off the lathe .
When the lathe has stopped, turn the cross slide out one full turn on the handle. Then put the lathe into reverse.
Watch the threading dial. As soon as it comes around to the number 2 (or whatever you chose), engage it again. If you miss, don't wait wait for it to go around again. That will not work. Instead, put the lathe in forward until it passes the mark, and then try again. The goal is to put it back exactly where it was. Not one turn more or less. Exactly where it was. This is the same as having left it engaged but without the stress of missing the thread relief.
After it is back where it was on the threading dial, you leave the lathe in reverse until the carriage is back before the start of the thread ready for another pass.
At that point, you stop the lathe, crank the cross slide in one turn to zero, advance the compound a reasonable amount (say 10 or 15 thou), and put the lathe in forward to cut another pass.
About your slip. For most cross slides and most thread sizes, one full turn is enough to clear the threads. I like to set my zero with the crank handle facing down. That way I can crank it out one turn to down again, dial a bit on the compound, and then crank it in one turn to zero. One turn is hard to overdo or underdo. But you can also tell just looking at the tool as it approaches the part that you are off and stop the lathe in time.
About the fit. If you have thread wires, there are charts that list the optimum diameters for all thread sizes. You don't really need a test nut. Just use the wires to measure the threads.
If you don't have wires, make some. But instead of using the charts, use your old bolt and duplicate that measurement on the new bolt with the same wires.
Then go get some thread wires or a threading Micrometer when you have time.
I deleted my previous post after redoing it here.
About metric threading. No, it isn't as simple as choosing only the number one. It actually doesn't matter what number you choose as long as you never change it. However, for metric threads it's a bit more complicated.
Most references will tell you that you should never disengage the half nut at all when threading in metric. That way it's always in the same place so thread indexing is never lost. It's not about always using the same number, it's about never disengaging it! It must stay exactly where it was for all cutting passes. But I find this too difficult to do. It also requires a very large thread relief.
I accomplish the same thing by disengaging it when I reach the relief, letting the thread dial move however much it moves (but never more than one turn) and then putting it back where it belongs when that setting comes back when I reverse.
The process goes like this:
Take your first pass with the half nut engaged at the location of your choosing. For discussion purposes, let's say we choose position 2.
When the relief is reached, disengage the half nut. The cutting tool will stop moving, but the threading dial will start moving until you shut off the lathe. Shut off the lathe .
When the lathe has stopped, turn the cross slide out one full turn on the handle. Then put the lathe into reverse.
Watch the threading dial. As soon as it comes around to the number 2 (or whatever you chose), engage it again. If you miss, don't wait wait for it to go around again. That will not work. Instead, put the lathe in forward until it passes the mark, and then try again. The goal is to put it back exactly where it was. Not one turn more or less. Exactly where it was. This is the same as having left it engaged but without the stress of missing the thread relief.
After it is back where it was on the threading dial, you leave the lathe in reverse until the carriage is back before the start of the thread ready for another pass.
At that point, you stop the lathe, crank the cross slide in one turn to zero, advance the compound a reasonable amount (say 10 or 15 thou), and put the lathe in forward to cut another pass.
About your slip. For most cross slides and most thread sizes, one full turn is enough to clear the threads. I like to set my zero with the crank handle facing down. That way I can crank it out one turn to down again, dial a bit on the compound, and then crank it in one turn to zero. One turn is hard to overdo or underdo. But you can also tell just looking at the tool as it approaches the part that you are off and stop the lathe in time.
About the fit. If you have thread wires, there are charts that list the optimum diameters for all thread sizes. You don't really need a test nut. Just use the wires to measure the threads.
If you don't have wires, make some. But instead of using the charts, use your old bolt and duplicate that measurement on the new bolt with the same wires.
Then go get some thread wires or a threading Micrometer when you have time.
Last edited:
This link is useful for thread sizes & provides input/output for wire measurement sizes. There is a metric version in there somewhere as well.
Disclosure: I am from the thread mic owners fraternity. As opposed to, once again I dropped a fr*cken wire in the chip pan fraternity
theoreticalmachinist.com
Disclosure: I am from the thread mic owners fraternity. As opposed to, once again I dropped a fr*cken wire in the chip pan fraternity