Threading away from the chuck on my lathe...

[schemer]

The easier and safer but slower way to cut a thread from a shoulder is to cut to the right, leave the half nuts engaged for the entire job, and back the tool up under power to near the shoulder with the tool away from the work, and then move the tool up against the shoulder by rotating the chuck by hand. Dial in the feed for the cut and run the cutter toward the tailstock. Repeat as required.

Bob, I checked out your thread in that link and those are some tiny threads but they turned out great. I have read about leaving the half nuts engaged for metric threads but this would be another good use as you have pointed out. On metric threads though, why can't the numbers be used anyhow?? And here are the pics of my rough looking threads as I mention a post or two above:

 

[RJSakowski]

Imperial lead screws are nice even tpi threads; 8, 12, 16. My grizzly 602 has 12 tpi and 48 teeth on the threading indicator gear with 12 marks on the indicator. Rotation of 1/4 turn means the position of the carriage relative to the lead screw has moved by exactly 1 inch. The thread indicator has move exactly 3 marks. All imperial threads have an integral number of threads in an inch so the pattern repeats every inch and moving 1 inch puts the cutting tool in the same relationship to the thread being cut as it was before moving. For odd half integral tpi, the pattern repeats every 2 inches so I have to move the carriage by 6 marks or a half dial rotation. For thread pitches divisible by 3 (6, 9,12, etc.) the pattern repeats every 1/3 inch and I can engage on any mark.

My Atlas has a 16 tpi lead screw and 32 teeth on the the thread indicator gear. One complete revolution of the thread dial moves the carriage exactly 2". There are four marks on the the dial so moving 2 marks moves the carriage exactly 1 inch. To cut even tpi, I can engage on any mark since the pattern repeats every half inch. To cut odd tpi, I have to engage on every other mark, either even or odd. To cut an odd half integral thread, I have to engage on the same mark for each pass. 8 tpi lead screw machines function in much the same manner as the Atlas.

Unfortunately, metric threads are not integral multiples of 1 inch. For a 1mm pitch, the pattern doesn't repeat for 100m or 3,937 inches. for .7mm, you would have to move 700m or 27,559". You have to engage in exactly the same position on the lead screw to ensure the cutting tool maintains its relationship to the thread. This is why we are advised to not disengage the half nuts.

Tom Lipton, OxTools has a good video on a method of cutting metric threads with a lathe with an Imperial lead screw where he does disengage the half nuts but then re-engages them in the same position. The advantage is that you don't have to rely on the lathe coasting to a stop when threading to a shoulder. There have been several references to this video on this forum and a quick search will pull it up.

 

[mksj]

There is a lot of technique to cutting a decent thread, and type, pitch and depth are all factors. The material/diameter/cutter used also plays a significant part of the equation. So there is a bit of trial and error to figure out what works. I have my lathe setup with an electronic stop, so the half-nut is not disengaged until you have completed the threading process, which makes metric threading much easier. An electronic stop also allows you to thread up to a shoulder, as the repeatability exceeds 0.001". Without this, there are some merits to threading away from the headstock as shown, but it takes some experience to hit the threading dial at the correct position when cutting something like a 8 TPI at 60 RPM. I usually thread steel at 120-150 RPM, and softer materials at 250 RPM. So each pass occurs in 1-2 seconds.

The thread cutter/insert also plays a big part in cutting the thread and the final finish. Specific to threading inserts/cutters there is a specific TPI cutting range that is specified. I recently had to cut a M50-1.5 thread in 8620 steel, and had a few bobbles in getting it cut right. I had not turned this steel before nor threaded it, and found that it did not take well to positive rake tooling. Also you need to be very methodical in the depth of cut on each pass, and look at the curls/finish on each pass. In this case, I got a little too aggressive on the depth of cut and the positive rake threading insert blew apart, the thread also got a bit mangled. I switched to a neutral rake threading insert and took smaller cuts and was able to come out with a decent final product. On smaller threads, I will often cut about 95% of the thread on the lathe and finish the final sizing with an adjustable die or tap, this gives a very precise and polished finish cut.

In your case, you need a sharp cutting tool, good cutting fluid, and to take progressively lighter cuts per pass the deeper the thread. Another common mistake, is that some threading dials may not be perfectly aligned with the indicator line. So one may think they are engaging it at the same mark, but in reality they are either just ahead or behind the indicator and do not perceive the difference. This can easily happen with the lower TPI threads because of the dial speed.

M50-1.5 thread in 8620


In this case both internal and external 1" threads were cut, but required different threading inserts because the thread depth was >0.07". These where cut at around 250 RPM.

 

[Doubleeboy]

heard about turning upside down and left to right years ago but never tried it till I saw Joe P's Youtube video. I am a believer, not only is drama free it is also a very easy way to pick up a thread to do repair work.

 

[schemer]

Imperial lead screws are nice even tpi threads; 8, 12, 16. My grizzly 602 has 12 tpi and 48 teeth on the threading indicator gear with 12 marks on the indicator. Rotation of 1/4 turn means the position of the carriage relative to the lead screw has moved by exactly 1 inch. The thread indicator has move exactly 3 marks. All imperial threads have an integral number of threads in an inch so the pattern repeats every inch and moving 1 inch puts the cutting tool in the same relationship to the thread being cut as it was before moving. For odd half integral tpi, the pattern repeats every 2 inches so I have to move the carriage by 6 marks or a half dial rotation. For thread pitches divisible by 3 (6, 9,12, etc.) the pattern repeats every 1/3 inch and I can engage on any mark.

My Atlas has a 16 tpi lead screw and 32 teeth on the the thread indicator gear. One complete revolution of the thread dial moves the carriage exactly 2". There are four marks on the the dial so moving 2 marks moves the carriage exactly 1 inch. To cut even tpi, I can engage on any mark since the pattern repeats every half inch. To cut odd tpi, I have to engage on every other mark, either even or odd. To cut an odd half integral thread, I have to engage on the same mark for each pass. 8 tpi lead screw machines function in much the same manner as the Atlas.

Unfortunately, metric threads are not integral multiples of 1 inch. For a 1mm pitch, the pattern doesn't repeat for 100m or 3,937 inches. for .7mm, you would have to move 700m or 27,559". You have to engage in exactly the same position on the lead screw to ensure the cutting tool maintains its relationship to the thread. This is why we are advised to not disengage the half nuts.

Tom Lipton, OxTools has a good video on a method of cutting metric threads with a lathe with an Imperial lead screw where he does disengage the half nuts but then re-engages them in the same position. The advantage is that you don't have to rely on the lathe coasting to a stop when threading to a shoulder. There have been several references to this video on this forum and a quick search will pull it up.

Thanks RJ for the detailed explanation on the half nuts and metric threading. I will check out Tom Liptons video. I have watch some of his stuff in the past.

 

[schemer]

There is a lot of technique to cutting a decent thread, and type, pitch and depth are all factors. The material/diameter/cutter used also plays a significant part of the equation. So there is a bit of trial and error to figure out what works. I have my lathe setup with an electronic stop, so the half-nut is not disengaged until you have completed the threading process, which makes metric threading much easier. An electronic stop also allows you to thread up to a shoulder, as the repeatability exceeds 0.001". Without this, there are some merits to threading away from the headstock as shown, but it takes some experience to hit the threading dial at the correct position when cutting something like a 8 TPI at 60 RPM. I usually thread steel at 120-150 RPM, and softer materials at 250 RPM. So each pass occurs in 1-2 seconds.

The thread cutter/insert also plays a big part in cutting the thread and the final finish. Specific to threading inserts/cutters there is a specific TPI cutting range that is specified. I recently had to cut a M50-1.5 thread in 8620 steel, and had a few bobbles in getting it cut right. I had not turned this steel before nor threaded it, and found that it did not take well to positive rake tooling. Also you need to be very methodical in the depth of cut on each pass, and look at the curls/finish on each pass. In this case, I got a little too aggressive on the depth of cut and the positive rake threading insert blew apart, the thread also got a bit mangled. I switched to a neutral rake threading insert and took smaller cuts and was able to come out with a decent final product. On smaller threads, I will often cut about 95% of the thread on the lathe and finish the final sizing with an adjustable die or tap, this gives a very precise and polished finish cut.

In your case, you need a sharp cutting tool, good cutting fluid, and to take progressively lighter cuts per pass the deeper the thread. Another common mistake, is that some threading dials may not be perfectly aligned with the indicator line. So one may think they are engaging it at the same mark, but in reality they are either just ahead or behind the indicator and do not perceive the difference. This can easily happen with the lower TPI threads because of the dial speed.

M50-1.5 thread in 8620
View attachment 243561

In this case both internal and external 1" threads were cut, but required different threading inserts because the thread depth was >0.07". These where cut at around 250 RPM.
View attachment 243563

MKSJ,
Looking good. I thought about the threading dial and maybe me missing the mark one time but although the line doesn't line perfectly up with the mark, it has to move quite a bit before it locks in the next tooth. I am pretty sure I can adjust it by loosening the gear on the dial and lining it up when the half nuts are engaged and re-tightening it, or I could just make a new temporary mark. That electronic stop is nice but I do not have one. I hope it is reliable or...ouch. I am threading unknown DOM tubing drops from ebay for my part. I believe it can likely only be one of a handful of steel types like A513 in 1020-1026. So it should be able to be threaded ok. I was using that dark sulphur cutting fluid and have others to try too. I don't think I am going to buy a 2 1/4-8 die for my most likely one of project as in reality, the threads don't have to be perfect (although I want them to be). It is just to adapt a chuck and face plate to a welding positioner. But I believe I messed up my threading tool with too deep a cut although I don't think I moved the compound more than .005 but maybe that is too much once you are in the groove deeper than the start. The best teacher is experience but not with large diameter steel chunks for the scrap barrel. I am going to test on some 1214L but it is only 1 inch diameter. I am sure it will cut like butter and I will smile, but I want that same smile on the DOM @ 2 1/4-8. Thanks for all the feedback.

 

[RJSakowski]

The OEM thread dial on my G0602 was not concentric with the shaft. It was far enough out that it would be uncertain as to which tooth of the gear to engage. I made a new dial for my lathe and everything line up perfac5tly now. The sordid details are here:http://hobby-machinist.com/threads/threading-on-a-grizzly-g0602.34230/

 

[mksj]

The electronic stop in reality does not fail, it uses an electronic proximity sensor similar to what is used on CNC machines for homing and stops. If they fail the machine would brake to a stop, it is also backed up by a mechanical limit switch which engages within 0.05" after the P sensor triggers. The hard stop is at 0.07".

The depth of cut you are taking is quite large, relative to my experience. I take a scribe pass, and at most will take a 0.002" cut, then 0.001", then back down to 0.0005" as I get closer to target and then a spring pass in particular on a smaller less rigid lathe. Depends on the material, thread and diameter. With coolant and a bigger machine you can get away with deeper cuts. Some steels like 1018, at least for me seems to tear and chatter when cutting threads. Smaller thread cutters/inserts are easily damaged from too deep of cut, especially as you get deeper into the thread. If you need to do deeper threads on larger stock (over 1" diameter), you might consider a beefier threading tool, you will still run into problems with deep threads. The maximum depth for my threading inserts is ~0.07" and you are looking at twice this. Example below is the Carmex lay down inserts which I use for internal threads, and I am pushing them at 0.07" thread depth. I think most people would grind their own HSS thread cutter in this application.

 

[woodchucker]

Yeah, I have always used the compound method but it seems I should try the cross slide as it sure does seem like it would be easier. I have a Grizzly G4003G so I think is will be stout enough to do what I need even if I have to take smaller cuts to do it. I just wanted to try higher speeds for better finish and use a carbide instead of HSS.

I prefer the compound method. And I have cut a couple of times with the tool upside down and running in reverse out toward the tailstock. Met a guy a few years ago who told me about it, that it was the way to go when you needed to get close and would crash if you didn't stop in time.. I told him I wouldn't be needing that, I don't think I'll have that problem... (NEVER SAY NEVER)... And when I had to do it, I clearly remembered his sage advice. He obviously walked away shaking his head saying stupid guy you'll need it one day.

 

[schemer]

The OEM thread dial on my G0602 was not concentric with the shaft. It was far enough out that it would be uncertain as to which tooth of the gear to engage. I made a new dial for my lathe and everything line up perfac5tly now. The sordid details are here:http://hobby-machinist.com/threads/threading-on-a-grizzly-g0602.34230/

You did a nice job on that. I think my problem is user error. I think my lathe is ok except the threading dial out of line by a smidge.