Looking for advice on tramming a small mill

Note for mods: I read Machinery Discussion Guidelines and noticed that y'all want to discourage threads about Home Shop Grade import machines. Unfortunately, my thread is about a Home Shop Grade import machine . That being said, I think my thread is more about technique and less specific to the machine I have. At the end of the day, if this post is deemed inappropriate for this forum then please remove and I'll understand. Thanks!


So, anyway, hello! I'll go ahead and say right up front that I'm an absolute beginner and rank amateur. I'm operating entirely under the assumption that most of my problems are a result of user error and/or ignorance, haha. I purchased a small, benchtop mill from Little Machine Shop (specifically this one) and I've been having a blast learning how to use it. My problems started when I noticed that my 3" face mill was double cutting (I'm not sure if that's the correct term but it was cutting on the advancing side and then again on the trailing side). A little bit of research indicated that my mill was probably out of tram so I decided to take care of that. Now I'm way, way down the rabbit hole and want to make sure I'm not just spinning my wheels.


With that out of the way, here's some info about my setup and what I'm trying to do so I can (finally) ask my questions.



[The Setup]
I've read that the first step to doing a full tram is making sure you have parallelism between the head and the column. For reference here, I'm attempting to use Rollie's Dad's Method which I've seen all over the entire internet.

I'm measuring this by attaching a test indicator (with half 0.0001" graduations) to the column and measuring the runout (on the X axis side) of the straightest 1/2" rod I could find held in a collet. With the spindle, X, Y, and Z carriages locked, I'll measure the runout very close to the collet and center the indicator's dial so the that the middle of the sweep is 0 - by that I mean if the total runout of the shaft is 4 thou then the indicator will sweep from -2 to +2. Once I've recorded that, I'll unlock the Z axis, move the head up as much as I can - due to the length of the shaft in my setup, that's about 4 inches vertically - re-lock Z and measure the runout again. Obviously, the total runout further away from the collet will be greater. However, if the head and column are parallel, the sweep of the indicator should still be centered on zero. If the sweep is not centered then I'd need to adjust and/or shim the head in whichever direction it was tilted to bring it parallel. Once that's done, I'd move the indicator to the front (Y axis) of the shaft and repeat the process.

Finally, I have my indicator mounted thusly:



[My questions]
Since I realize I wrote something of a novel here, I've bolded my actual questions to make them a little easier to find.


Section 1: Is this even what I'm supposed to be doing?!
Visual Aid: Here's a video showing varying amounts of deflection based on how hard I lock the Z axis:

1. Obviously, I want to maximize rigidity when I'm taking a measurement - that means locking down everything that can be locked (right? right?!? ). However, I'm getting between 0.001-0.004" of deflection when I lock the Z axis depending upon how hard I crank down on the lock handle. I'm guessing that some slight deflection is unavoidable but is this a normal amount of deflection to expect?

2. How can I ensure that I'm applying consistent locking pressure? If, when locking the Z axis, I do it lightly and only cause 0.001" of deflection for the first measurement but then, after moving the head up to take the second measurement, I lock it more tightly and cause 0.003" of deflection then my measurements will appear to be off. In reality, however, the measurements could indicate parallelism and the actual issue was how tightly I locked the Z axis. Am I missing something or overthinking it?



Section 2: This seems wrong but I don't know enough about it to say why ?!

3. I'm noticing that the range swept by the indicator during a single spindle rotation is not always the same and that seems very strange to me. Specifically, up near the collet, I have about 0.001" of total runout, which seems super good to me! However, once I zero the dial in the middle of the sweep, the indicator does not always show +0.5 to -0.5. It seems to cycle between +0.5/-0.5 and +0.25/-1.25 almost like there's a wobble or eccentricity somewhere.

Stranger still, the wobble seems to happen every other spindle rotation. I've tried 3 different shafts in the collet as well as about every combination of lock/unlocking the Z axis, the spindle lock, and the handle lock- no matter what I do, the wobble persists. What could be causing this inconsistent reading?

Here's a video showing the wobble. I drew a black dot on the spindle to help stay oriented and I'll break down the readings for each rotation of the spindle.

These measurements were taken with the drive belt disconnected to ensure the motor wasn't somehow influencing the readings.

Rot 1 - Indicator sweeps up to +2, down to -2, and back to 0. My assumption here is that this would continue to happen every single rotation.

Rot 2 - Needle is noticeably shy of +2 (maybe +1.8?), goes down to -2.5, returns to 0.

Rot 3 - +1.8, down to -2, back to 0.

Rot 4 - +2, -2.5, 0

Rot 5 - +1.5, -2, 0

Rot 6 - +2, -2, 0

Rot 7 - +1.5, -2.5, 0

Rot 8 - +2, -2, 0


At first I thought I was somehow influencing it by how I was turning the spindle (from the top, by hand, with the drawbar wrench) but I've tried several different methods (including other people) and the pattern is always the same with the wobble.



Just in case it helps, here's a full video of the measurements both near the collet then with the head cranked up a few inches.




[Wrap Up]
Overall, I'd appreciate any sort of confirmation that I'm on the right path, guidance as to what I should be doing differently, and advice on how to proceed.

It occurs to me that I'm likely overcomplicating or overanalyzing things but I don't really have anyone I can ask in person so here I am.

For anyone who actually managed to read all of that, please have mercy and thanks in advance!

Quite a novel you wrote - I did not read it fully. I'm a self taught amateur with 15 yrs more experience than you, but since no one else has responded, I figured I would. Maybe someone else will come along and correct us both. My response only deals with my tramming set up, not why you are seeing such variation in readings. The images were found on-line to illustrate my words - not my exact setup.

I have no experience with how you are trying to tram your mill. Your set up appears to be overly sensitive to me. I have only read of Rollie's Dad's method for use on a lathe to check for machining an unintended taper (checking the twist one's lathe bed). I have never read/seen the method you are showing in images used to tram a mill (that could be simply my limited experience).

I used to use an Indicol mounted to the spindle of my mill. https://external-content.duckduckgo...id=OIP.nLpqdLpZEtG41DCLXMP8mgAAAA&pid=Api&f=1 I recently added a spindle speed sensor and can no longer attach and rotate anything mounted to the spindle so now I insert the shaft of the Indicol in a collet. I use an 8 inch dia ground disc on my mill table to avoid the need to raise the indicator up as I sweep across the mill table slots (because I already had the disc on hand). Others use a large dia bearing race.
Tramming a BP head Many hobbyist use a new brake caliper. https://external-content.duckduckgo...id=OIP.IMwteD83G1S-C4ctym2b2wHaEK&pid=Api&f=1

I still want to buy one just to see how bad it is and what I could actually do with it...think 17-4 or inconel 718 lol

That is about as bass akwards a way to tram a mill as I have ever seen! WTF are you doing? checking collet run out? With a piece of round stock, not even a ground pin, size pin or the like? Do you know what tram is???? A mill is tramed when the quill is square to the table in the X and Y axis. Most toy mills are not even adjustable for tram and they flex so bad on even the lightest cuts that it does not matter if they are tramed or not. A guy I knew that had one bolted the column top and bottom to a big column that was holding the building up. Still a flexy flier. Get a real mill not a MSO.

moonlight machine said:

That is about as bass akwards a way to tram a mill as I have ever seen! WTF are you doing? checking collet run out? With a piece of round stock, not even a ground pin, size pin or the like? Do you know what tram is???? A mill is tramed when the quill is square to the table in the X and Y axis. Most toy mills are not even adjustable for tram and they flex so bad on even the lightest cuts that it does not matter if they are tramed or not. A guy I knew that had one bolted the column top and bottom to a big column that was holding the building up. Still a flexy flier. Get a real mill not a MSO.

Click to expand...

Mso...machine superior to operator?

I am not sure what you are doing but that is not the way to tram a mill. All you are measuring is the run out of the rod which is probably not even straight.

To tram you put the indicator in spindle and measure to the surface of the table in four places 90 degrees from each other.

Years ago I had a CHINA made mill and the only way to tram it was to remove the column from the base and shim it. It was a big pain in the ass and took hours to do.

Go to Youtube and look at videos of tramming.

He is measuring alignment of the spindle axis to the z-axis. It’s not an invalid test but without knowing the z-axis is perpendicular to the table, it doesn’t address the performance of a face mill or fly cutter. Seems better to directly measure that in the convention way, by indicating from an offset position rotating with the spindle, to the table.

plastikdreams said:

Mso...machine superior to operator?

Click to expand...

I think that was a joke, but if not, Mill Shaped Object. *.S.O. is a commonly used acronym for saying something is a piece of ... I mean, that it's subpar.

plastikdreams said:

Mso...machine superior to operator?

Click to expand...

Mill shaped object!

moonlight machine said:

Mill shaped object!

Click to expand...

Sorry I looked it up on the Google and like 90 someodd choices came up, but not that one lol...I picked the closest

plastikdreams said:

I still want to buy one just to see how bad it is and what I could actually do with it...think 17-4 or inconel 718 lol

Click to expand...

the only thing that will impress you will be the amount of corners cut when trying to make that machine to a 300-500$ price tag that they could be bought in China, same machine in either US or EU will be around 1500-2000$ mark

TLDR. Looking at the photo it's nuts. Unplug the mill. You mount an indicator on a rod that's offset from the collet a couple inches or more, so it can measure downwards to the table/vise. By hand, rotate the spindle to the left and zero the indicator against the table. Rotate the indicator to the right and note the reading. It should be zero. (lift the indicator point so it doesn't drop into the table slots and get broken off.) Adjust the tram until the reading is equal on both side. No idea how you do that on your machine. Shims? You should also check the same way to the front and rear. Job done.

Ummmm, Am I missing something?

The OP said, in the first paragraph, that a three inch face mill was cutting on both the leading and trailing edge. I didn't read the novel, because that is usually how I know that my mil is IN tram.

fciron said:

Ummmm, Am I missing something?

The OP said, in the first paragraph, that a three inch face mill was cutting on both the leading and trailing edge. I didn't read the novel, because that is usually how I know that my mil is IN tram.

Click to expand...

that is more or less when the machine is rigid, in OPs case all bets are off, with a 3" face mill it most likely is flexing in all possible axis, so even if trimmed to 0 and everything is indeed parallel and perpendicular where necessary, the lack of the rigidity in the machine won't allow him to make a flat surface anyway

actually in might be better for him to intentionally tram the head slightly to one side, then the load of the cut may straighten it and he'll get a better result, but that will only be true under certain load and in one cutting direction, and everything else done on the machine will come out slightly off

there is a reason why threads related to problems caused by quality of that grade of machines aren't worth the storage bytes on PM servers...

on my universal Mikron weighing in 4 tonnes I simply stick noga base to something in the spindle, move the arm away so the indicator travels in aprox 200mm circle and slide a gauge block under the indicator so as it doesn't catch the edges of T slots in the table, tap and tighten, whole procedure takes maybe 3-4min to do and most certainly isn't worth writing an essay about... would take probably 2nd paragraph to describe it if the head had the "nod" adjustment like Bridgeports have, but I don't think OPs machine has the "nod" adjustment either

For one thing, did anyone else notice that the indicator appears to be positioned so that the rod is contacting halfway down the indicator stylus rather than at the ball contact? Unless it's a perspective issue. But anyway, that ain't no good...

Next, It has already been said, but I'll say it again. Tram is most easily and quickly checked to a good level by sweeping an indicator across the table as described by multiple guys here. If your machine isn't square, you'll drill, bore and ream holes that aren't square to the surfaces you mill. If you have a worn machine or a cheap one, indicate the surface you are going to tram using table travel first, then set the head to those numbers if they aren't zero.

Taking a heavy cut on a small machine will often flex the machine enough that it might cut on the trailing side, that's normal. A machine trammed at zero might still leave very light drag marks on the trailing side, so some guys like to leave just a tiny bit of "out-of-tram" on purpose and finish in only the one direction. Usually even if you're set pretty close to zero you can do that, since often nobody spends the time to get this dead perfect.

The problem with many MSOs like that is that there are two Z trams in the x plane, one where the mill head rotates on the mounting to the vertical carriage and one on the base of the mill column. You have isolate the tramming work to ensure that you are only adjusting one tram at a time.

The first tram has to be the spindle relative to the column, not the table. Here's what I do:

Extend the quill and lock it. Mount an indicator on the column and zero it near the bottom of the quill. Then run the Z axis head elevation down until the indicator is near the top of the spindle. That measures the tram of the spindle relative to the column. Fix that, and then move on to tramming the column relative to the table in x and y. Unfortunately there is no simple way to fix the tram of the spindle relative to the column in the y plane.

When tramming the column to the table, replace the supplied column bolts with new bolts from a reputable manufacturer so you don't end up with a broken bolt. Use a torque wrench to tighten the bolts to the recommended torque for the bolts you are using, and use that value every time. The tram in the y axis can be adjusted with shims, but do not measure unless the bolts are tightened fully. Once the shims are determined for the y axis tram, move on to the tram in the x plane.

fw3 said:

...

Stranger still, the wobble seems to happen every other spindle rotation. I've tried 3 different shafts in the collet as well as about every combination of lock/unlocking the Z axis, the spindle lock, and the handle lock- no matter what I do, the wobble persists. What could be causing this inconsistent reading?

!

Click to expand...

Big kudos for seeing this.
Depends on bearing size but there is cycle.
Even on high precision machines people often do one turn and do check more.
The balls and races cycle and this is not in one rev.
Sometimes my workheads need one or two full revs before the next part to get back to home for the bearings.
Which is best to tram too, spindle axis to table top or slide travel?
Bob

QT Op:Quote Originally Posted by fw3 View Post
...

[Stranger still, the wobble seems to happen every other spindle rotation. I've tried 3 different shafts in the collet as well as about every combination of lock/unlocking the Z axis, the spindle lock, and the handle lock- no matter what I do, the wobble persists. What could be causing this inconsistent reading?]


Re: What could be causing this inconsistent reading?]

If the shop window is facing south in the Northern Hemisphere and it is a cloudy day heat can vary usually in an up and down direction. Heat makes things swell often from larger to smaller, or from smaller to larger and even from longer to shorter. . using a long test bar and a long indicator arm can even increase the outcome of a test to the point of confusion. Because if this it is best to tram a Mill on Wednesday.

Plus Z axes cut a lot better if you take time to sharpen them.

Ref: Rag Trick: Great technique to sharpen/reprofile a convex edge! Every ax user should know this Trick - YouTube

Oh, and if looking for a few tents, I just heard there are a lot of tents at Venus beach California nowadays.

Tent city Kalifornia.
Los Angeles, COVID, and a homeless beach tent city

michiganbuck said:

QT Op:Quote Originally Posted by fw3 View Post
...

[Stranger still, the wobble seems to happen every other spindle rotation. I've tried 3 different shafts in the collet as well as about every combination of lock/unlocking the Z axis, the spindle lock, and the handle lock- no matter what I do, the wobble persists. What could be causing this inconsistent reading?]


Re: What could be causing this inconsistent reading?]

If the shop window is facing south in the Northern Hemisphere and it is a cloudy day heat can vary usually in an up and down direction. Heat makes things swell often from larger to smaller, or from smaller to larger and even from longer to shorter. . using a long test bar and a long indicator arm can even increase the outcome of a test to the point of confusion. Because if this it is best to tram a Mill on Wednesday.

Plus Z axes cut a lot better if you take time to sharpen them.

Ref: Rag Trick: Great technique to sharpen/reprofile a convex edge! Every ax user should know this Trick - YouTube

Oh, and if looking for a few tents, I just heard there are a lot of tents at Venus beach California nowadays.

Click to expand...

Everybody likes some ass, most of the time nobody likes a smart ass. ROFLOL!!!!