How not to...

[kvom]

... tram a mill.   

It has been 6 months or so since I first trammed the head on the Bridgeport.  The last time I flycut something I didn't get the trailing cut lines, so the head was definitely off tilt a bit.  I had needed to advance the ram to drill a larger than normal piece that also necessitated removing the vise from the table. 

My bright idea  was to place my surface plate on the table and use it as the reference for tramming.  I got the tilt adjusted to less than .001", then replaced the vise and trammed the fixed jaw to .005" according to my DTI.  I was a happy camper.  Today I decided to check the "nod" of the head, and it was way out.  I then decided to measure the vise ways in the Y-axis, and was getting .010" over 6" travel.  Put the surface plate back onto the table and got a bit less, but still over .005" over the table width.

After that, I turned over the surface plate to see if there was swarf underneath, and saw the bottom of the plate has a rough surface, certainly not a good reference plane.  When I laid a 123 block on the table and indicated it, I had less than .001" over 3".

The upshot was that I needed to retram the head using parallels on the table.  I also discovered that my cheap DTI is a PITA, and totally unreliable; I used the better quality DI to tram the nod of the head.  OTOH, I found that the vise was in fact trammed correctly, so at least I didn't have to do that again.

So I will be in the market for a good quality DTI soon, and I won't be tramming using the surface plate again. 

[chuck foster]

thanks for the heads up on the surface plate..........................i was just looking at mine and thought it would be good for tramming the mill.

chuck 

[bogstandard]

When tramming or levelling a machine, the less you have between the bed or ways the more accurate the job will be.

By using things like surface plates and parallels, you are introducing automatic errors. You only have to look at the tolerances supplied with 'add ons', with parallels it is usually 0.0002". It doesn't sound much, but when you are after perfection, buy NOT having it there certainly helps.


Bogs

[Darren]

Does it really matter? You once told me that using a mill to surface grind was pointless because it couldn't hold the accuracy required.

How accurate is a milling machine, 1/2 thou, 1 thou over what distance?

[John Stevenson]

A while ago I bought an arbor press off Ebay from a firm that was closing down. When I went to collect it there were three surface plates laid outside red rusty.

The guy said I could have them as they were going to sell them but not now some goon had left them out in the rain.
Threw them in the Donald, did a detour round the grinders and gave the Lumsden man £20 to regrind all three.

Usual practice, he upside downed them licked the feet then skimmed the top to just clean up. Seeing as this machine regrinds it's own bed from time to time it has to be very close to perfect. Sold one and still got two.

I know these are just ground and not scraped in but neither is a sheet of plate glass that many of us started with and as an aside are gramite plate scraped or just lapped which is only grinding anyway

For what I do I'd have no compunction on using these to tram the Bridgy, hell I don't even know how true the table is after 30 odd years ?

John S.

[bogstandard]

Darren,

There is a great difference between surface grinding and setting a machine up to basics.

A standard mill cannot give you the continual accuracy of a surface grinder. It is all to do with the way they are made and work.

But when setting a machine up to basic settings (tramming, levelling etc) you should always use, if possible, the machines original ways or table. If say you put a surface plate onto the mill table, any error in that surface plate will be directly transferred to the machines settings. So say you tram on a surface plate that is sitting on the table, and that surface plate has a 0.002" error on it, when you tram, you will be automatically setting your machine up with that error.

If you are happy working with that sort of error, then fine, it is you that has to make the parts fit together.

I was just informing the members that errors in setting your machine is reliant on getting the most accurate results you can, otherwise, why bother in setting it up in the first place.

As an example, if your tram is out, how can you expect to bore or drill a hole truly vertical if using the quill?


John

[Darren]

Oh I agree with you about 2 thou, but you were talking about 0.2 thou. It was such a small error I couldn't see it rearing it's head in the scheme of things?
Not even my callipers can measure down to that. Although my dti can.


It's odd though, we can set the machine up to the ninth degree, use it and we loose that ninth degree pretty quickly.

[bogstandard]

Darren,

The tramming tool that I made can easily detect 0.0002", it just depends on the accuracy of the measuring instrument.

The last time I trammed and adjusted my mill was about 3 to 4 months ago, but because it is so easy for me, I check the tram before the start of each major job. Up to now, the tram hasn't moved at all.
It all depends on the amount of very heavy cutting that you do as to whether it gets knocked out or not.

John

[Darren]

Fair enough,

I was just thinking 0.0002 was so damned small....

I need to make one of your tramming tools methinks...