- Joined
- Feb 20, 2019
- Messages
- 335
There is often times a clutch on the feed mechanism, not going to be a very good thread when your clutch slips.
-
Welcome back Guest! Did you know you can mentor other members here at H-M? If not, please check out our Relaunch of Hobby Machinist Mentoring Program!
Power Feed to Make Threads (English or metric threads) Curiosity
- Thread starter B2
- Start date
-
- Tags
- power feed thread chart tpi
Thanks. True on my PM1440GT also. Interesting thought! I had not though of this. I guess one has to be aware of it for sure. The way the manual describes it ("The clutch comprises a pair of spring loaded steel balls bearing on a detent disc driven by the saddle feed gearbox."). It does not sound like a high performance device and that a variable load might make the phase of the feed rod motion modulate without actually slipping out of the detent. If it is like the detent mechanism in my speed control lever it is not all that great. So I will have to engage mine to see what it sounds like when activated. I suspect it makes a clicking noise. I had thought that there was one also on the lead screw, but I guess not. It would just crash a tool into the spindle if it were not for my proximity sensor system. I suppose with the proximity sensor in place one could get rid of the Feed shaft clutch or redesign it. One might ask, "Does one need a safety clutch on the slow feed mechanism why is there not one on the lead-screw as well?" Maybe just for this reason!
Dave L.
Dave L.
- Joined
- Feb 1, 2015
- Messages
- 9,631
- Joined
- Feb 20, 2019
- Messages
- 335
What happens when a piece of swarf triggers your proximity sensor and your in the middle of say a $795 carbon fiber barrel blank thread? Not enough material for a do-over. You just created a very expensive tomato stake. Might not ever happen until Mr. Murphy pay a visit.
I thread away from the chuck and don't have to worry about crashing. If you have a chuck that is not screwed on, I don't see why anyone would ever feel they need the stress.
Joe says he's been doing it this way for 30+ years, so if good enough for him, I know it works for me.
Try cutting an internal Acme 4 thread into a blind hole at 60 rpm.
Joe Piecynski single point threading playlist
Last edited:
- Joined
- Nov 23, 2020
- Messages
- 788
Gears that are not involute move at different rates of speed depending on tooth engagement. The saddle would speed up and slow down very slightly throughout the cut, but the spindle would be uniform. It would cause some thread deformity. Would it be inaccurate enough to ruin a thread? Probably not. Could be for fine pitched threads though. I don’t think the clutch slipping would be an issue though. I would think single point threading would be less taxing than rough cuts with power feed.
Thanks guys,
@xr650rRider
Yes, I like the inverted threading concept and know that you are a believer! However, because the chuck spins on and off to mount/dismount on my old lathe spindle I was afraid of it causing the chuck to come loose so have not moved over. However, more modern lathes do not have spin on chucks so maybe I can give it a try one of these days on my PM1440GT. I may have to get some extra tooling.
.
Yes, I have been concerned about the magnetic metal cutting debris getting into the inductive (magnetic) proximity sensors that are used by essentially everyone. For this reason I have been thinking about building a new sensor which is more like a limit switch but more precise. Most of the VFD conversions discussed on HM are using mechanical relays rather than solid state electronics. For this reason, the relay contact bounce at the VFD latching relay input can result in slow relay response or even chatter of the relay, so a mechanical switch's performance might not be great for repeatability, reliability of the relay, nor over all performance. However, my solid state VFD conversion and its transistor latch system does not have this slowness limitation. The solid state latch is so fast that it would unlatch during the first part of any contact bounce. So while any debris might still be a concern about repeatability, at least it would not prevent it from working all together.
@Christianstark
I just went an looked at my PM1440GT. I cannot say that the lathe bed track uses involute thread shape, but they are some what like them. I would say they look more like triangles with the tips cut off. (There is probably a name for this style thread.) Meanwhile the Lead screw thread is square. In is my understanding that an involute tooth shape is used to prevent wear of gears. A round gear against a round gear will have no/little friction, during the gear turning, at the contact point when an involute gear is used as the surfaces are essentially rolling over each other. No friction means no wear. However, in order not to have back lash they have to be very tightly fitted, which we never see so wear does develop. So the straight feed lathe bed gear teeth are on an essentially infinite radius circle. For this am not for sure that an involute gear tooth shape would work well and raises the question which gear tooth size should even be used? Since my lathe is pretty new, I feel no backlash on my apron when the Power Feed lever is engaged but I have not measured it. So maybe the motion uniformity is pretty good??! On the other hand perhaps any backlash would not even be due to this gearing interface. Their are more gears inside the apron! I suppose gears in the lathe gear box also have lots of backlash so maybe the motion of the lead screw is not perfect either when under load?! So maybe it goes like this.... when threading course threads where the load is usually high the errors in the gears do not mater much as a percentage of the thread length. When threading fine threads the load is always small so maybe the back lash does not really kick in much. Win-Win?
Dave L.
@xr650rRider
Yes, I like the inverted threading concept and know that you are a believer! However, because the chuck spins on and off to mount/dismount on my old lathe spindle I was afraid of it causing the chuck to come loose so have not moved over. However, more modern lathes do not have spin on chucks so maybe I can give it a try one of these days on my PM1440GT. I may have to get some extra tooling.
.
Yes, I have been concerned about the magnetic metal cutting debris getting into the inductive (magnetic) proximity sensors that are used by essentially everyone. For this reason I have been thinking about building a new sensor which is more like a limit switch but more precise. Most of the VFD conversions discussed on HM are using mechanical relays rather than solid state electronics. For this reason, the relay contact bounce at the VFD latching relay input can result in slow relay response or even chatter of the relay, so a mechanical switch's performance might not be great for repeatability, reliability of the relay, nor over all performance. However, my solid state VFD conversion and its transistor latch system does not have this slowness limitation. The solid state latch is so fast that it would unlatch during the first part of any contact bounce. So while any debris might still be a concern about repeatability, at least it would not prevent it from working all together.
@Christianstark
I just went an looked at my PM1440GT. I cannot say that the lathe bed track uses involute thread shape, but they are some what like them. I would say they look more like triangles with the tips cut off. (There is probably a name for this style thread.) Meanwhile the Lead screw thread is square. In is my understanding that an involute tooth shape is used to prevent wear of gears. A round gear against a round gear will have no/little friction, during the gear turning, at the contact point when an involute gear is used as the surfaces are essentially rolling over each other. No friction means no wear. However, in order not to have back lash they have to be very tightly fitted, which we never see so wear does develop. So the straight feed lathe bed gear teeth are on an essentially infinite radius circle. For this am not for sure that an involute gear tooth shape would work well and raises the question which gear tooth size should even be used? Since my lathe is pretty new, I feel no backlash on my apron when the Power Feed lever is engaged but I have not measured it. So maybe the motion uniformity is pretty good??! On the other hand perhaps any backlash would not even be due to this gearing interface. Their are more gears inside the apron! I suppose gears in the lathe gear box also have lots of backlash so maybe the motion of the lead screw is not perfect either when under load?! So maybe it goes like this.... when threading course threads where the load is usually high the errors in the gears do not mater much as a percentage of the thread length. When threading fine threads the load is always small so maybe the back lash does not really kick in much. Win-Win?
Dave L.
- Joined
- Jun 12, 2014
- Messages
- 4,821
There is a reason why you never see a feed rod used for threading, it is not accurate, has too much play and the loading would probably change the position. My lathe in feed mode has about a 1/4 turn wind-up in the chuck before it engages the carriage going from reverse/forward, and cannot be used with a threading dial if one was to use it. Clutch only engages with a hard stop, it doesn't do little slips unless one is doing really deep cuts, I have yet to get to that kind of load. Fundamental difference is that feed rates are an much slower then thread rates, the maximum feed rate on my lathe is 0.033 IPR, which would be a 30.333 pitch. If you want to go faster, you may be exceeding the design limits of the feed gear train. If one wants a zillion threads, then go to an ELS, or buy a lathe with more capabilities.
Use of a proximity stop has no effect on screwing up threads, it is only the user that screws up threads. Upside down reverse threading you can't see the chips/what the cutter is doing, and metric threads would be even more difficult. Make a chip shield if you are worried about chips triggering the sensor. I am not looking for a Rube Goldberg machine to do things I will never need.
Use of a proximity stop has no effect on screwing up threads, it is only the user that screws up threads. Upside down reverse threading you can't see the chips/what the cutter is doing, and metric threads would be even more difficult. Make a chip shield if you are worried about chips triggering the sensor. I am not looking for a Rube Goldberg machine to do things I will never need.
- Joined
- Nov 23, 2020
- Messages
- 788
- Joined
- Feb 13, 2017
- Messages
- 2,138
There are several methods for threading vs carriage travel. The two I have are vastly different. Either one works well enough for small threads. The smallest(?) thread I have cut on both machines is 80 TPI. They will both cut smaller, the problem is my skill grinding tools.
The Atlas 12 inch (Craftsman version 101.27440) cuts both feed and threading from one point, the half nuts. There is a follower gear/dial that is geared to the lead screw but I have taken it off and use motor reversing for the subsequent passes. Metric gears on a U S lead screw must do so. I do both for convenience to me and just ignore the dial. Plus, the machine is old and sorta "loose".
The earlier model Grizzly (G-1550) is a 9X19 Asian model. Essentially a G-4000 but made in Taiwan rather than China. I have personal perceptions of the Grizzly vs a Horrible Fright equivilent, having had both. But they are not the point here. The Griz has a "feather" that runs in a slot in the lead screw and is fed from a different lever than the longitudinal feed. I changed one of the gears in the outboard chain to get finer threads. 80 TPI is the smallest I dare cut. It will go to 160, I think.
There are larger machines that have two lead screws. One for feed, one for threading. I would love to have such a machine, but am limited by weight in a wood frame shop (old house) in addition to cost. The Atlas is heavy enough to rate reincorcement under the floor, I don't dare put a heavier machine in place. From experience, I consider the Atlas to be a "light" but quite servicable machine.
The Atlas is the key to the original question. It cuts threads using the same process as when doing longitutinal feed. On that machine, the only difference between the two is that the threads cut on the lower end have a sharper, well pointed tool and lighter cuts. The idea of cutting threads with the "feed" function would be limited only by the pitch of the thread. TPI is a function of how far the carriage moves per revolution of the spindle. Feed is no different from threading beyond the depth at which it is cut. At 40 TPI, the carriage is moving 0.025" per revolution. The depth of cut is on the order of a couple or three thou per pass for threading. And, in my case, the speed at which the spindle is running. I'm an old man and don't think/move that fast, so need to run the machine slower. Beyond that, as long as the feed gearing will align with the proper pitch, there is no reason not to cut threads with longitudinal feed.
.
The Atlas 12 inch (Craftsman version 101.27440) cuts both feed and threading from one point, the half nuts. There is a follower gear/dial that is geared to the lead screw but I have taken it off and use motor reversing for the subsequent passes. Metric gears on a U S lead screw must do so. I do both for convenience to me and just ignore the dial. Plus, the machine is old and sorta "loose".
The earlier model Grizzly (G-1550) is a 9X19 Asian model. Essentially a G-4000 but made in Taiwan rather than China. I have personal perceptions of the Grizzly vs a Horrible Fright equivilent, having had both. But they are not the point here. The Griz has a "feather" that runs in a slot in the lead screw and is fed from a different lever than the longitudinal feed. I changed one of the gears in the outboard chain to get finer threads. 80 TPI is the smallest I dare cut. It will go to 160, I think.
There are larger machines that have two lead screws. One for feed, one for threading. I would love to have such a machine, but am limited by weight in a wood frame shop (old house) in addition to cost. The Atlas is heavy enough to rate reincorcement under the floor, I don't dare put a heavier machine in place. From experience, I consider the Atlas to be a "light" but quite servicable machine.
The Atlas is the key to the original question. It cuts threads using the same process as when doing longitutinal feed. On that machine, the only difference between the two is that the threads cut on the lower end have a sharper, well pointed tool and lighter cuts. The idea of cutting threads with the "feed" function would be limited only by the pitch of the thread. TPI is a function of how far the carriage moves per revolution of the spindle. Feed is no different from threading beyond the depth at which it is cut. At 40 TPI, the carriage is moving 0.025" per revolution. The depth of cut is on the order of a couple or three thou per pass for threading. And, in my case, the speed at which the spindle is running. I'm an old man and don't think/move that fast, so need to run the machine slower. Beyond that, as long as the feed gearing will align with the proper pitch, there is no reason not to cut threads with longitudinal feed.
.
Hi Bi11
I too have had need for and cut some high density treads in my optics work while at the Eastman Kodak Research Labs, as well as since then. It turns out in some optics applications you wish to turn or move a lens to adjust a very tight focus, especially in applications like microscopy. I first ran into this when I had responsibility to find materials to use for optical disk recording back in the early 1070's (think Blue Ray DVD type devices). Here one employs the focusing of a laser spot down to the "diffraction limit" to make a dot or stripe exposure on the medium/disk surface. The depth of field over which the beam can be focused to that limit is approximately the wavelength of the light/laser being used. For green or blue light that is approximately 1/2 of a micron. So you wish to be able to move a lens on the 1/2 micron scale or even better. In my case, I had to do this in the total dark when doing the recording as some of Kodak's potential materials were very light sensitive. There are optical tricks, primarily looking at the reflected portion of the beam, to determine if you have reached this focal point or not. Anyway, I had access to a really competent technician who was also an excellent machinist, and he built me something that at that early stage of my life I would not have been able to do. (I am sure that even now I could not make as fine a device even with more than one try!) I also had the help of a very good mechanical engineer, who the technician worked regularly for. He designed a spring loaded air pressure floating foot that would stay at, and repeatedly go back to a constant distance from the surface it was setting on. (Think of a game table where you float a hockey puck like object around on the compressed air coming out of the holes in the table.) This worked well enough that I could find a focus position while in room light, then in the dark I could pull back on air-spring device and insert fresh materials to be exposed. The air foot on the device was less than 2" by 2" with air ports at the corners and the then lens mount in the hollow center. The material was on a cheap, motorized, X transnational stage table which conveyed the material to be expose... so I could make a string of exposed spots by moving the stage while modulating the laser on and off at very short, micro-second variable, exposure times. The lens barrel was to be mounted into the air spring foot and was turned in a threaded mount to adjust the focus. The Tech. took the microscope lens that I was to use and machined off most of the barrel of it and re-threaded it with a very fine thread and, of course a new outer barrel to receive these threads, which was held in the air-spring mount. He also machined a hardened steel wheel handle for the lens which was about 1.5-2" in diameter and he put 360 cut lines on the outside of the wheel. Into these he fashioned a spring loaded detente point. So as you turned it one turn it made 360 clicks, which you could feel and count. As I recall, each click represented a motion of the lens of less than or about 1/2 micron.
It turned out later that a couple of the best materials I found for this application was very thin films of Bismuth and amorphous Tellurium and Selenium which were not really light sensitive at all. They formed images by melting the material and making tiny holes in the thin films. Hence, to this day, Bismuth is one of my favorite materials. Everyone should have a favorite material! =-) !
So after all of these years, if I still have that right, a rotation of that wheel by one full turn would have moved the lens in less than 180 microns. At 40 microns per 0.001" that meant that it moved 0.0045" per turn ... which is 222TPI! Maybe my memory of the numbers is off by a factor of 2, but even if so, this is some pretty fined threads in that brass barrel! More a kin to what Bi11 was talking about.
I got to work on a lot of neat projects while at EK! So I was able to learn a lot of stuff.
By the way, the Norton gear box of my Old South Bend 10K heavy that is in need of a rework has 10 fundamental threads and 7 factors of 2x. Hence, the highest thread density listed on the plate is 480TPI or 0.0021" per turn with a corresponding feed rate of 0.0007"! That is without any exchange gears! I am sure that position has seen very little use! None by me. Love that Norton Gear box and wish I had one on my PM1440GT!
Thanks to everyone for their responses. I think I got what I needed to know. If others have had experience with Feed threading I would still be interested in hearing about it.
Dave L.
Thank you for your response! It is very through, useful and very informative. Base upon the input from the forum, an especially your experience, I have decided I will include the possible threads one can get from the FEED position in my over all charts. It seems clear that some folks have lathes which can be used for this and their operators are sufficiently skilled to make use of the information. Others may not need it or be sufficiently skilled to use it. My spread sheet tool will also be constructed so that if one does not want the FEED thread info they can always leave the data out of the search or the overall threading table.
I too have had need for and cut some high density treads in my optics work while at the Eastman Kodak Research Labs, as well as since then. It turns out in some optics applications you wish to turn or move a lens to adjust a very tight focus, especially in applications like microscopy. I first ran into this when I had responsibility to find materials to use for optical disk recording back in the early 1070's (think Blue Ray DVD type devices). Here one employs the focusing of a laser spot down to the "diffraction limit" to make a dot or stripe exposure on the medium/disk surface. The depth of field over which the beam can be focused to that limit is approximately the wavelength of the light/laser being used. For green or blue light that is approximately 1/2 of a micron. So you wish to be able to move a lens on the 1/2 micron scale or even better. In my case, I had to do this in the total dark when doing the recording as some of Kodak's potential materials were very light sensitive. There are optical tricks, primarily looking at the reflected portion of the beam, to determine if you have reached this focal point or not. Anyway, I had access to a really competent technician who was also an excellent machinist, and he built me something that at that early stage of my life I would not have been able to do. (I am sure that even now I could not make as fine a device even with more than one try!) I also had the help of a very good mechanical engineer, who the technician worked regularly for. He designed a spring loaded air pressure floating foot that would stay at, and repeatedly go back to a constant distance from the surface it was setting on. (Think of a game table where you float a hockey puck like object around on the compressed air coming out of the holes in the table.) This worked well enough that I could find a focus position while in room light, then in the dark I could pull back on air-spring device and insert fresh materials to be exposed. The air foot on the device was less than 2" by 2" with air ports at the corners and the then lens mount in the hollow center. The material was on a cheap, motorized, X transnational stage table which conveyed the material to be expose... so I could make a string of exposed spots by moving the stage while modulating the laser on and off at very short, micro-second variable, exposure times. The lens barrel was to be mounted into the air spring foot and was turned in a threaded mount to adjust the focus. The Tech. took the microscope lens that I was to use and machined off most of the barrel of it and re-threaded it with a very fine thread and, of course a new outer barrel to receive these threads, which was held in the air-spring mount. He also machined a hardened steel wheel handle for the lens which was about 1.5-2" in diameter and he put 360 cut lines on the outside of the wheel. Into these he fashioned a spring loaded detente point. So as you turned it one turn it made 360 clicks, which you could feel and count. As I recall, each click represented a motion of the lens of less than or about 1/2 micron.
It turned out later that a couple of the best materials I found for this application was very thin films of Bismuth and amorphous Tellurium and Selenium which were not really light sensitive at all. They formed images by melting the material and making tiny holes in the thin films. Hence, to this day, Bismuth is one of my favorite materials. Everyone should have a favorite material! =-) !
So after all of these years, if I still have that right, a rotation of that wheel by one full turn would have moved the lens in less than 180 microns. At 40 microns per 0.001" that meant that it moved 0.0045" per turn ... which is 222TPI! Maybe my memory of the numbers is off by a factor of 2, but even if so, this is some pretty fined threads in that brass barrel! More a kin to what Bi11 was talking about.
I got to work on a lot of neat projects while at EK! So I was able to learn a lot of stuff.
By the way, the Norton gear box of my Old South Bend 10K heavy that is in need of a rework has 10 fundamental threads and 7 factors of 2x. Hence, the highest thread density listed on the plate is 480TPI or 0.0021" per turn with a corresponding feed rate of 0.0007"! That is without any exchange gears! I am sure that position has seen very little use! None by me. Love that Norton Gear box and wish I had one on my PM1440GT!
Thanks to everyone for their responses. I think I got what I needed to know. If others have had experience with Feed threading I would still be interested in hearing about it.
Dave L.