Tramming my PM-833TV

[Tipton1965]

All of my experience in the past was on full size Bridgeport mills. I'm discovering that these column bed type mills are a different animal as far as tramming goes. I've done some research but would like to hear from others the techniques they use. Here is what my gut and a little research tells me to do. I see so many videos where people simply tram the head to the table in the X and Y and call it a day. It would seem like you would want to start with tramming the actual column or carriage on the column to the table first using a granite square? Using shims for both the X and Y axis where the column meets the base. Then tram the head to the table using the tilt feature of the head for the X and using shims where the head meets the column carriage for the Y? You end up with more shims but this seems like it would be the best practice. Am I thinking correctly for this tramming task?

 

[Pcmaker]

I just tram the head to the table

 

[Aukai]

Where/what are you referencing the square to, to tram Z? If you square X with the head swivel, and square Y by shimming the nod, then Z can only be compensated by shimming at the dove tail bolting interface if it's out in vertical travel? Interesting, I have not given much thought to tramming the Z through its vertical travel.

 

[Tipton1965]

Where/what are you referencing the square to, to tram Z? If you square X with the head swivel, and square Y by shimming the nod, then Z can only be compensated by shimming at the dove tail bolting interface if it's out in vertical travel? Interesting, I have not given much thought to tramming the Z through its vertical travel.

Yes, using the square to tram the Z (column) to the table in both the X and Y axis. If this isn't done you can lose your 0 reference even if the head is trammed to the table. Imagine working on a multi height piece. If the Z axis is not in tram with the table you will lose your 0 when you raise the head. If you just tram the head to the table I would think the head would only be accurate at the height you trammed it unless by chance the column is in perfect tram with the table from the factory.

 

[Aukai]

I can understand the vertical travel tramming, I'm not getting where, how the plate is referenced to be able to tram Z, unless your tramming X, and Y first for the plate to be level/square? It would seem that there is going to be some tail chasing going on?

 

[Tipton1965]

I can understand the vertical travel tramming, I'm not getting where, how the plate is referenced to be able to tram Z, unless your tramming X, and Y first for the plate to be level/square? It would seem that there is going to be some tail chasing going on?

I'm talking about tramming the column (Z) with the table. You would tram the head afterward. The head could be turned at an angle and you could still tram the Z axis (column) to the table using the granite square. Then come back and tram the head to the table.

Take a peek at this video.

 

[davidpbest]

I have trammed the column and head on about 10 square column mills similar to the PM-833. Two of them were my own machines, the others were owned by friends who followed me down the path of buying and fitting up a benchtop mill for precision work. This is the process I follow.

Make sure the dovetail gibs on the head-to-column and both X and Y directions on the table are properly fitted. Pull out the gibs, stone them and check for flatness, then reinstall and adjust the gibs so the surfaces move but exhibit minimal slop when force is applied. If you really want to go all out here, you would clean the dovetail and gibs, blue up the gibs and check the fit that way. Depends on the level of precision you are after. But the point is to start by making sure the dovetail surfaces and gibs are properly fitted. You don't want to start screwing with the column alignment, then discover the head-to-column connection is sloppy because of loose fitting gibs. Examples of using indicators to check gib fit are here.

The next step is to tram the column. This can be checked with a granite or machinist square on the XY table, and indicating from the head to the vertical square face and moving the head up/down. Adjustment is done by breaking the column to base connection and shimming. Be sure to check column alignment in both X and Y axis directions and shim accordingly. And consistently lock or leave loose the head-to-column connection. I prefer to move the head, lock it with the provided levers, check the indicator, move/lock/check again. A quality head-to-column dovetail connection will alter head nod by 0.001-0.002" at the spindle end when the locking levers are secured – no more than that.

For shimming the column, you will need access to some really thin shim stock to do it well. I use cigarette papers and aluminum foil for the final tweaking – 0.0003" level of adjustments. Some people will set the column to base connection in epoxy to take up the gaps, but that's a pretty tedious job. Example of that can be seen here:

The next step is to check that the quill travel is coplanar to the head travel on the column. Assuming the column is trammed to the XY table, put an indicator in the spindle and indicate again to the square on the XY table and move the quill up/down. If you are lucky, the quill travel will remain consistent in the Y-axis. If not, then some kind of adjustment will need to be done where the head attaches to the swivel base at the column. Only the "nod" is important here since the "tilt" can be adjusted with the swivel mechanism. In two situations, I had to remove the head, scrape the swivel interface to get the quill travel trammed to the head travel on the column.

Final step is a conventional tramming operation in the X-axis using an indicator in the spindle against the XY table and adjusting the head tilt at the head-to-column swivel. On most of the benchtop mills I have set up, I added adjustment tramming bars to both sides of the head for fine adjustment with screws. Example of that can be seen here:

Full video on how to make these is here.

Hope this helps.

 

[Tipton1965]

Excellent Mr. Best. Looks like you and I were on the same page minus checking the quill travel......didn't think of that.

 

[Tipton1965]

Examples of using indicators to check gib fit are here.

I watched your videos and had a question. When checking the saddle to the X/Y table you push and pull vertically. When checking the saddle to the knee you push and pull horizontally. Why wouldn't you push and pull horizontally in both setups? Or push and pull vertically in both setups?

Also what type of stone do you recommend for lightly stoning the gibs? I have knife sharpening stones but as you would guess they certainly aren't flat from years of use.

 

[davidpbest]

I actually checked the X-axis in two ways, but filmed only one. The biggest challenge with the baby-bridgeports which have long tables and relatively narrow saddles is "end sag or droop" when the table is cranked out in a cantilevered manner. That was the primary issue I was trying to correct, in fitting and adjusting the gib in the X-direction and why I concentrated on that in my video. When adjusting a gib, it's usually assumed that the gib is properly fitted to the dovetail and making consistent contact along the dovetail surfaces when dialing-in the gib screw for proper table tracking and movement. Having helped a few others with their benchtop mill gib adjustments, I have come to know that the gibs are often poorly fitted in the first place.

My PM935 exhibited far more table sag up/down than it did lateral table twist. I was having a terrible time trying to attain vertical tolerances on long part machining (20"). Even when the gib was fully wedged and the locking levers cranked down hard, I was getting teeter-totter movements that I felt were excessive. In my tests, the X-axis gib was found to be contributing to that issue far more than the Y-axis gib. By bluing up the gibs and looking at the interface in the dovetails, I was able to determine the X-axis gib taper and flatness was many thousandths off from one end to the other, and the gib was also slightly twisted. It was obviously never properly fitted to the dovetail in a precise manner. I was able to get that sorted by first properly fitting the X-axis gib (filing and stoning and checking flatness) then turning to the Y-axis gib (which turned out to be a minor contributor). I spent most of a day doing all this, an in my moment of joy getting this sorted, neglected to film and discuss the X-axis vertical check.

But assuming the gibs have been properly fitted, the recommended check in the X-axis is to indicate from the saddle to the side of the XY table and apply lateral force as you suggest. Same is true for the Y-axis check. I ended up doing both vertical and horizontal checks, but the gib fitting culprit on my mill was the most vexing, and pulling up/down on the XY table was the best evidence check. Once I got that aspect resolved, the lateral movement check showed no further issues.

Here is a link to one of the very few gib adjustment checks I have seen documented that attempts to make the process empirically based, rather than just based on "feel." Check out page 6 in that document, and the diagrams on page 5 that illustrate the indicator setups. Most typically, the gib adjustment process is as described in the PM-935 manual:

"Aim for the best compromise between firmness and reasonably free table movement. Too tight means accelerated wear on the ways, leadscrews and feed motors, if installed. Too free means workpiece instability, inaccuracies and chatter."​

I prefer the Wells-Index method.

Here is another video on the topic of ADJUSTING gibs using the Wells-Index method - again, assuming the gibs are fitted properly to begin with.

As for fitting a gib, in my case, the gib needed a lot of work - far more than I have seen needed on another other machine I've worked on. Most of the fitting was done with smooth files, and the stoning work was mostly for final flattening and burr removal. Keith Rucker goes through the process of making and fitting a new gib in this series, episodes 19-21. I don't have access to a surface grinder, so I did as best I could with files and stones. It's not perfect but it's a lot better than what I started with and sufficient for my needs.

As for stones, most of mine are Norton benchstones, with two grits. I have three identical stones, one is a beater, the other two I work hard to keep flat. I also have a pair of precision ground stones that are the same Norton units ground to tight tolerance, which is what I use to stone the table top, vise base, etc to remove any dings or irregularities. The precision ground versions are available from Lance Baltzley here. That rectangular stone on the far right is what I have. More on the precision stones is here.

Hope this helps.