Pm45m Cnc Conversion With 7800 Rpm Belt Drive 3 Hp Upgrade, Bt30 Spindle, Power Draw Bar + Enclosure

gt40

Registered
Registered
Joined
Feb 8, 2015
Messages
107
I have this previously posted on another forum and Darkzero thought it might be of interest to folks with Rf45 type mills.

Here are couple of pics of the machine in Jan when it arrived back in 2012:
newmill2.jpg
Welded a basic stand:
overallshot.jpg
At this point the mill was just a variable speed PM45 mill.

I went with DMM Tech 900 watt 3 axis ac servo kit. It seems more elegant than doing the stepper route. As I mentioned in the lathe thread, they have provided me with great service and tech support and the servos have been flawless on reliability.

I ordered (2) 5/8" 5 tpi ball screws for x&y axis and (1) 3/4" ball screw for the Z from Roton. I also ordered double ballnuts for each axis. In hindsite, I would recommend 3/4" for x and y and 1 inch or larger for z. You basically are limited to anything larger on these machines because of space considerations:

The ball screws from Roton x and y were 5/8" x 5 tpi 15" and 48" respectively to be cut to size and z axis was 48" to be cut to size but 3/4"/ 5 tpi

5/8 X .200 Right Hand Recirculating Lead Screws & Nuts for Power Transmission - Roton Products, Inc.

3/4 X .200 Right Hand Recirculating Lead Screws & Nuts for Power Transmission - Roton Products, Inc.


I also ordered (8)12mm double row angular contact bearings + (4) 15mm from

http://www.vxb.com/page/bearings/CTGY/Double-Row-Angular-Contact-Bearings

I machined my own bearing blocks with 2 bearing per side on each side of the block.

I picked up 80lbs of aluminum rem stock from industrial metal supply. Mostly 2 1/2" x 3" thick and 1/2" thick plate. There is also 20lb chunk of steel rem for the z axis ball nut mount. All of the aluminum was at 1.69lb and the steel was 1.39. Basically a $150 for stock metal costs for the conversion:

metal-rems.jpg
Got to use the new Glacern Face mill. It really is awesome!

I should have snapped a pic right after I milled it- shiny metal and dead smooth flat. I took this one 3 hours later after measuring the mill a and stock to check fitment and design. This chunk is a a bearing block to be.
left-side-x-axis-bearing-block.jpg


Here is a pic with the basic shape cut and one side bored for the double row angular contact bearing. There are to be 2 double row bearings on each block in a sandwich to support each end of the ballscrew.
left-side-x-axis-bearing-block.jpg
I mount ad block like this on each side of the x axis and then a 1/2" thick plate for the offset to mount the motor.

I machined to identical bearing blocks and will use the stock bolt holes for coplanar alignment. Here are some pics to show perspective better:

Width of bearing block:
6.jpg

I basically cut the blocks to size and then drilled through the block with a 1/2" drill to to provide a way to locate and bore the bearing holeon each side of the block and have them be coplanar.
5.jpg
Checking fitment:
4.jpg
gang machining to make them identical:
2.jpg

finished blocks with 2 bearings per side:
7.jpg
I remember the day everything arrived:

dmm ac servo kit, antec 65v psu, ballscrews, ballnuts...
everthing1.jpg

ballscrews:
ballscrews.jpg
dmm control board:
normal_circuit-board.jpg

i removed the table and measured for the x axis ball nut mount:

saddle.jpg

Here is a pic of my X axis setup. The DIY pre-loaded double nut seems to work. It currently has 2 50lb bellville washers per screw to provide pre-load force.

In this pic, you can see the nut loose and the parts, there are actually 2 opposing bellevilles per bolt like this () They generate 50lbs wer washer x 8 washers:

x-axis-ballnut-block3.jpg

All tightened up:

x-axis-ballnut-block4.jpg

Head on view:
x-axis-ballnut-block1.jpg

It fits!
x-axis-ballnut-block2.jpg

Fyi, the preloaded ballnut setup with 2 cheap roton ballnuts (24.00 each + 5/8" ballscrewat $11.00 a foot) seems to have absolutely no backlash with 300lbs of preload from belleville washers.
testfit1.jpg

Test fit of the gear side:testfit5.jpg

Tail side:
testfit4.jpg
BTW, this is one of my favorite tools. I always hated tapping and these spiral flute taps are awesome. The chips come through the flutes and they just cut better. I got 4 sizes so far all on ebay:
testfit6.jpg
Close up because I really really like these taps!
testfit7.jpg
X axis is nearly done:

20120428_163115.jpg
Y axis ball nut block with boss:
20120511_185754.jpg
I had to take the mill head off. Considering the weight, this was done by disassembly. I added bigger motor anyway:
20120511_184843.jpg
Gears- I am getting rid of these and converting to belt drive:

20120511_184818.jpg
Z axis lower bearing support block. Alignment of Z axis was a little tricky. In the end, i snuck up on it but had to make several of these to get the fitment right. It was hard to take accurate measurements inside the column:
20120526_163032.jpg
z axis ballnut block. Basically a 3"x8"x5" block with a boss cut on the end. I am really happy with the fitment as there is absolutely no slop:
20120526_163046.jpg
Here is the opposite view with the boss cap:
20120526_163054.jpg
3 hp inverter duty motor and drive. It is heavy at 130lbs and makes the AC Tech vfd look tiny:
motor~0.jpg
motor-tag.jpg
I was going to use the pm factory dc motor but I scored the inverter motor for 180 dollars new surplus and the motor that came with the mill is a "tad" optimistic in its "3hp" rating. The lincoln motor is really a monster in comparison and 4x the size. It will be a little interesting shoe horning it on top of the head...

On another note, I was able to cut grooves in the ways and drill/tap x, y and z axis using the bridgeport at roadrace engineering. They work on my car and let me use it after hours. First time using one and it cut like butter:
grooves.jpg
I performed surgery on the top of the mill, opening the top up for my flush, undermount bearing block. Not the prettiest cut but it will be covered entirely by the servo mounting plate. I basically climbed up on the mill stand and cut it out with my plasma. I think I won't do that again :p
top-cut.jpg

More to follow...
 
My favorite way for boring the bearing blocks is proving to be a surplus 1" diameter mt3 drill bit. Much faster making the hole and then finishing off with a boring bar:
drill2.jpg
drill1.jpg
X, Y and Z axis ballnut blocks, angular contact bearing blocks on both ends a and x and z servo motor mounting plates are finished. I also was able to test fit the 3 hp inverter duty motor. It is a beast and I think you can't fit a much larger motor without mounting it offset so I am extremely happy. It clears the column with a full 1/2" to spare with it's designated mounting location:
motor-test-fit1.jpg
I cut the side plates and hope to mount it tomorrow then make the the 2 step poly v pulleys. That should allow for 2 speed ranges up to 7000 rpm with the vfd.

Here is a shot of the z axis bearing blockand double ball nut with 300lb belleville washers to prevent backlash:
z-axis.jpg
Keep in mind the bearing block is 8" x 4" chunk of aluminum....


Here is a preliminary setup for the motor mount. Given the size and extra power, I wanted a really stout mount. The horizontal plate on top will bolt to the motor and pivet to adjust the belt tension. The top plate is bigger because the motor diameter is 2 inches larger than the with of the head. I will make the pulleys now I just got my 8 rib poly v belt from Mcmaster Carr.
motor-mount.jpg
side view:
motor-mount2.jpg
I cerakoted the parts sniper grey to match the hammered grey finish i painted the mill. I am looking forward to assembling it and start things moving finally. This was a bigger project than I thoughteek.gif

Parts ready to go back on the mill:
parts.jpg
I completed the VFD conversion by machining 2 identical 8 rib poly v pulleys exactly .0921 distance between 40 degree .072 deep v grooves and mounted the motor:
vfd.jpg
I hooked up the z axis servo and was able to test it with the heavy motor on it. I only got to run it at less than 50% speed as I was scared I would have a crash otherwise.


I got the belt drive working and AC Tech VFD wired and controlling the motor. I need to adjust the pulley's though as I have a little wobble on the motor side.

I am extremely happy with AC Tech. I hooked it up in 10 minutes and it just works!
normal_actechvfd.jpg

I was going to go with a separate box but ended up re-purposing the stock electronic box instead. Everything fits together except for the pc.

I also re-purposed the stock master switch after seeing it was rated for 30 amps. I also mounted an e-stop switch on the side. I will add another on the front later.

Power and e-stop:
power-and-estop.jpg
Here is the control box. It has the 3 drives, breakout board and antek power supply. I was able to set up the power switch so I have one 240v going into the box. The power switch controls everything so turning it on sends power to the vfd and spindle via relay and also to the 21 amp power supply:
control-box.jpg
I cut all the wires to length and soldered/shrink wrapped things and anchored things.

I got the machine working with Mach 3 and Z axis is now functional. E-stop, limit switches work and the DMM servos are moving the heavy motor and head with authority:

First chips- Here is a video using the mach3circular hole wizard to drill 9 holes:

On another note, accuracy is looking pretty good on simple test calibration:

Mach 3 has a built in DRO/ Jog function with "go to capabilities. You can also add a pendant which is basically a hand held controller to do this while you look at the mill instead of the computer screen if you want. So basically, I have a 3 axis power feed mill with jog capabiility to 10th via Mach 3. Since as I am still learning how program it, this is actually really cool because I can move the table faster and better than I could before and cutting/ drilling is smoother than manual. The go to and zero functions are pretty neat as you can tell it to move all 3 axis to a given point.

Repeatability seems to be pretty good too. Below is a video where I zeroed the dro in Mach 3, zeroed the indicator and then moved the table and then hit move to zero button on mach 3:

I ran the belt driven spindle and measured 7680 rpm with a laser tach. Shortly there after the stock bearings started to smoke from the speed. New angular contact bearings coming tomorrrow. I can verify the following sizes for my PM45M mill that I measured and ordered:

Lower bearing: Kit10853
7207CYP4 Nachi Angular Contact Bearing 35x72x17 Abec-7

Upper bearing: Kit10852
7206CYP4 Nachi Angular Contact Bearing 30x62x16 Abec-7

In addition for the belt drive bearing tube that is per GD Marsh's design, I ordered:

(2) Kit8891
7007B Angular Contact Bearing 35x62x14 Ball Bearings

FYI, make sure you install the bearings correctly facing the correct direction. I managed to do both wrong and killed a set of cheap angular contact bearings from vxb by installing them with the cage plate facing out which is backwards + tapping them in with a dead blow hammer.

Don't beat on bearings. In hindsight, this was completely obvious :p

Handle with care and use a press etc and properly support them by putting the load on the outside only and not the floating assembly in the center.


Good video of the process from Hoss:

http://www.youtube.com/watch?v=hP-XvbV7Pc4]G0704Spindle Bearing Replacement - YouTube


I ended up buying the fancy abec7 bearings because I want the machine as quiet and smooth as possible.


Basic cutting test:

Made a new stand is 4" x 4"x 1/2" thick tubing with 2"x 2x 1/4" thick crossmembers. The top is 3/8" thick 30" x 30" and weighed 96lbs by itself. All in all, I got 422lbs worth of steel from the rems pile at industrial metal supply at 68 cents a lb.

newtable1.jpg

retractable levels:
newtable2.jpg


I added 6" steel casters and it rolls really nice even though it is over 400 lbs. Added Barrymount Leveling Mount LM5-B isolation pads that are 6" diameter and weigh over 5lbs each. they are offset the wheels so you can roll the mill around and then level the table after you get it where you want. I will be mounting the flood enclosure on top of the stand.

I got xyz touch off probing to work. Thanks out to erniebro and this thread:

http://www.cnczone.com/forums/mach_w...tml#post436238

I installed the Mach Blue Probing by Big-Tex screen set and got it to work with the DMM tech DMB4250-8B Breakout board. I need to make an electrically isolated probe next so the stock doesn't need to be taped up, but the probe works. For connections, I connected a wire from the DMB4250-8B Breakout board's "a limit" in and vout terminals to make the circuit and then used the following settings under input signals in ports and pins in mach 3:




Probe:
Enabled
Port#1
Pin# 13
Active low

These settings worked with the dmm tech board. I know this has been out on cnczone for a while but it is still awesome.

Video:

Update:

I added an ethernet smooth stepper to the DMM Tech servos and it worked just fine. I needed additional imputs beyond the DMM tech board's for touch probe + hall sensor based tach and spindle speed control etc.

I was able to use the a limit input on the DMM tech bob for either a probe input or a spindle tach with a hall sensor but not both. Adding the Smooth stepper, I gained 2 more parallel ports worth of inputes + extra precision and reliability of a dedicated motion control board.

I also started work on my flood enclosure. First I had to make a bending brake. The bending brake is basically 3 6ft 1/4" angle iron pieces, some black pipe and 2 3/8 thick pieces for the arms. Plans I got from here:

www.ch601.org/tools/bendbrake/brakeplans.pdf

Pics:
bendbrake2.jpg
I could bend 1/8" thick aluminum just fine:
bentpiece.jpg
This was the first time I ever bent sheet metal but the brake works great:
traypic1.jpg
progress on the coolant tray:
coolanttray1.jpg

Flood enclosure base is done and the mill re-assembled. I added an IMachIII M1 pendant from Vista CNC and also calibrated the axis with my dro.

The big news is I have achieved my accuracy goals using the Roton ballscrews and double nuts= .0001" and .0002 backlash total. Considering the ballscrews so inexpensive, this is pretty cost effective. I am hooking flood coolant and building the top of the flood enclosure next.

Video:

I received my 30 taper spindle from Industrial Hobbies. FYI, 100 bucks and they fit the pm45 as well as the ih and zx45 clones of the rf45. Installation was straightforward and it is a drop in. Press the bearing on and reinstall. I decided to go with 30 taper because of cheap bt30 er20, er32 and er40 collet chucks for 25 bucks and the ability to add an automatic tool changer at some point. The 30 taper is beefier and you gain almost 2" on z axis compared to using an er collet chuck. I also picked up a full 1/32" increment set of er20 collets (15 in all) + a set of er32 collets that are all techniks.

Pic of the two spindles. The one on the top is the new 30 taper spindle. I had just pressed one of the bearings on:
bt1.jpg

That's it for now. I will post part 2 tomorrow.

Thanks for looking
 
Great job on the CNC conversion. For the Sheet Metal Brake you build, did you use any plans?
 
Part II:

BT30 conversion with pnuematic drawbar

Here is a business end shot between stock R8 spindle and and the BT30:
bt2.jpg

Industrial Hobbies said they have over a 100 30 taper spindles and I can confirm they fit the PM45m mill.

To convert the manual draw bar to pneumatic you have to work within the stock spindle measurements . The drawbar hole diameter is .684" with the top of the cone at .703 for a short bit. That is because it is NMTB 30. My plan is to cut a relief above the cone where the stud would sit so my gripper can release when the drawbar is pushed forward and put a triple valve spring on top of the the drawbar in place of bellevilles. Because my motor is so large, the air cylinder will be mounted upside down in front of the head and push up on a pivot arm to push down on the drawbar to release.

I took the IH NMTB 30 spindle and bored the straight section deeper by an inch while keeping the diameter to 0.72". This allows me to fit a ball type gripper where the straight section would be. Next I cut groove a 0.25" wide and a depth of 0.95". When the grippers balls are aligned with the groove, the larger diameter allows the pull stud to be released. I made an aluminum version of the spindle and it released properly. The trick is measuring everything right.

I will post some more detail later but here are some pics of the new drawbar. It is O1 steel, properly hardened and tempered to around 55 rc. (PID kiln)

Gripper Close up:
gripper.JPG
Gripper on drawbar with bt30 toolholder:
gripper2.JPG

Gripper parts:
gripperparts.JPG

There is no room for springs inside the IH nmtb spindle, I am putting the spring on the top of the pulley with a top hat and fabco 2 stage 4" cylinder- 1500 lbs of spindle clamping force.

Meanwhile, took 8"x 3" thick diameter round of 7075. This is inspiring me to make a new head assembly with an extension. I already cut the top plate by the y axis to the vertical wall of the base and that freed up a lot of travel. Adding an extension

Having a head extension allows up to 12" y axis travel on the RF45 type mills.
head%20extension1.jpg


Finished part bored for (4) 5/8" bolts:
head%20extension2.jpg

Back to the spindle, I took the plunge literally and cut the inside of the spindle to make way for the pull stud gripper. This entailed slightly boring past where the taper terminates and then cutting an internal groove for the balls in the gripper to drop into. I test fit and with the balls in place, tool holders are held and then if you allow the drawbar to move down approximately 1/2", the tool is released.


I spent some time further test fitting the spindle. I found I was .0006 too narrow when the gripper was retracted with all 4 balls in. During my initial test fitting, I only put 2 balls in. With all 4 in and the gripper assembly retracted, the tool holder wouldn't retract all the way into the spindle. As a result, I had to slightly enlarge the bore in the spindle. After that, the gripper works smoothly.

Here are a few pics:

Inside the spindle:
gripper%20balls.jpg


Tail end:

tail%20end.jpg

Basically you need to machine the part in the oval, make a gripper and draw bar.

spindle%20cutaway.jpg

After finishing the plate for the 2 stage air cylinder, I took some cuts. I put a tool in, fully inserted it using the gripper with the balls and then manually tightened the bolt on top of belville washers with a wrench. No problem taking .2" cut with a glacern 4" face mill- it holds just fine.

The power draw bar works. I am really happy because bt30 tooling is plentiful and cheap smile.gif

Video below:

I cleaned up everything with the enclosure and things are starting to come together. A few pics of the mill in it's current state with a power drawbar, one shot oiler, enclosure and diy fog buster with pressurized coolant:

Shot of the mill head with the bt30 power drawbar installed:
millhead.jpg

Belleville springs:

bellevilles.jpg

Overview shot showing the enclosure. You can also kind of see the 3.5" head extension. I am getting a full 12.75" of Y travel now. I will need to make a fixture plate to take full advantage. Considering these mills have 8" of y travel, this is really awesome:

millenclosure.jpg


Power drawbar switch. I ended up using a foot pedal momentary pneumatic switch. Seems more robust than the toggle type and the price is right:

footpedal.jpg


Diy fog buster built out of scrap 1/8" npt plumbing parts I had around and a mig nozzle tip. I forgot to snap a pic of the pressurized coolant tank but it is a whole house water filter housing modified with a pick up tube. compressed air goes in and forces the coolant out through the pick up tube connected to the top. The needle valves allow for both pressure and fluid level adjustments. Works really well:

fog%20buster.jpg


Sample part being machined:


Thanks for looking...
 
In addition for the belt drive bearing tube that is per GD Marsh's design

I will be doing the belt drive conversion soon. Can you provide more detail on the drive bearing tube?

Thanks,

Jay
 
Great work! I have one question, why did u use pulleys for the servo motors instead of direct drive? What are the benefits of using pulleys in this case other than speed of course


Sent from my iPad using Tapatalk
 
Great work! I have one question, why did u use pulleys for the servo motors instead of direct drive? What are the benefits of using pulleys in this case other than speed of course


Sent from my iPad using Tapatalk

Timing pulleys provide an accurate and shock absorbing method of power transmission. They also were a way to make the foot print more compact. The main reason though was to make the resolution as high as possible. I forget the exact number of teeth but it is 1.8:1 ratio which increased the accuracy and you also increase the available torque because of the lower gear(at the expense of speed). The new servos though that are out are 65000 steps per resolution so If I had to do it over, I might just do direct drive. I am building a large vmc using epoxy granite and 36mm and 40mm ground ballscrews and going direct drive with it. There is more than one way to skin a cat.

There is a good lowdown on this subject on cnccookbook: http://blog.cnccookbook.com/2014/01/12/ultimate-benchtop-cnc-mini-mill-part-2-cnc-conversion/
 
Timing pulleys provide an accurate and shock absorbing method of power transmission. They also were a way to make the foot print more compact. The main reason though was to make the resolution as high as possible. I forget the exact number of teeth but it is 1.8:1 ratio which increased the accuracy and you also increase the available torque because of the lower gear(at the expense of speed). The new servos though that are out are 65000 steps per resolution so If I had to do it over, I might just do direct drive. I am building a large vmc using epoxy granite and 36mm and 40mm ground ballscrews and going direct drive with it. There is more than one way to skin a cat.

There is a good lowdown on this subject on cnccookbook: http://blog.cnccookbook.com/2014/01/12/ultimate-benchtop-cnc-mini-mill-part-2-cnc-conversion/

Thanks, this clarified everything for me


Sent from my iPad using Tapatalk
 
Back
Top