Mill Power Feed Weight Question

[whydontu]

Question regarding adding a power feed to a smallish bench mill. I have a BB CT129 (same as a Grizzly G0704). I measured the operating torque under a reasonable cutting load, and I need about 7Nm / 60 in-lb worst case.

I have a nice new NEMA 34 stepper motor, puts out 12Nm / 100 in-lb locked rotor, so lots of safety factor.

My concern is weight - the motor weighs about 5 kg. Should I be concerned about hanging this much weight off the end of the table? I know that if I run the table to the end of stroke and then really lean on the table, my dial indicator will show maybe 0.003” deflection.

Or just add some weight to the opposite end of the table to compensate for the motor weight?

 

[Darren]

Is the dial indicator mounted to the column? or elsewhere on the bench? In other words, is the table itself deflecting the entire 0.003"? If so, I wouldn't be adding any weight, and would want to reduce the weight on the ends. I'd also want to lock the x gib completely to make sure i'm seeing the results correctly.

On Bridgeport type mills, table sag is a real issue, to the point that you don't want to leave heavy accessories on the table hanging over the edge. The weight of the table alone is enough to bend the table over decades of time. On a lighter mill, i'd be more concerned for sure.

 

[whydontu]

I moved the table close to the column, and moved the table all the way to the right. The mag indicator holder was set on the base #76, with the arm arranged to put the indicator stem as close to #56. Then I put as much weight as I could on #56 and read the deflection.

 

[Darren]

locking up the gib for testing seems like the next logical step. A small bit of play at the gib could certainly account for the 3 thou as you saw on the indicator. At any rate, 11lbs on the end of that small of a table seems like a lot. Perhaps an anti gravity device could be employed if actual table deflection is found.

Edit: maybe lock up both x and y gibs to be certain.

 

[jcdammeyer]

That's a really good question and an interesting problem.

This may be the sort of situation where using a size 34 stepper is the wrong approach creating more difficulties.

AC servo's tend to be much lighter and have higher RPM ratings. Imagine mounting the motor on the front of table and using toothed belt pulleys to bring the speed of the motor down and the torque up to the required level. The other advantage of the servo is that it's locked torque is way higher than constant torque so where a stepper motor might lock up and stall the servo just draws some extra current for a short time and overcomes the load.

Now most of the weight, which is less than the size 34 stepper, is closer to the center and you get that torque advantages of a servo.

 

[slow-poke]

Another thing to consider is the torque of that stepper at the RPM that you would be happy with. On my old mill the steppers were more than up to the task of moving the tables at slow speeds < 20IPM, however could not be trusted at higher speeds. Micro-stepping lowers the torque even more. The ac servos I'm using on my new mill are about the same size as the old steppers however can effortlessly move the larger table on the new mill at say 200IPM, which is scary fast. The torque vs. RPM curves of steppers vs. DC servos vs. AC servos are vastly different. I see StepperOnline have new low cost ac servos (they say pre-order) 750W for about US$200 for motor including drive. Their tech support sucks but if you can get past that, ac servos are becoming really affordable for all the other benefits they offer for example a 17 bit encoder.

I just looked at a typical 12Nm stepper and at 650RPM it can only provide 2Nm of torque. While the ac servo can deliver > 7Nm @3000 RPM and the AC servo is smaller and about 2 lbs lighter.

 

[mbond]

Probably you aren't going to be cutting at the very end of the table where the sagging is happening. So that indicates that the main concern should be the angle of rotation that this kind of sagging causes. Assuming that the unsupported length of your bed is 10" (probably it is longer) a displacement of 3 thou implies a change of about 1.3x10^-10 at the center of the table.

 

[whydontu]

Probably you aren't going to be cutting at the very end of the table where the sagging is happening. So that indicates that the main concern should be the angle of rotation that this kind of sagging causes. Assuming that the unsupported length of your bed is 10" (probably it is longer) a displacement of 3 thou implies a change of about 1.3x10^-10 at the center of the table.

That’s sort of what I’m thinking. The cutter is always centred over the table so trig says the actual deflection related to the cutter must be pretty small. Probably time to sacrifice some aluminum stock and cut some long bars with and without the extra offset weight and see what happens.

 

[whydontu]

When all else fails, cheat. NEMA 34 motor in hand, dial indicator, scale.

Ran the mill table almost all the way off to the right. Set up dial indicator and zeroed against table using nearby drill press table as reference. Plunked the motor on to the table overhanging the mill handwheel. 0.002” deflection at end of mill table.

Moved the dial indicator to the center of the mill table, directly under the quill. Referenced off the mill base. Zeroed. JPG 3915. Plunked the motor down again. Can’t see any dial movement. As predicted, referenced to the quill, adding the motor weight to the end of the table has little or no influence on the relative defection between quill and table. JPG 3917.

 

[Susquatch]

Set up dial indicator and zeroed against table using nearby drill press table as reference.

Normally, this isn't a good idea because it measures EVERYTHING between the drill press and the mill table (floor, base screws, isolators, stand/base, etc. That 2 thou might be less if measured between the mill head and the table instead of using the drill press as a reference.

I'm guessing you don't really care though cuz it isn't more than that!

 

[whydontu]

Normally, this isn't a good idea because it measures EVERYTHING between the drill press and the mill table (floor, base screws, isolators, stand/base, etc. That 2 thou might be less if measured between the mill head and the table instead of using the drill press as a reference.

I'm guessing you don't really care though cuz it isn't more than that!

Nah, the drill press and mill are both bolted to a bench made of 6000lb pallet racking, with a bolted deck made of two layers of tongue and groove fir 2x6 and a top layer of 3/4” marine plywood, connected by 3/8” carriage bolts on 12” centers. Coated with half a gallon of spar varnish. My Toyota flexes more than my work bench.

 

[Susquatch]

Nah, the drill press and mill are both bolted to a bench made of 6000lb pallet racking, with a bolted deck made of two layers of tongue and groove fir 2x6 and a top layer of 3/4” marine plywood, connected by 3/8” carriage bolts on 12” centers. Coated with half a gallon of spar varnish. My Toyota flexes more than my work bench.

Ya, maybe so. But EVERYTHING moves. It's not if, it's only how much.

Even so, it has to be less than the 2 thou you measured so no issues. Be interesting to see how much it does move though.....

 

[whydontu]

I love questions.

Dial indicator base on drill press table, tip against mill head. Mill table run almost to end of travel. Zero indicator. Place stepper motor on table. Measure deflection. Roughly two thou deflection, which has to be total deflection of milling machine relative to drill press caused by movement of the milling machine in response to the offset weight addition.

 

[Susquatch]

Roughly two thou deflection, which has to be total deflection of milling machine relative to drill press caused by movement of the milling machine in response to the offset weight addition.

Well..... That didn't answer the question. You get to love some more.

which has to be total deflection of milling machine relative to drill press caused by movement of the milling machine in response to the offset weight addition.

Yes. But I want to know total deflection of milling table relative to milling head, not relative to drill press.

What you measured is total deflection of milling table vs mill plus deflection of bench.

Do the same thing, but measure vs the spindle or mill head.

It should be less than 2 thou. But how much less is the question.

 

[whydontu]

Or lots more, but entirely unrelated to the weight of the stepper. My mill is a BB CT129 dated 2008, and I'm the second owner. I don't do serious precision work, mostly use it to locate holes in panels and machine pipe fittings. Rarely do I need to be within 5 thou on anything I do, so I never bothered measuring tolerances.

Turns out that table rocks 3 to 5 thou at each end of travel. I set up a dial indicator about 2-1/2" right of the quill, zeroed it, and ran the table to both ends of travel with and without the stepper motor weight. And replacing the 5kg stepper with two 3.5kg bricks and a 5kg anvil only added maybe another 1 thou of deflection. and just to be sure I did the measurement at both the outer rim of the table as well as one of the centre ribs.

So I guess while I'm pulling it apart to install the stepper motor power drive I should maybe see if the gibs needs some attention.

 

[Susquatch]

More questions since you love them..... LOL!

What would the readouts be for two indicators - one mounted right of the quill and the other mounted simultaneously on the Drill press?

I'm skeptical of the higher reading. Although I suppose it's possible that some movements are cancelling out.

 

[whydontu]

Today's work, 90% done.

Tale of mild woe and greater success.

Sent Busy Bee a request for spare parts for my mill, I wanted a new end plate as I never want to drill holes into the original parts when I do machine mods. Never got a response. So went to Grizzly Industrial and took a chance and ordered the parts I wanted from them, making an educated guess that the parts for Grizzly G0704 and BB CT129 would be similar. Had the parts from Grizzly one week later. Still no response from BB, going on six weeks now. And I've been in BB's computer system since they first opened their store in Vancouver some time in the late 1980s.

Grizzly end plate is aluminum compared to the original cast iron, and a tiny bit shallower than the original plate, but the mounting holes line up and it works fine.

Cobbled together a mounting bracket from some 1/2" aluminum plate and a heavy stainless steel bracket, originally intended to mount an electric motor on a 4" butterfly valve. Use what you got for testing...

Coupled to the 10mm feed screw using a set of Lovejoy flex couplings, also from stock. (Never throw anything out.)

Installed a NEMA 23 425oz-in stepper motor.

DM860A stepper driver, 60 VDC power supply, and a strange little stepper control board off Amazon. $11.

https://www.amazon.ca/dp/B08THJ615X?psc=1&ref=ppx_yo2ov_dt_b_product_details.

Control board provides variable pulse step signal, Enable, and Direction signals.

NEMA 23 won't cut it, too wimpy to drive the feed screw.

Installed a NEMA 34 motor that I originally bought for @jcdammeyer 's ELS lathe controller, 1600 oz-in torque. Had to make up a new aluminum plate, and add some spacers.

Works just fine! Instant reverse, variable speed, just needs some tweaking of pulse frequency to get the feed rates a bit slower.

One super nice aspect of the controller, if I click the Enable button to OFF, the stepper driver removes power from the stepper motor and I can easily rotate the feed hand wheel. No need for a clutch!