[Shiny]

You will be pumping out beauty parts soon and having fun doing it.

You'll install a rolling mill and polymer extruder in your neighbors' garage any day now.

Programming using parameters sounds familiar but I never used a microcontroller. Seems we used "integrated" controllers with PC interface when "cost was no object" and were basically sending commands with parameters using a simple programming protocol. The controllers had all the smarts to interpret and the manufacturer provided the "program" environment and utilities to edit on the PC and send to the controller.

IMO you won't need encoders unless you start moving too much mass too fast to control. Do your drivers and software support "micro-stepping"? I remember that being pretty slick.

[dataway]

Yep those machines you used, with that display screen, they have both the g-code editor and the microcontroller (or probably a PC back then) made into the unit. I use a laptop as the g-code editor and "sender" ... which sends to the microcontroller (tiny CPU) and that little CPU has firmware I loaded that can convert the g-code into the signals needed by the driver boxes. The driver boxes are pretty incredible for the price, 10 or 5 volt input (for input from the microcontroller), configurable current limits, micro-stepping up to crazy stuff like 2000 steps per rev. I'm using 400 steps, which gives me a resolution of .0004" at the table. As you increase micro steps you decrease torque, and holding torque, so I'm keeping it conservative. Up to about 800 it doesn't hurt much.

[Shiny]

The last time I delt with it (~2018) we were using Applied Motion motors and drives for test automation. Out of box solutions with support.

[dataway]

I'd be lost in an actual, real industrial environment when it comes to CNC stuff. Big money, big HP, lots of damage can be done

[Shiny]

Oh yes, hobby vs making money!

There's a fishing reel manufacturing company here in our small town. We toured the factory and it was filled with relatively small CNC machines but impressive. Far from high-tech or high volume but fun to see. Bars of aluminum in one end, machined fly reels out the other.

When you are pumping revenue through machines, the QA better keep you from pumping garbage and the machines better not break.

[dataway]

Got most of the welding done on the column ... will probably add some webbing to the back side after it cools and I check the squareness.

Had to undercut the bead around the outside of the base to make sure the bead didn't interfere with the bolt holes.

MIG welded the inside of the column rather than TIG as none of the beads will be in the way of anything going forward.

Going to have to try to rig it up in the Bridgeport to square the column base with the column face. It's showing about 0.020" out in the 16" length. Going to have to try to fixture it on the front facing edge of the BP table with the bottom of the column facing up ... I've seen it done, but I've never done it myself. Has to take forever to get it fixtured square.

Might be a better idea to just shim it square .... I'm not the best at squaring things on the mill if the past is any indication

[Pav8427]

I suggest you tickle cut the rail mout surface o few thou.
Gives you a good flat surface to work off of.
Being that this is a home shop set up, would gues your collection of long end mills is weak.
One option to square up the mout face would be to trim the rail mount surface and side cut the base in the same set up.
Might be able to borrow a long end mill from a local shop.
Or atleast have them do a quick trim of what you have.

[dataway]

I've got two issues right now ... the rail mounting surfaces are about .005" out in depth compared to each other. ( I used your previous suggestion and cut them down a bit to leave an alignment ridge) Meaning ... I put an indicator on one carriage and slide it down indicating off the top of the other rail and I show a total of .005".

I'm good on parallel, about .001". So your idea worked nice there.

The base is 4" x 6" so I'd need a pretty long mill to cut the whole thing ... not to mention that might be beyond the ability of my clapped out BP to maintain any tolerance with that kind of stick out. Most of my mills are surplus Boeing stuff I got at an auction years ago .. hundreds, most brand new ... but lots of them for aluminum of course. Longest with a 3/4 shank around 2.5" of flute.

I've been using a high quality 3" face mill to do most of the work surfacing pieces ... works great in the Y axis ... but I need to work on the BP tram in the X.

My "plan" at the moment was get the rail surfaces cut as flat as I can, bolt on the rails and clamp the whole thing to the front edge of the table with the bottom of the column mount facing up, using the rails to provide the reference, then I could indicate off the spindle to get the bottom of the mount square with the spindle ..... which "should" be square with the table. Going to have to check my knee gibs .. I think I've got a bit of sag in the knee.

I can slide the ram out, and swing it over the reach I'm pretty sure.

And yes it did occur to me to take it to the local machine shop and have it squared up, but I know the guy, and he's not the sharpest knife in the drawer ... likes to charge stupid money for stuff like this too.

[Pav8427]

That should work. When you go back in and recut the linear mout surgaces, I would still take a tickle cut to the column so if any other work has to be performed you know that it is all square and on the same plane. Then you could set up w/o bolting rails on. When you get everything hanging off the table, pull off the part in Y and sweep the table top in with indicator. That should square up the head and therefor make the column square. Should take care of any 'slop' in your BP. Just remember to sweep it back in once the head is swung back to normal position.
May want to even qualify the backside of the column before you remachine the rail mount surfaces.
Then you arent pulling any warp out when clamping and then unclamping only to have it come back.

[dataway]

OK ... I get it. That will give me a reference I know is square to the rail mount surface. Very good .... obviously I'm not a machinist

[dataway]

PAV ... here is what I'm looking at now.

After spending some time tramming and adjusting the BP I've got the areas seen in green flat and parallel to each other.

Next I plan on putting the column face up, supported by parallels laid flat under the green notches. That gets the column face parallel to the table and facing up. Then in that position I can remachine the rail mounting surfaces flat and true to each other in the Z axis.

Problem now is that in order to do that I need to get the column aligned perfectly in the X axis so the step that I use to align the rails are parallel. I can't use the sides of the column as they are nowhere close to true in any axis ... if they were I'd just use some 1-2-3 blocks and align with the edge of the table.

I guess I could indicate off one of the steps, since the steps that are there are parallel to each other in the X, but not "flat" in the Z. But if I indicate off the vertical edge of the step the Z error shouldn't matter .... ?

I know I'm using the word "flat" wrong ... the green areas are parallel to each other, and the table.

Hope this makes sense

[dataway]

Jeez ... just occurred to me ... I can bolt on one of the rails and indicate X off that, then take it back off and machine the depth to get the two rail mounting surfaces aligned in the Z. ( goal is to not screw up the X parallelism of the steps but to get them correct in the Z)

[Pav8427]

Thats what I would do. One thing you can do to get the depth of the rail pckets equal is set up a dial indicator someplace on the BP head in vertical postition.
Set it up so when you get done re cutting the first one that it is 'zeroed'
Before you move anything.
Then instead of bringing the quill up, drop the table.
Move to your second pocket and start the mill and crank table back up so indicator is zeroed again.
As long as you have a center cutting end mill.
As long as you use the same surface you used to set up the indicator, it should make the pockets on the same plane.
Kind of an old skool CNC.
Make sure to sneak up on zero while clamping table because these will change when locked.

[Pav8427]

The pockets wont need to be cut to the exact width as before due to the chamfer that is on the rails.
A few though narrow will kepp you from possibly cutting them over width and loosing a nice snug fit.

[dataway]

We have a plan! .. I've got a nice 3/4" carbide end mill I can use.

I can get to the second pocket without dropping the table or lifting the spindle ... just enough travel to sneak around the end of the column. Past experience with this machine tells me that will be within .001" in the 4" width of this column ... maybe better since I snugged up the knee gib. Y axis has the typical .001" dip in the middle of a well used machine.

[dataway]

Column is done. Turned out very nice. The rails are true height wise to about .001" in 16" ... don't think you can get much better with these Chinese rails. They are dead on parallel .. because you can tap them that way and then tighten down.

Column base is about .0005" square with the surface of the rails. I used that "tickle" strip you suggested to true up the column upside down, mounted on a 90 degree block, then used a face mill on it. Took a picture of the mill in that position because I doubt it's ever been like that in it's 65 year life. Certainly didn't feel like it had when I was cranking the ram out that far. Also rotated about 30 degrees to the right.

Pictures showing the rail height at top and bottom, and the CNC mill in it's present level of completion.

[Pav8427]

Nice job. Actually just tore my set-up down like that.
We mount a vice on it side for putting holes in the end of longer stuff.
Used to be you used the most precision machines to build a more precision machine.
Not so much in this case.

[dataway]

Yeah, I'd be real happy if the little CNC comes close to the accuracy of the BP .. which isn't that great. The thing with the little machine is it can be shimmed into pretty good accuracy whereas the BP .. it would be an extremely expensive endeavor to increase its accuracy.

Of course as you shrink a machine down ... technically you should also be shrinking down the error for it to be valuable as a small machine. Not sure that will be the case here either.

Will be spending a lot of quality time with the mill on a surface plate the next couple of days to find out exactly where any error is.

[dataway]

Another week or so and I'll be testing on a work piece.

Here are the mechanicals basically all done. Just need to make up some brackets for the limit switches and come up with some way covers, and route the water cooling lines.

[dmorg1]

Very nice and I am sure a great learning experience thanks for sharing.


Possibly your next project will be adding ball-screws and a control to that knee mill of yours?


Dave