Milling Machine Trammel

[KWILParticipant @kwil]

[KWILParticipant @kwil]

ME 4495. 14th November 2014. The article describes the making and assembly of a useful and adequate Machine Trammel.

However the time taken to ensure the accuracy by the stated method of post assembly machining is, I suggest, excessive and whilst it will produce a useful tool, such a tedious approach is unecessary.

Make the base and fit the pillar as accurately as you reasonable can.

Fit both clocks and tighten the pinch bolts. Mount the trammel in the collet or chuck on the machine spindle, lower to the reference surface so that the measuring tips touch. Set trhe FIRST clock to a known value, note this value. Rotate the trammel 180 degrees so the the SECOND clock tip lies exactly where the first tip had previously touched.(perhaps a mark will have helped to identify the exact spot.) Set the SECOND clock reading to be EXACTLY the same as the FIRST [if your pillar is far out of square you may have to release and retighten the SECOND pinch bolt to achieve the same value].

You now have a very accurate trammel where the measuring tips are located precisely in the same plane and at right angles to the pillar

Now proceed with the HEAD alignment as described.

[Eric CoxParticipant @ericcox50497]

Why not use a single clock, rotate the trammel 180 degrees and see if you get the same reading in both positions.

Alternatively can you use a digital angle gauge, Set the gauge to zero on the mill bed, lower the quill and place the gauge on the same and see if it reads 90 degrees or isn't the angle gauge accurate enough.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Eric Cox on 13/11/2014 09:47:02:

Alternatively can you use a digital angle gauge, … or isn't the angle gauge accurate enough.

.

Eric,

I would not trust a [typical "hobbyist"] digital angle gauge for this job

… The accuracy and repeatability are simply not good enough.

Of course: That might be how they assemble them wrong in the first place

MichaelG.

[VicParticipant @vic]

Posted by Eric Cox on 13/11/2014 09:47:02:

Why not use a single clock, rotate the trammel 180 degrees and see if you get the same reading in both positions.

That's what I use. It probably takes longer but should be more accurate as you're measuring over a much larger distance.

[Gary WoodingParticipant @garywooding25363]

Eric,

The normal digital angle gauge (Wixy type) has a claimed accuracy of 0.1 degrees, which translates to an error of almost 0.007" in 4". That's nowhere near good enough for a mill.

With regard to the MEW article, what's wrong with holding the pillar in the lathe chuck and taking very small cuts from the bar? Yes, I know its an intermittent cut but that shouldn't present any problems.

[Nigel BennettParticipant @nigelbennett69913]

Gary's suggestion is valid if you can swing the bar in your lathe! The longer it is, the more accurate it is. Per contra, the more pain it is to get the head set level with a large distance between clocks when you use it…

[Gary WoodingParticipant @garywooding25363]

The suggestion in the original article was to mount the bar on a faceplate and trim the pillar, so it obviously must be possible to swing the bar.

The longer the bar the more accurate is the trammelling.

[Neil WyattModerator @neilwyatt]

I agree with Ken, all that matters is that both clocks read zero at the same point, then assuming they are accurate when they read the same the spindle is vertical.

The reason for using two clocks is two-fold;

1. It's easy to upset the position of the clock when spinning it over the t-slots. using a sheet of paper helps, but if your radius is such the clock goes off the side of the table, it isn't very reliable. I made a small clip out of apiece of aluminium to let me raise the probe while I swing the clock round.
2. You have to cut and try – move they pillar, swing the clock, try to adjust by half the error, swing back the other way, ditto repeato ad infinitum…

The process can take ages! I estimate about 10 minutes last time I did it, but I was taking extreme care to get it dead right as I wanted to flycut a 2 1/2" x 3" surface really flat.

With the dual clock method I imagine you can always see exactly where you are and trim the last thou away with ease.

Sadly I only have one dial gauge so I'm stuck with the slow way for now.

Neil

[Gary WoodingParticipant @garywooding25363]

Posted by Neil Wyatt on 13/11/2014 14:47:28:

I agree with Ken, …

Sorry Neil,

I must have missed something – who's Ken?

Anyway, yes, you can do it with one clock but, as you clearly stated, it creates problems. The only objection to the 2 clock method is that it requires 2 clocks. They don't have to be identical, just accurate enough, Second hand ones will do.

[Neil WyattModerator @neilwyatt]

Sorry KWIL!

Mine is a 1960s vintage KS one from the DDR.

Neil

[LimpetParticipant @limpet]

For those with plenty of the folding stuff I've seen adverts for a dual dial gauge to set up the mill trammel from Chronos £80 or £90 from memory

Lionel

[VicParticipant @vic]

Can I ask why these tools are made so short? Surely it would be more accurate if the horizontal bar was at least 10 – 12 inches long? The head on my mill only tilts L/R so if I was to make one of these that's what I'd do.

[I.M. OUTAHEREParticipant @i-m-outahere]

I use one of those Edge technology units and it works well, they supply a small magnet that is placed on the table and you zero the gauges off the top of it then lower the spindle until one [ or hopefully both] gauges read zero. if they don't then adjust the tram until they do . The only thing that irritates me is that the magnet will pick up any chips or fine iron dust and getting it off the damned magnet is near impossible !

I found also because of the crud sticking to the magnet it was best to slide the magnet into position to ensure any junk caught under it will get wiped off then wipe the top surface as well , I then sweep the indicator nib across the top of the magnet to make sure nothing is caught between the two before setting zero , I also sweep the indicator nibs around on the table surface when I lower the unit down to take a reading – sounds like a hassle but once you have done it a few times it becomes second nature and the readings will be consistent .
Now I just need to get the fine metal dust off that magnet and I will be happy , may try compressed air .

Ian

[VicParticipant @vic]

Thanks John, I've got a Mitutoyo indicator somewhere that I never use so perhaps I'll see if I can get a twin for it and make a rig.

[John HaineParticipant @johnhaine32865]

I use a Verdict dial indicator using one of the attachments gripped in a collet. It sweeps about a 6 inch circle. To avoid the problem of the tip catching in the tee slots, I place a plate of 6 mm float glass (my first surface plate) on the top of the table for the indicator to sweep. Float glass is dead flat and parallel, as you can check by spinning the plate whilst keeping the indicator still, the needle doesn't move. It seems pointless to me to spend money, or time making, a gadget that you can equally well improvise and only use occasionally.

[John HaineParticipant @johnhaine32865]

Getting dust off magnet – try duct tape.

[Douglas JohnstonParticipant @douglasjohnston98463]

Why use a magnet on the mill when a disc of non magnetic metal would work just as well. If it was a vertical surface I could see the point, but otherwise magnets just make life more of a hassle.

Doug

[JasonBModerator @jasonb]

Magnet ensures you set the gauges in exactly the same place, same as KWIL suggested to mark the table.

I did see a suggestion to use an outer shell from a large bearing so you can swing a single gauge without the slots getting in the way. I think the guy used a 8-9" one that was also surface dround to make sure it was not distorted when the bearing was taken apart

J

[John HaineParticipant @johnhaine32865]

Jason, why go the trouble when float glass is is so chap, flat, and self checking?

[JasonBModerator @jasonb]

Is it that flat? Take Pilkington "Plateau" which is there high quality extreamly flat float glass and that can have upto 1.5mm per metre allowable bow. Standard float glass would be even more.

J

[Anonymous]

Bother, JasonB beat me to it. The Pilkington spec on thickness isn't that good either, ±0.13mm I think for 10mm sheet. I rarely tram my mill; when I do a check I use a DTI and four parallels. And before anyone asks the parallels were made by the late father of a friend of mine, who was a tool maker. I can't detect any variation in thickness across all four, using a micrometer. So probably within a tenth or so.

Andrew

[Tony Pratt 1Participant @tonypratt1]

You guys are making a simple task complicated. Start with a clear milling table and stone off any high spots, centralise the spindle approx. middle of table.

Set up a simple ‘L’ bar & DTI, swing the spindle until the bar is in the full left position. Place some sort of ground part [vee block, slip gauge, parallel etc.] between DTI and table, raise table to zero DTI.

Swing DTI through 180 degrees and slide ground part gently under DTI, adjust vertical head angle if necessary.

Repeat until both DTI readings are zero/zero or what you judge is good enough.

Tony

[VicParticipant @vic]

You don't even need a DTI. I was taught to use a "z" bar and two fag papers to check the mill head for vertical, but that's another story.

[I.M. OUTAHEREParticipant @i-m-outahere]

Posted by John Haine on 14/11/2014 10:33:48:

Getting dust off magnet – try duct tape.

Thanks John !

Now why didn't think of that !

Now i can clean off my magnetic welding clamps as well!

Back to the OP

The magnet wont move around when you rotate the tram head and it also draws the steel nib down onto it ensuring accuracy.

Ian

[Author]

Posts