Screwcutting problem on Colchester Student 600

  1. esb New Member

    Messages:
    8
    Location:
    Somerset, UK
    Can anyone help with advice on the following? I am having problems cutting threads successfully due to the tool seemingly cutting a different path on each pass. The result is a mess of multiple pitches, rather than a constant pitch thread. I have tried keeping the half nut engaged throughout and reversing the machine for each pass and also tried using the threading dial. Only tried cutting using the English Thread settings. The pictures attached of the magnified "thread" show an attempt at 12TPI on a piece of 3/4" brass rod.

    There seems to be an issue causing a lack of repeatibility in subsequent cutting passes. I can't find play on the slides, but wondered if the problem could be a missing tooth in the screwcutting gearbox, or play elsewhere (leadscrew) which could be the cause? Any pointers would be most welcome!

    Lathe is a Colchester Student 600, Mk2 square head.

    IMG_2600.jpg IMG_2601.jpg IMG_2598.jpg
     
  2. grim_d

    grim_d Unlikeable idiot.

    Messages:
    3,093
    Location:
    Scotland - Ayrshire
    How are you advancing the cutter?
     
  3. Matchless

    Matchless I started with nothing, still have most of it left

    Messages:
    1,625
    Location:
    Essex UK
    are you disengaging after each pass or just pulling out and reversing the motor? I am no machinist but I was told to pull out the tool, stop, and reverse the motor before having another go, I have a well worn roundhead Colchester student, and a Harisson 12" , not tried pulling out with the 12" yet, waiting for the 3 phase socket to be wired up,
     
  4. Screwdriver

    Screwdriver Forum Supporter

    Messages:
    6,895
    UK London
    Zero the cross slide and advance at 29 degrees with the top slide for your first scratch pass.
    Wind out the cross slide for clearance as you run the saddle backwards, then set the cross slide to zero and advance on the top slide.

    Looks like you might be feeding with the cross slide after winding out the top slide.
     
  5. slim_boy_fat

    slim_boy_fat Forum Supporter

    Welderpaul likes this.
  6. Pete.

    Pete. Member

    Messages:
    9,808
    Location:
    Kent, UK
    When you are keeping the half-nuts engaged, how are you backing the tool up to the start of the cut? Not with the tumbler reverse I hope?
     
    Screwdriver and Brad93 like this.
  7. Seadog

    Seadog Forum Supporter

    Messages:
    6,968
    Location:
    NE London - UK

    That's for metric or Unified. The compound should be set to around 27° for an Imperial thread.
     
  8. Brad93

    Brad93 M J B Engineering

    Messages:
    10,372
    Location:
    Essex
    You need to run the whole gear train backwards. That means not changing gears, no disengaging the gearbox from the headstock, not using tumbler reverse.

    the only thing you can use is the motor.
     
    esb and Screwdriver like this.
  9. esb New Member

    Messages:
    8
    Location:
    Somerset, UK
    Good point Brad, I had been using the tumbler to reverse the geartrain. Didn't realise the implication of this so I tried again, this time stopping the lathe with the start/stop lever and brake, then backing the tool out of the work, stopping then reversing the motor to traverse back to the start point, re-setting the cutter depth, stopping the motor, then re-starting in the opposite direction and running a subsequent pass with pressure on the apron handwheel to limit any play. All the time with the leadnut engaged. I achieved a good result doing this, so very pleased that I have managed to achieve someting at last! Thanks

    Although this system has worked, it is quite a laborious procedure and relquires a lot of stop-start on the motor which concerns me a little with regard to windings overheat. I could use the motor reverse switch with the motor running (3ph motor), but I feel that probably puts more load on the windings than the stop-start method.

    When cutting imperial threads with an imperial lead-screw, I understood that when cutting an even TPI it should be possible to disengage the leadnut at the end of each pass, manually wind the carriage back, then re-engage the leadnut on any division on the dial. This was my initial method and created similar appalling results. With my first good thread completed, I looked again at engaging the leadnut on each division in turn and ran the tool up to but clear of the threads to check the accuracy. This seems to create completely random results, with the tool tip meshing anywhere between a crest and root on each separate attempt. Clearly something is wrong, either with the lathe, or more likely my understanding. The leadscrew is 6TPI, so definitely an Imperial screw.
     
    Hopefuldave likes this.
  10. Seadog

    Seadog Forum Supporter

    Messages:
    6,968
    Location:
    NE London - UK

    I assume you don't have VFD? If I ever need to use the reversing method I just flip the VFD into reverse, it ramps down stops and then goes into reverse. It's a very quick process and you can even accelerate the return if your coordination is up to it.
     
    fizzy, Welderpaul and Hopefuldave like this.
  11. Pete.

    Pete. Member

    Messages:
    9,808
    Location:
    Kent, UK
    For 6, 12, 18, 24, 30, 36 tpi etc (i.e. any multiple of the leadscrew pitch) you don't need a threading dial - you can engage the half-nuts any time you please so for your 12TPI thread you should not have been able to get it wrong (except by using the tumbler reverse).
     
  12. Pete.

    Pete. Member

    Messages:
    9,808
    Location:
    Kent, UK
    I just thought - if you have metric translation gears installed, then your 12tpi would be double-cutting. Check the tooth counts on your end-gears.
     
    esb likes this.
  13. esb New Member

    Messages:
    8
    Location:
    Somerset, UK
    Seadog, no VFD on the lathe, just a rotary phase convertor. VFD definitely the way to go!

    Pete, thanks for that. I did wonder if there may be a mix of imperial and metric gearing but not sure how to check.

    I'm missing something obvious wrt the threading dial. I ran the saddle back and forth by hand with the threading dial engaged and measured the displacement per division at .500 with repeated accuracy of less than 0.002". There is 0.010" lateral play due to the leadnut and 0.023" lateral float on the leadscrew, so a total of .033 backlash. Seems a lot to me, but may be as expected? I would have thought it should not matter in any case if it's always allowed for prior to engaging the cut.

    However, I also tried engaging the leadscrew on random divisions after running the lathe back by hand by different amounts and when running back up to the thread under power, meshing accuracy with the test thread was completely random and seldom in the correct place. Your theory about metric translation gears sounds quite plausable. Not sure what I should have fitted, can you shed any light? Currently the driving gear is 22T, idle gear 120T and gearbox input driven gear is 35T. The only documentation I have states that the normal gearing should be 42T driving and 35T driven, but not sure if that's reliable info. It looks like I may have a 22T gear in place of either the 42T or 21T options the gearbox ratios are designed for. Not sure why they would be like that unless a previous owner had a need for a specific thread. I've found a 42T gear so will try that and report back!
     
  14. Agroshield Member

    Messages:
    1,365
    esb likes this.
  15. grim_d

    grim_d Unlikeable idiot.

    Messages:
    3,093
    Location:
    Scotland - Ayrshire
    Just whitworth.

    Standard metric and imperial is 60 degree Included angle.

    Acme 29 degree Included angle, trapezoidal 30 degree.
     
  16. esb New Member

    Messages:
    8
    Location:
    Somerset, UK
    Thanks Agroshield, useful info. Looks like I need to source a 21T driver gear. Meantime I should be able to use my 42T gear in it's place and set the gearbox for twice the TPI required. Excellent forum, been a good learing day, thank a lot everyone!
     
  17. grim_d

    grim_d Unlikeable idiot.

    Messages:
    3,093
    Location:
    Scotland - Ayrshire
    What the pitch of the changewheels? If its 14DP@14.5° pressure angle then I can make you that gear. (As long as it doesn't have any funny splines in the bore!)
     
  18. Pete.

    Pete. Member

    Messages:
    9,808
    Location:
    Kent, UK
    What end gearing do you have right now. The 21/42 thing should not cause your problem above.
     
  19. esb New Member

    Messages:
    8
    Location:
    Somerset, UK
    TGrim_d, it is splined, but thanks for the offer. I'd like to find a 21T gear, but not sure how common they are.

    Pete, what do you mean by end gearing?

    Currently running a 22 driver to a 35 driven. I was originally trying to cut a 20 TPI internal thread for a collet closer. Just experimented with the 12 TPI to give some clarity on the problem when things were not working out. When using the Colchester TPI formula with the 22/35 ratios I get 19.1 TPI rather than 20 TPI and around 11.45 TPI instead of 12 TPI. Would those errors not cause the problems I've seen when trying to use the threading dial? I'll have another go later with the 42T gear and let you know.
     
  20. Pete.

    Pete. Member

    Messages:
    9,808
    Location:
    Kent, UK
    Yes they absolutely would.
     
    Hopefuldave likes this.
Advertisements