Harrison M300 lathe - part 3 of thread pitch problem

Final part of my 3-parter on problem with long nuts jamming and possible thread pitch mismatch. No conclusive answer.
Part 1: • Harrison M300 lathe - ...
Part 2: • Harrison M300 lathe - ...

[@samuelfielder]

Twice in this video I say that I measured the actual pitch of my lathe and found it to be 1.74 mm instead of the nominal 1.25 mm. I mispoke on both occasions. I should have said 1.274 mm. Also, I was shooting for a pitch diameter of 9.188 which I took off an internet table. But Machinery's Handbook says pitch diameter should be 9.042 to 9.160 mm, so that fits better with my experience.

[@derekpeters8946]

I just wanted to say I have enjoyed the series on the Harrison lathe. I really do like the fact you have included all the bits good and bad. I look forward to your next project. Thanks Derek

[@wizrom3046]

I think you have proved very nicely why bolt and threaded rod manufacturers make their thread just slightly undersized compared to the "official" thread dimensions. 👍

[@stevechambers9166]

It’s not easy to get a perfect thread on a long thin diameter single pointing everything is against you you did a good job showing us just how tricky it is thanks for your time 👍👍👍

[@nheng6913]

There are endless videos you can make, in your style, which is thorough and entertaining. Some examples: 1. knurling 2. machining thin stock, glued against faceplate 3. boring 4. means of holding small stuff, awkward stuff, thin walled tubing 5. holding small diameters in chuck without chatter (hint: wrap soft aluminum or business card around it) 6. milling (end milling, keyway cutting, etc.) 7. polishing (just polished a bunch of 1-1/4 (20mm) copper tube for new boiler install, ten foot lengths passed thru headstock and were polished with ScotchBrite, looks like a micro brewery 8. swaging (smooth crimp on end of tubing, mount a ball bearing as the tool for the process) 9. spinning (first make forming tool out of turned wood, use long wooden stick, secured near end, as spinning tool) spin aluminum, copper, brass 10. threading with tap (centered with tailstock, then turned by hand die holder or very low RPM) 11. turning plastic without melting or deforming ,,, and much more 12. Fix things ... have made parts for car, lawn mower, tractor, brush cutter. Make obsolete parts to keep things running.

[@samuelfielder]

Some good topics. Not sure what you're saying here. I am not a machinist and don't feel qualified to make tutorials. I just document what I'm doing, which is usually blundering about and learning by experience.

[@HaxbyShed]

Thanks Samuel. I don't do much single point threading but as you say it's more of an art than a science. I always have to cut a bit deeper than the theory suggests. Threading on the lathe for me would most likely be a vice nut or perhaps my mill table nut, something you either can't buy anymore or would cost a fortune if you could. I've never had any problems with the pitch on the lathe not being right (if I set it up correctly). Cheers

[@samuelfielder]

Yes, I was only limbering up for cutting a new lead screw for the top slide, but worried now it won't be any good!

[@smallcnclathes]

Did you notice that all those ring gauges were quite short? Their length did not appear to exceed 1.5 the thread diameter. I have a M10 go gauge and it hardly makes it to 10mm thick. My advice is to cut threads to suit the job in hand. Stop trying to measure things so scientifically, get out of the laboratory and into the workshop. When I bought my M10x1.0 thread gauge I was amazed that a thread that measured 10.00 on the OD would not go anywhere near the gauge. A 9.95 thread would go in the gauge but the thread was quite firm, no looseness what so ever. So the theoretical sizes don't work without some clearance. Enjoyed the video.

[@jonsworkshop]

Hi Samuel, found your channel today and subscribed, have been watching your journey with interest. I have worked in Engineering all my career and have experienced many peculiarities like this even on high end CNC machines. Based on your evidence, I would say your problem in words is singular (variables), which in reality are many. Your M300 has unfortunately had a hard paper round, you have variable leadscrew wear, which means you also have lead nut wear (variable but over a much shorter distance) and those two don't play nicely together. With the wear in your bed, I also suspect you have some saddle crabbing close to the chuck which is another source of variation. You also changed your depth of cut on your last test (I don't blame you by the way), please see my workshop tips video on balanced cuts, and my other one on lathe push off (they go hand in hand) as you are introducing more variables through changing depths of cut and number of spring passes. I suspect your long nut is tapped?? There could also be pitch error in this, all taps are made to a tolerance etc. As you move closer to your tailstock, your variables reduce and your results improve, similarly when you cut in reverse. I love your theoretical approach, very close to where aerospace manufacturing is these days where I have spent a lot of my career, however, when dealing with 30 year old manual kit in a home shop environment, theory is rarely achieved (I realise this will not be pleasant reading😇). My one piece of advice would be "to achieve consistent outputs, control and lock down as many of your input variables as possible per my tips videos" . I hope you find this helpful. I have really enjoyed a few hours of your content today, and your M300 series is superb reference that I am sure I will be watching again in future to help me with my own. Well done! Jon

[@samuelfielder]

Thanks. I'm a subscriber of yours - watched your coolant video yesterday - and good to get advice from someone who really knows what they're talking about. I think I've watched your lathe push off video, not the balanced cuts one (but Stefan Gotteswinter, another proper professional, made one on that topic). Keep up the good work! How about a video on where to use different shaped inserts (squarish CCMT/WNMG versus pointy DCMT/VCMT ones? I got into the habit of using pointy inserts on the chinese minilathe, because I thought they would have less tool pressure on that rubbery thing, but I don't really know what I am doing on the M300 with insert selection. Annoyingly, your M300 doesn't seem to be much worn! Yes, my long nut is tapped. I agree: only change one thing at a time, but I don't always follow this advice! I generally overthink things and waste time when I should just push on and make it work somehow or other. If you have time, watch my kzbin.info/www/bejne/iYfZpXikp6d_n6s and tell me what I should do with this pile of scrap.

[@jonsworkshop]

@@samuelfielder Hi Samuel, I watched your 'what shall I do' video yesterday as part of my binge. It's a really difficult question to advise on, however here is my two pennies worth: Firstly, what do you want to make regularly on your lathe (tolerances etc) and are they generally within what you can currently achieve, if the answer is yes, I would suggest do nothing and work round what you already have. If no, then my fear is another second hand lathe is not going to do it for you as they all come with apologies in some form. If you really want repeatable accuracy over a long time then a new M300 would be your best option, the Taiwanese stuff is certainly better quality than Chinese rubbish but still not up to British/European standards. The challenge you have is working out the difference in what you want, and what you need as these will be different. I wouldn't go for a regrind on yours purely and simply because you would only be addressing one element in a system of many worn elements so whilst it will improve, it will never be perfect. If your 'want' is more important than your 'need' and you can afford it, go for a new M300 and break your existing one. Sorry no perfect answer. Cheers, Jon

[@nheng6913]

I don't recall if you have any digital readouts. You can pick up a 24" digital caliper for cheap, secure one end and the other end secures to your cross slide. Now you can read displacement over two feet with < 0.001" error.

[@dickmick5517]

I very much enjoyed your discovery. Maybe I have learned something, from you, in the process. Best of luck

[@andrewdolinskiatcarpathian]

Close up the thread looks pretty darn good. I would be overjoyed with the result you achieved. One question I would ask, if the commercially available standard nut goes on ok, how likely are you to ever thread on that longer cylindrical nut? A fascinating discourse which I have thoroughly enjoyed. Thank you for sharing. 👏👏👍😀

[@samuelfielder]

The long nut has no real use, just something to make my life awkward. But the fairly long nuts that engage with the leadscrews in the top-slide or cross-slide could be a problem if I remade their lead screws.

[@andrewdolinskiatcarpathian]

@@samuelfielder Ha ha. That long nut has indeed made your life awkward, but in turn has provided a thought provoking talking point. I like what others have said. The problem you are trying to resolve only manifests itself in a production environment. When making one item, for example a replacement leadscrew you would be making it to fit your existing nut, so you would approach the pitch diameter with that in mind and not some arbitrary theoretical dimension. Thank you for posting such interesting videos. 👏👏👍😀 Andrew

[@ThePottingShedWorkshop]

Your problems with a carbide insert are not surprising. That type of moulded insert has a relatively blunt edge and requires a certain amount of tool pressure to cut. Its a common problem with trying to creep up on a target diameter, the tool suddenly bites and chews off too much. If you grind up a 60deg HSS tool the really sharp edge will allow you to shave off half a thou at once.

[@samuelfielder]

That may be the cause of non-repeatable results. I was using it because it is a full profile insert and don't know how I could grind an HSS to full profile. Also lazily not using the 29 degree topslide increases tool pressure.

[@ThePottingShedWorkshop]

@@samuelfielder Don't worry about full profile. Just grind a 60degree tool and give the sides about 8 degrees of clearance angle and use an oilstone to lightly round the tip (which will otherwise be fragile and snap halfway through cutting the thread). Use the 29.5 degree offset method to feed in to reduce pressure on the tip. I've done many threads this way and you can sneak up on the exact PD you want. Don't worry about the crest profile as it adds very little to the strength of the thread, just wipe a smooth file over the revolving thread and wire brush clean before measuring the PD. I haven't seen how you're measuring PD but if you haven't got a set of thread wires, buy some as they're cheap and work fiine. Just takes some pratice unless you have 3 hands. As for cutting depth, I usually feed the compound in by about 5 thou (0.127mm) for the first 25 thou, then reduce the cut depth by half, ending up with an infeed of about 1 thou until the final pass. Don't also forget the tolerance range on the PD, aim for the middle. Commercial bolts and nuts are made to fit this entire range, hence they're a bit sloppy.

[@chrismayhew3651]

Hi Samuel, The pitch diameter you quote is the maximum, the range is 9.188 to 9.070. Someone else made a comment about deburring which can make a real difference. Also, the deeper you go with each pass the more engagement with the work piece which will create more tool pressure. You mention about taking a deeper first cut then taking smaller cuts as you get in which I think would help. I also think it would be worth while trying to cut the thread at the 30 angle. My thinking behind this is that there will be less tool pressure when cutting using the "traditional" way. A CNC lathe cuts the thread square on using carbide inserts and I have seen other people do this on a manual lathe and I do it myself so I don't consider it an "improper" method. You know you have wear in the bed ways, you probably have some in the saddle and cross slide and / or the gibs are worn or not adjusted quite right thus allowing the saddle or cross slide (or both) to move some small amount resulting in these different results between one attempt and the next even though you are following the same process. Using the 30 degree method will lesion the tool pressure and might result in a more consistent result from one sample to the next. If you don't have a copy of Machinery's Handbook I highly recommend getting a copy. It has a great section on threading and it you study the diagrams you will get what I am banging on about. Regards, Chris

[@samuelfielder]

Good to know there's an acceptable range of pitch diameter. Agree using the 29 degree thing gives lower tool pressure. I did use that to begin with (in first two videos, so it didn't solve the problem) but got bored with it. I do have a copy of Macinery's Handbook, year 200 edition, but have not opened it! Such old fashioned, forbidding, typography!

[@samuelfielder]

Actually I've just checked the Machinery's Handbook for Metric external threads M10 1.25mm threads and it gives pitch diameter 9.042 to 9.160 mm, so notably smaller than I was working to.

[@davemason6273]

Hi Samuel, I don't see you using thread pitch wires. My apologies if I missed it! Used with international standards from charts such as in the Machineries Handbook you will know exactly if the threads are within specification.

[@samuelfielder]

I was using a Mitutoyo thread micrometer, which I hope is as good as thread wires and has advantage you don't need three hands. That was covered in the first video kzbin.info/www/bejne/omavpZKBds1km5o

[@davemason6273]

@@samuelfielder Even better. That makes more sense, I should never have doubted it :-) apologies I missed it. If the thread is within spec then that would surely suggest the internal thread that you're testing it on is not!

[@sigurasg]

Burring of the workpiece may produce inconsistent results. Look at Joe Pie's KZbin channel and see how he goes about testing threads. I seem to remember there's some deburring in there. Also the carbide thread cutting tools will have a minimum reasonable depth of cut. Below that, you'll be rubbing or burnishing the work, whereas once you exceed it by a little bit, the tool may dig in and produce a deeper cut than what you dialed in.

[@NightsReign]

Ok, I posted a comment on the first video, but after watching all 3 parts, and reading every single comment (it's entirely possible I missed a few), I am scratching my head as to how not a single person, in watching you vanish down this rabbit hole, thought it prudent to try recapping the basics with you. Occum's Razor appears to have been eschewed, in favor of watching you turn all of your 10mm bar stock into swarf. So, now I'm forced to tsktsk the regular viewers. __ Since I'm still over a year late regarding commenting on this video, I can only hope someone has since pointed you toward measuring runout on your spindle, and your 3-jaw chuck (I don't know if you have a 4-jaw or a collet chuck for verification purposes), but all of the calculations you ran through seem to indicate (on top of the tool/work deflection you already confirmed) significant runout needing addressed. Perhaps a bolt/bolts have been under/over-tightened, perhaps the chuck, spindle, or mounting bracket need to be reground, perhaps one of the jaws is dinged or dented or uneven. We won't know which issue(s) might exist until you can get an indicator mounted up and begin chasing down leads. Have you been able to verify that your toolpost is correctly leveled on center? It's not likely to be the sole cause of your plight, but it could be 0.001-0.005mm above/below centerline, and/or even tilted at an angle (due to wear in your compound and/or cross-slide ways). Since nobody suggested these very basic sanity checks be conducted, I'm concerned...

[@samuelfielder]

I can hardly remember now without rewatching my own videos which I don't fancy doing. As far as I remember I thought the problem was pitch not taper. I think in one of these videos I eliminated the possibility of taper by cutting 1cm off the end of the thread and demonstrating that the problem recurred 1cm further along the thread.

[@davidmartin805]

Hi Samuel, threading at this level is not all that scientific, my advice would be buy a thread file and some emery tape and make each part 'bespoke' Unless you have to manufacture quantity it really doesn't matter. Commercially available Nuts and bolts aren't all that precise and for general hobby purposes a few 'thou on nuts and bolts is usually ok, screw cutting comes into its own on bigger sizes. Be assured that your filming is very entertaining and I always watch. Just for fun, if you have an ER 32 collet chuck, make a test piece for the nut to screw on, from memory I think it's M50x2mm pitch. feel free to ask anything, I may be able to help, regards David

[@samuelfielder]

I do have an ER32 chuck. Cutting a bigger diameter thread would be fun, and have advantage that at given spindle speed the cutting speed would be faster and hence more appropriate for the carbide insert.

[@janstaines5989]

Hi Samuel, as I understand it, the diameter of the crests of the threads matters little (within reason), provided it smaller than the ‘roots’ of the thread in the nut. What is crucial is the pitch diameter as I’m sure you know. Wear in the ways will make both vary along the length, it might be worth checking along the length. Additionally, runout in the chuck causes misalignment of the axes of the shaft and the thread flank helix, which can make threads measure on spec, but not fit in the real world. I did not watch your preceding videos, so my apologies if you’ve covered this already!

[@samuelfielder]

Thanks but I don't understand this: "runout in the chuck causes misalignment of the axes of the shaft and the thread flank helix". Surely, however much runout the chuck has, anything turned on it is coaxial with the spindle axis - so long as you leave the part in the chuck.

[@janstaines5989]

@@samuelfielder yes absolutely, I must have missed the part where you made that clear, apologies. Depending on the condition/runout/side play of the spindle bearings, there could be some intermittent error there too?

[@daveharriman2756]

I've enjoyed your journey discovering the ins and outs of thread cutting, but as Mr Pete would say, "Well, I think I've beat this subject to death!" seriously though, you've ended up with a nice looking thread on that close up. Did you compare your finished part thread diameter with a commercial 10mm bolt? I remember in 1972 when our factory started using metric instead of Whitworth, BA and BSF, we thought there was something wrong with the metric fasteners, as they all seemed a bit sloppy, but we found out that that's just how they are, anyway best regards, Dave

[@steamtrainmaker]

Hi, Is there wear in the lead screw , nut engaged no lathe running will the saddle move at all it should not. Try it up the tail stock. The half nut has some adjustment on them by way of letting them shut tighter on the lead screw.. All so try hss tooling. . Check pitch 1 turn of chuck with DTI on saddle see it moves correct amount, try different pitch 2mm Is there end moment on the lead screw. Steve

[@samuelfielder]

I don't see that backlash between the lead screw and the half nut could affect the pitch, provided one engages the half nuts with the tool a small distance off the workpiece so that any slop is taken up before the tool syatys cutting. I did check the pitch using a dial gauge and another method in part 2. It's actually hard to do with sufficient accuracy for the very small difference I observed: 1.25 - 1.247 = 3 microns. Probably this difference doesn't matter.

[@billyboybenson8584]

Hi Samuel would you consider cutting a imperial thread and compair, it might be the case that your Harrison lathe is more accurate cutting an imperial thread? I have used a Harrison m300 at work on and off for 30 years and I think that they are a great little all round lathe so I would keep as you would have to spend a lot of money to get a little better!! Keep up the good work and don't loose heart as it is difficult to screw cut perfect threads .

[@samuelfielder]

Mine is a metric lathe. I did work out the gear ratios in part 2. They are exact: 4.8 spindle revs for one leadscrew dev. Leadscrew pitch 6mm, 6/4.8 = 1.25 exactly. I've never cut an Imperial thread yet, but would have to keepo the halfnutsw engaged.

[@neffk]

I'm not sure about your pitch diameter measurement. By measuring the OD aren't you getting an indirect reading of the pitch diameter? I suppose you might be OK if you're using an insert that forms the crest of the thread. Also, different depths of cut will give you different results. Consider a solid tool post riser instead of the compound. It is traditional to have a compound mounted at all times but there's also a million little tricks that machinists have learned to deal with the lack of rigidity that it introduces. Buying an over-spec lathe is probably the first trick. But I don't want to learn the tricks---I just want to make parts. Get out a 0.01mm indicator and find the slop in your system. Clamp a 300mm long bar in a tool holder, mount the indicator on the ways, and measure the deflection of the tool, tool post, top of the compound, the cross slide, and carriage. If there's any slop, you'll find it. Measure in all directions. Pull and push on the bar, watching to see if the indicator springs back or if there is hysteresis. Some spring is inevitable but if it doesn't spring back, there is is motion that is not due to the intermolecular forces in the iron lattice of the lathe's components. It's a bit of a project but I don't know any way to avoid it. I added some oil ports to my cross slide and when I re-assembled, there were vibrations that I thought I'd already fixed. Sure enough, the tool post bolt wasn't quite snug enough.

[@samuelfielder]

I used a Mitutoyo thread micrometer to measure the thread pitch. I was measuring the thread o/d because my lathe cuts a taper due to bed wear. Don't think lack of rigidity is a big problem for these threading tests, although it clearly is if one gets chatter or has trouble parting off.

[@tiisetsomolokoane4198]

Good day it's Isaac can you please help me cos I used to cut threads on my centre lathe but now it doesn't wanna to change pitch it only allows to cut 1 pitch what can be wrong

[@samuelfielder]

Well what lathe do you have? The pitch of thread that it will cut is determined either by (a) change gears between the spinlde and the lead screw, or (b) by a gearbox if there is one.

[@ianthompson367]

Hi Samuel, I'm not sure if this has been mentioned or indeed if it is practical, but what about swapping the leadscew end for end .. It's not going to be worn at the tailstock end,,,

[@samuelfielder]

I had not considered that. I guess I must have a video of the leadscrew lying on the floor somewhere but I can't find it. I'm pretty sure it isn't the same at both ends, but uncertain if the ends could be modified to make a swap feasible.

[@ianthompson367]

@Samuel Fielder Well, I think that there's half a dog clutch at the gearbox, and probably a plain shaft into the bearing at the tailstock end. I would estimate that a man of your intellect and insight might come up with a method to modify it. Though you are likely to need a working lathe so to do... ;->

[@samuelfielder]

@@ianthompson367 Could probably machine the lead screw in my M300. It would fit through the headstock, and only a small amount at each end would need machining. I think it's an idea to be held in reserve. At moment I'm not really convinced that the lead screw is worn at all, or sufficiently to cause the problems I met. I just need to cut the threads to the right depth and all will be well for most applications. To cut a new lead screw for the top slide or cross slide I'd have problems with turning it all to the same diameter, but that's down to the worn bed not the lead screw.

[@ThePottingShedWorkshop]

As I write I've not watched the first two parts, so apologies if this has already been suggested, but have you checked for wear in the leadscrew? You can do this fairly easily by rotating the chuck by hand and measuring the travel with a dial indicator. Quicker and more quantifiable than cutting a thread. Wear on the bed would change the pitch dia, but not the pitch.

[@samuelfielder]

I have checked for leadscrew wear in earlier episode but it is hard to do accurately due to published maximum permissible error in Mitutoyo dial gauges. The pitch error I measured was almost within the dial gauge limits: 1.250 - 1.247 = 3 microns. Either not really an error, or too small to account for nut jamming, I think.

[@Walter-v5j]

Surely it is configured for 20 threads per inch, greetings from Uruguay.

[@samuelfielder]

ha ha

[@bobastill8538]

have you checked the pitch in the long nut? i did a similar thing as i did the long bush with a tap but thread cut the shaft turned out i had used a 1.50 tap rookie mistake but i just had to laugh.

[@samuelfielder]

Yes it is correctly made with a 1.25mm pitch die, and fits the commercial M10 1.25mm bolt.

[@ginozorzetto6497]

2 possono essere le cause, o stai commettendo un'errore in fase di filettatura o quando hai rimontato la scatola del cambio qualcosa non e' andato al suo posto....il mio e' perfetto in tutti i passi. Verifica anche tutti i laschi del carro longitudinale e trasversale.

[@toolman7540]

Tap and die set work well for small work like this

[@samuelfielder]

Yes indeed, but I wanted to practice single point threading whilst not chewing up too much stock.

[@user-jm4nh7um9b]

Это влияние тонкого вала?

[@samuelfielder]

Возможно, посередине образовалась выпуклость из-за изгиба вала. Но имел живой центр в более поздних экспериментах.

[@davesden]

Where are you Sam?

[@samuelfielder]

Near Southampton.

[@davesden]

😂. I meant we are waiting for more videos. We miss your content.

[@samuelfielder]

@@davesden Oh, I see. Sorry for the interruption of service. I'm not sure what else to do.

[@davesden]

Make something. It's nice to watch you work.

[@StraightThread]

This fellow explains the wear on the leadscrew that may interest you. kzbin.info/www/bejne/nJixhYSwhbCdjKM