Likes Likes:  0
Needs Pictures Needs Pictures:  0
Picture(s) thanks Picture(s) thanks:  0
Results 1 to 8 of 8
  1. #1
    Join Date
    Mar 2017
    Location
    Brisbane
    Posts
    245

    Default Tolerances on scraping dovetails?

    So I'm tackling the table of my Chinese mill for scraping practice (HM-46).

    The table is flat and I've progressed to scraping the flats of the dovetails parallel.

    They are now parallel to the table top however have some rise/fall along their width, and I was wondering what is acceptable. It is tricky getting it parallel in x and y. One always seems to be out!

    One side goes from "0" on the indicator to "+2" tenths. The other then goes from "0" to "-4". This is measuring off a dead accurate 1-2-3 block to average the height. As the block can't get under the dovetails, it would probably be a little worse than this.
    Do I need them within say 2 tenths across their width?
    On a positive note across the length of the table - 730mm or so - they are parallel to the table top within 1 tenth.

    Picture to attempt to explain what I'm illustrating, dimensions obviously exaggerated:


    edit/update - I think I know what to do.... keep it parallel to the table but make them flat across the width. Height doesn't matter so much. Matching it to the saddle would be easier with different heights so long as the widths were parallel. Any tips on how to fix widths would be helpful..... I think my tactic needs to be scrape out the middle first, then the far end.

  2. #2
    Join Date
    Aug 2008
    Location
    near Rockhampton
    Posts
    6,216

    Default

    Are you meaning to say you have twists in each way.

    Generally tolerances would be so much over a distance as nothing can be made perfectly flat.

    But yes they need to be just flat and both ways pointing in the same direction. Using your picture, the ways could look like that and still work perfectly adequately but they have to be pointing in the same direction.

    When scraping remember to be careful when spotting, it is very easy to rock the straight edge and you end up low at the ends. and high in the middle. Also when doing narrow ways with a wide straight edge to roll the edge so to speak so you get a false reading..

    I am never afraid to spot something twice or three times without scraping to see if a different approach to spotting makes a different spotting pattern.
    Gold, the colour of choice for the discerning person.

  3. #3
    Join Date
    Mar 2017
    Location
    Brisbane
    Posts
    245

    Default

    Quote Originally Posted by .RC. View Post
    Are you meaning to say you have twists in each way.

    Generally tolerances would be so much over a distance as nothing can be made perfectly flat.

    But yes they need to be just flat and both ways pointing in the same direction. Using your picture, the ways could look like that and still work perfectly adequately but they have to be pointing in the same direction.

    When scraping remember to be careful when spotting, it is very easy to rock the straight edge and you end up low at the ends. and high in the middle. Also when doing narrow ways with a wide straight edge to roll the edge so to speak so you get a false reading..

    I am never afraid to spot something twice or three times without scraping to see if a different approach to spotting makes a different spotting pattern.
    Thanks RC for the reply. No it's not a twist in each way, it's just although they are flat down the length of the table, they are inclined towards and away from the centre of the table a very small amount.

    You did just make me realise that they might not be parallel all the way down the table though! (highly unlikely they are) Which would make matching the saddle to them impossible.

    So, I am fairly sure now I need to make them flat across the width as well as down the length of the table. I am not sure however that the ways need to be the same height from the table top (the bottom of the picture here). I think it's fine for them to be stepped, so long as each plane is parallel in 3D to the table.

  4. #4
    Join Date
    Mar 2017
    Location
    Brisbane
    Posts
    245

    Default

    My next difficulty is step scraping for fine adjustments. I am pretty good at step scraping to adjust things out by a thou or 3-5 tenths, but to get them right when they're only out a tiny amount is tricky.

    One seemingly good method I tried is I blued up the surface then read the height of the blue roughly. Usually the difference from blue to non blue is about 2 tenths. Then you scrape off all the blue and you drop the surface by 2 tenths. But how do you do 1 tenth? Do you do a pattern and not knock it all off? Or do you perhaps blue it up, then rub all the blue off your straight edge, and use it to shine the high spots - then knock those off? Or perhaps use a very light blue so the difference from blue to non blue is 1 tenth....

  5. #5
    Join Date
    Aug 2011
    Location
    Melbourne
    Posts
    4,779

    Default

    If you are only out by 2 tenths then I would leave it alone. With a geared head mill, the cutting forces from moderate a cut will flex the colunm by more than 2 tenths anyway.

    Seriously, if it's within 2 tenths then that's bloody good.

    Edit: for further reading, try looking for a youtube video by Mueller Nick. Has an excellent video on interpreting the blue transfer patterns.

    Simon
    Girl, I don't wanna know about your mild-mannered alter ego or anything like that." I mean, you tell me you're, uh, super-mega-ultra-lightning babe? That's all right with me. I'm good. I'm good.

  6. #6
    Join Date
    Aug 2011
    Location
    Melbourne
    Posts
    4,779

    Default

    This is the video I was thinking of. It may help you.
    https://www.youtube.com/watch?v=y2AUew5Evxc

    Simon
    Girl, I don't wanna know about your mild-mannered alter ego or anything like that." I mean, you tell me you're, uh, super-mega-ultra-lightning babe? That's all right with me. I'm good. I'm good.

  7. #7
    Join Date
    Mar 2017
    Location
    Brisbane
    Posts
    245

    Default

    Hmm, no doubt true for some cuts but for all the crap these mills get (and deservedly so for the build, christ) some people achieve quite a lot on them. Stefan Gotteswinter for one. I think they could be accurate if your finish cuts were on the light side and you got one to the state they arrive from Optimum in Germany - that is, better ground ways and a higher quality of build.

    I spotted up the saddle I will have to post a video, it's hilarious. It ROCKS by a thou or more. I think the ways which should have been parallel varied by something like 7 thou.

    I plan on boxing in the whole z axis and adding plates at a later stage. This guy did it - Ultimate rf-45 video for laughs - https://youtu.be/k2F6100RKOE?t=479
    In the future it's likely I'll just use this as a learning project and put more time and effort into a different machine later on. All depends on size / weight I can have at my premises really. I think these are decent mini mills once scraped in.

  8. #8
    Join Date
    Aug 2011
    Location
    Melbourne
    Posts
    4,779

    Default

    Quote Originally Posted by j3dprints View Post
    Hmm, no doubt true for some cuts but for all the crap these mills get (and deservedly so for the build, christ) some people achieve quite a lot on them. Stefan Gotteswinter for one. I think they could be accurate if your finish cuts were on the light side and you got one to the state they arrive from Optimum in Germany - that is, better ground ways and a higher quality of build.
    I think you have misunderstood me. I'm not canning these mills. I have one and have been using it for 8 years. I have done lots of stuff on it and learnt much about machining on it. But, it does have limitations, most of which I have been able to allow for or work around.

    I never checked the flatness of the ways or how parallel they are/were because I was scared I wouldn't like what I found. Ignorance is bliss! However, I know for a fact that with a DI attached to the head and touching the table will show significant deflection (greater than your 2 tenths) with the pressure from one finger.

    Simon
    Girl, I don't wanna know about your mild-mannered alter ego or anything like that." I mean, you tell me you're, uh, super-mega-ultra-lightning babe? That's all right with me. I'm good. I'm good.

Similar Threads

  1. Dovetail straight edge & scraping dovetails
    By j3dprints in forum METALWORK GENERAL
    Replies: 4
    Last Post: 10th Apr 2018, 09:20 AM
  2. Thread tolerances ? 0.2% critical ?
    By steamingbill in forum METALWORK GENERAL
    Replies: 6
    Last Post: 25th Sep 2016, 01:04 AM
  3. B grade Engineers square, what are the tolerances?
    By simonl in forum METALWORK GENERAL
    Replies: 4
    Last Post: 2nd Aug 2012, 10:50 PM
  4. Swiss Scraping or scraping from the shoulder
    By Ueee in forum METALWORK GENERAL
    Replies: 3
    Last Post: 30th Jul 2012, 01:31 AM
  5. How to Id QC toolpost dovetails
    By Grahame Collins in forum METALWORK GENERAL
    Replies: 16
    Last Post: 6th Feb 2011, 10:16 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •