PDA

View Full Version : Tramming tool



Kd0afk
02-15-2015, 01:04 PM
I made a tramming tool
http://tapatalk.imageshack.com/v2/15/02/15/a26ff6d91846a28f9f69ecf32bf223f0.jpg
The way I made it was to drill the holes for the indicators and then the hole for the shaft mid way between the other two holes. I pressed in with loctite a shaft in the center hole, chucked it in the lathe and faced off the base so that it would be perpendicular to the spindle. I put an indicator on it while still in the lathe and it was dead nuts spot on.
I assembled it and zeroed out the indicators and mounted it in a collet in the mill and trammed the mill with it. Everything was in order but the first set of test cuts came out horrible. Long ridges told me that the mill wasn't trammed properly so I zeroed out the tool again and double checked the plumb of the spindle. It read right on but when I rotated the tool 180 degrees it was around .010" off. What the heck is up with this?

SirLesPatterson
02-15-2015, 01:25 PM
Just a novice thinking out loud here but maybe it's the drill chuck? I would try it in a collet.

EDIT: Nevermind, I see now you did use a collet. Sorry.

Kd0afk
02-15-2015, 01:30 PM
Only collets were used in the making.

lynnl
02-15-2015, 01:37 PM
......

I assembled it and zeroed out the indicators and mounted it in a collet in the mill and trammed the mill with it. Everything was in order but the first set of test cuts came out horrible. Long ridges told me that the mill wasn't trammed properly so I zeroed out the tool again and double checked the plumb of the spindle. It read right on but when I rotated the tool 180 degrees it was around .010" off. What the heck is up with this?
How are you zeroing out the indicators? Do you have a a plane surface that you know to be exactly perpendicular to the spindle? That would seem essential.

Maybe adjust the tram .005" in the other direction, make a trial cut, and if good re-zero against that surface.
You're probably seeing an accumulation of errors. ...that's my guess.

Kd0afk
02-15-2015, 01:39 PM
The tool is zeroed out by sitting it on a flat surface. I have a granite plate for that. The shaft indicated to be perpendicular to the base.

garyhlucas
02-15-2015, 01:49 PM
This reminds me of the guy who once asked me to borrow my level, to check his! I asked him how he knew mine was accurate and his was not? Of course a level is easy to check. Set it on a roughly level surface and observe the reading. Flip it 180 degrees and if it is in calibration it will read exactly the same.

I have seen these tramming devices and thought WTF? When you use two indicators like this you have introduced whatever error the two indicators may have, and whatever error the spindle, and collet may have. So in order to properly calibrate this device you need to swing it in a properly trammed spindle with the first indicator and zero it. Then you need to swing the second indicator and adjust it also to read exactly zero. As long as you leave it in the spindle it will be accurate. Take it out and all bets are off. There is a good reason that a single indicator is the most accurate way to tram a spindle.

lynnl
02-15-2015, 01:49 PM
The tool is zeroed out by sitting it on a flat surface. I have a granite plate for that. The shaft indicated to be perpendicular to the base.

Well, apparently something went awry. Either the "flat" surface isn't (maybe some dust), or the shaft isn't quite perpendicular.

Another possibility of course, is the assembly is somehow cocked in the collet so that the plane traced or swept by the two indicator tips is not perpendicular to the axis of rotation.

Kd0afk
02-15-2015, 01:51 PM
I will make a video and try and post it to explain what's going on.

Kd0afk
02-15-2015, 01:55 PM
I know how to check a level for accuracy and how to check and adjust a tri square for squareness, that is not the issue.

Toolguy
02-15-2015, 01:56 PM
All you have to do to zero it is put it in the spindle (preferably in a collet), zero 1 indicator, turn it 180 degrees and zero the other indicator. This way you are zeroing the indicators both on the same spot on the table, perpendicular to the spindle. Make sure the spindle is locked so it can't go up and down.

dp
02-15-2015, 02:28 PM
As a secondary test chuck it up in your lathe's tailstock and see if your stationary chuck face is square at each 90 point of rotation of the tool in the tail stock.

Kd0afk
02-15-2015, 02:39 PM
I'll try what's suggested.

TN Pat
02-15-2015, 02:51 PM
Yes. I made one of these devices as well, beautiful little helpers.

But the easiest way to use them is like this...

It helps to use a 1-2-3 block, thick parallel, gage block, small surface plate, some such - a good, flat surface, laid on the mill table.

Raise the table / lower the head so that one indicator touches the surface. Zero it. Preferably just touch it, without going more than a quarter-revolution, so as not to have to worry about how many revolutions and such.

Turn the tool 180 degrees, without moving the knee/head at all, and zero the other indicator on the same surface. Again, preferably without going more than one revolution, that cuts down on confusion...

You can then bring the tool down/table up to where they touch the table. See the differences, and adjust as needed.

Front-to-back (Y axis) tramming can get a bit confusing, as both indicators will move the same amount - just at different rates, until they finally do match.

Good work!

Kd0afk
02-15-2015, 03:04 PM
Zeroing each indicator on the same point by rotating 180 degrees worked. Thanks guys

LKeithR
02-15-2015, 03:30 PM
Personally I think one of those devices is a waste of time and money. If you read the directions for one of the professionally made ones they tell you to zero the unit simply by setting it on a "flat" surface. This works if the centrelines of the mounting shaft and the two indicators are all perfectly perpendicular to the base and parallel to each other--somehow or other yours aren't.

The easiest way to tram a mill table is to use a good back plunger DTI with a large (1/2" or greater) button point. Set it so the plunger just kisses the top of the table and it will glide over the slots no problem. This also allows you to swing the tool through a larger radius which will give you a more accurate tram...

Kd0afk
02-15-2015, 04:03 PM
The mill is trammed and I took some test cuts. I have no ridge now and I can see swirl marks going in all directions so I believe that might be a good sign

Kd0afk
02-15-2015, 04:04 PM
Thanks for all the help guys.

Toolguy
02-15-2015, 04:36 PM
Good job! You're welcome.

dp
02-15-2015, 04:39 PM
This works well if you have a small table in the Y dimension.

https://www.youtube.com/watch?v=OPMmGrdTqro

So does this:

https://www.youtube.com/watch?v=ZfioLDhBNBQ

Just don't drop the rotor like I did.

Kd0afk
02-15-2015, 04:41 PM
I've seen that video. Someone makes a tramming ring based on that principal.

ulav8r
02-15-2015, 07:31 PM
The tool is zeroed out by sitting it on a flat surface. I have a granite plate for that. The shaft indicated to be perpendicular to the base.

For that to work the bottom face has to be perfectly perpendicular to the shaft. The mill it was made on has to be perfectly in tram when it( the tool) was made or it would have to be trued up by some other method.

Kd0afk
02-15-2015, 07:47 PM
The tool was made on the lathe which I know is true.

Blogwitch
02-16-2015, 10:11 AM
Have a look at my post on here, and most of the way down the post it shows how it has to be done to ensure you have the spindle exactly vertical to the datum bottom face. It is the most important part of the build.

http://www.modelenginemaker.com/index.php?topic=416.0

Hope it helps

John

Lew Hartswick
02-16-2015, 10:15 AM
I made one of those two indicator tools but I still think the big disc and a single indicator is easier. It also give you a much longer "lever arm" in both axes.
...lew...

http://i233.photobucket.com/albums/ee238/LewHartswick/100_3521_zps1g6wkqya.jpg

http://i233.photobucket.com/albums/ee238/LewHartswick/100_3522_zpsvnbkalf1.jpg

J Harp
02-16-2015, 11:23 AM
Blogwitch,

Welcome to the forum. I have enjoyed your build articles on other forums, especially the swing up threading tool. I haven't made one yet , but it's on my to-do list.

dp
02-16-2015, 11:23 AM
There is only one simple requirement for precision and that is the three holes need to be perfectly parallel, in line, and the space between the center and each gauge mounting hole be identical. The bar itself imposes no requirement for precision and has no role in the device's accuracy. There is also no requirement that any of the surfaces on the bar be parallel or perpendicular to the arbor or table, and in fact it could be an ellipse or any other shape so long as the three holes are parallel and accurately spaced. The bar can even be curved - it just doesn't matter so long as the holes are parallel and properly spaced.

Alignment is simple and does not even require a properly trammed mill to accomplish. You set the depth of each gauge on a single fixed target to produce the exact same reading. There is no requirement the mill be trammed to do this calibration.

There are some considerations that need to be examined. If you have a knee mill then you cannot use your knee when drilling the holes unless you know for certain your head is already perfectly trammed using other methods for the simple reason you can't know that the spindle is not cocked presenting the drill at an angle to the path of the knee. If you use the spindle it won't matter if it is trammed or not. But in this case you cannot have the work mounted on a rotary table as the holes will not be guaranteed parallel. You must use your table to move the work left/right to bore the holes parallel.

Column mills come in a couple configurations - with a tilting head or a tilting column, or both, and with or without a moving quill. My mill is very simple and has a tilting column, no tilting head, and no Z motion in the quill. The entire head moves on the column. In my case the head is aligned to the column and so the drill path is that of the angle of the column. I can make this tool with this configuration even if the column is not perfectly trammed. The holes will be parallel to each other even of not perpendicular through the work (and that doesn't matter).

If your mill has z-axis in the quill then you can also make this tool by simply drilling the holes. No pre-tramming required. If your mill has a tilting head and the z axis is at the column and not your quill then your head and column need to be perfectly trammed so that your drill is not presented at an angle to the path of the head.

Blogwitch
02-16-2015, 06:34 PM
I would have to disagree with you on a few points there DP.
The tramming tool is originally set up with the dials on say a surface plate or mill table. If you did as you said, not exact quote, just drilled the holes and as long as they were parallel everything would be fine.
Wrong. The holes have to be perfectly square to the base, and that is its datum point, plus all three holes should be square to the base..
If you did as you say, when mounted into a collet in the mill after setting the clocks on the surface plate, and the main spindle wasn't square to the datum base, then the main bar would be sitting at an angle, and one dial would then be reading high or low, depending which one you were looking at.
If you care to read my post, it is all explained in there about how critical it is to get the centre main spindle square to the datum base.

Hi Jim, thanks for the welcome. I have come out of hibernation after many years looking after my late wife and getting myself almost fit again.


John

dp
02-16-2015, 10:38 PM
Wrong. The holes have to be perfectly square to the base, and that is its datum point, plus all three holes should be square to the base..

It isn't necessary to align to the base as a datum. You only need a single stationary target to align both indicators to. The target can be a V-block or a 1-2-3 block or a precision gauge block.

Insert the tool into a collet on the mill. Place a target block within reach of either of the indicator probes. Lock the quill's Z axis. Adjust the probe's depth in the bar to indicate 0.0. Without moving the block rotate the other indicator over the target and adjust it to 0.0. At this point the only thing that matters, the indicator, are adjusted such that when both indicators show the same reading, both probe tips are perfectly parallel to the plane of rotation of the quill. It is not important at this point that the quill and indicators' plane of rotation are not parallel to the surface of the table. With the tool in place bring the the two indicator probes in contact with the table directly or by way of two precision and identical targets. If the two readings are not identical they will indicate the degree of error in tram in the direction of orientation of the indicators.

It is never needed that the fixture's surfaces come in contact with the machine. It is only necessary to perform the single target test before each tramming operation to ensure the indicator dials are both aligned to 0.0 (or 1.0 or 2.0...). This doesn't mean one shouldn't build to high precision and align by bringing the data surface in contact with a surface - it just means there's more than one way to do it.

justanengineer
02-16-2015, 11:39 PM
JMO, but unless I'm seeing things wrong the tram setups shown arent likely to be even close, looks like theyre using standard 1", 0.001" division indicators.

Blogwitch
02-17-2015, 05:43 AM
I think you are getting my explanation wrong.
I am using a system which is foolproof and works around a datum, which if you notice with the commercial units, they do exactly the same thing, the datum is the bottom face of the tool, and once the gauges and holding spindle is set up to that datum, nothing on the tool requires to be touched ever again. No rechecking, resetting the gauges or any other voodoo charms that can be conjured up.
After the intitial setup (only needs to be done the once), just mount the tool spindle into the machine, and if the machine spindle runs true (your problem if it doesn't), touch the tips of the gauges onto the table so they are both reading, and adjust the tram until they both read the same. I can't think of any other easier way to do or explain it.
I am sure I would prefer to do it my way than any other ideas that have been mentioned.
Everything I do is KISS.

John

JCHannum
02-17-2015, 10:17 AM
For day to day tramming, I use a machinist's square against a ground test bar in a collet. On the rare occasion that more precision than that is required, I will use a DTI and a Pratt & Whitney toolmaker's flat.

I do agree that the double dial indicator setup does not require any more than the three holes be parallel, equally spaced, in line and zeroed at the same point on the table. Once set, it should retain its setting, but, as with any precision instrument, it is prudent to recheck zero with each use.

dp
02-17-2015, 02:16 PM
nothing on the tool requires to be touched ever again. No rechecking, resetting the gauges or any other voodoo charms that can be conjured up.

John

All you mentioned applies to what I have described. It is also a case of no resetting (except the dial ring and that applies to both designs) and surely involves no voodoo charms. Your concept and mine will always require rechecking, though, because we're all a bunch of clumsy bastids capable of accidentally bumping a gauge or dropping the fixture. The only difference between your datum method and my setup is you have more faith in the repeatability of your tool than I have in either of our tools. If I had one like yours I'd still check it before tramming. And I'd test is as I've described above. My test also verifies the arbor while yours ignores it. Since both fixtures are mechanically identical both are subject to the same opportunities for bumping a gauge or dinging the arbor.

Neither fixture design or construction has any advantage once they've both been calibrated and neither calibration method produces a more accurate or rugged instrument. And there is nothing wrong with precision grinding all the surfaces to within the flatness of an infinite circle if one wishes. It just isn't necessary and mortals can construct this tool with a drill press and a hunk of angle iron and get good results.

There is no life left in this discussion - we've beat it to death :)

Walter
02-17-2015, 05:32 PM
I surface ground a bearing race and set that on my table, actually vice and use a single test indicator. Ground race was checked for flatness after ground on a surface plate with a .0005 indicator. Its worked nicely for the past 4 years. I've given thought to a rig like the OP's but eh, what I have works.

Its not my fault, I didn't do it, I wasn't even there, and I wouldn't admit to it anyways.

dp
02-17-2015, 07:38 PM
I just recalled this topic came up a couple years ago here.

http://bbs.homeshopmachinist.net/threads/61769-EDGE-Pro-Tram-System

I drew this diagram of an esthetically unpleasing but functionally perfect tramming gauge set.

http://metalworkingathome.com/images/SkewedTramGage.jpg

dian
02-18-2015, 12:14 PM
this is what i tram with:

http://i973.photobucket.com/albums/ae218/romandian/Bild%20023_zpsrkkcoa1u.jpg (http://s973.photobucket.com/user/romandian/media/Bild%20023_zpsrkkcoa1u.jpg.html)

Kd0afk
02-19-2015, 03:11 PM
Now my question is; if my lathe made the datum faces perpendicular to the shank of the tool and was proved to be perpendicular with an indicator, why was it reading wrong when I calibrated it in the way this type of tool is supposed to be calibrated ie placing on a flat surface and zeroing out the dials? The holes for the indicator are perpendicular also.

Kd0afk
02-19-2015, 03:12 PM
this is what i tram with:

http://i973.photobucket.com/albums/ae218/romandian/Bild%20023_zpsrkkcoa1u.jpg (http://s973.photobucket.com/user/romandian/media/Bild%20023_zpsrkkcoa1u.jpg.html)
How is your tool used?

Rosco-P
02-19-2015, 03:19 PM
How is your tool used?

Before dian can tell you, you have to: sign a non-disclosure agreement; agree to destroy by smelting, crushing or granulating, your version of the "tram-be-damn".

dp
02-19-2015, 05:24 PM
Before dian can tell you, you have to: sign a non-disclosure agreement; agree to destroy by smelting, crushing or granulating, your version of the "tram-be-damn".

You chuck it up in the Bridgy and then deliver a ka-whack to it with a dead-blow hammer. Then you put a level on it and repeat until it's level. :)

dian
02-20-2015, 05:03 AM
interesting idea. i just chuck it up and put an indicator on it. hammer stays on the wall.