Wednesday, December 17, 2014

Lathe: Tailstock Setscrew

The lathe tailstock uses a set screw to hold alignment while tightening the bolt underneath the tailstock. The stock set screw uses just a slotted head and its dog point isn't very well made, so when you tighten it the tailstock tends to move out of alignment. I replaced the stock screw with a socket head cap screw, and inserted a ball bearing in front of the new screw. This allow the set screw to apply much more even and stable pressure as it's being tightened.

Also, make sure you don't have any paint over spray on the tailstock's mating parts.

Monday, October 27, 2014

Lathe: Arduino TouchDRO

Please read the complete article:

After finding myself with an extra iGaging digital scale I decided to convert to the lathe to use the Arduino and Android powered TouchDRO. I had already installed an iGaging scale on the Z axis (see previous blog post) and adapted a cheap Chinese caliper to the cross slide. Going to an Arduino powered TouchDRO meant replacing the caliper with the iGaging scale.

The scale's read head needed 1.5mm machined off the cover's mating surface. A new bracket holding the scale needed to be made as well. While I was at it I added a second mounting screw for the read head to insure it won't wiggle on a direction reverse as the scale's guide wear.

New iGaging scale mounted to cross slide.

 Unlike the mill's Arduino, I constructed this one using a prototype board. It's much cleaner and easier and I highly recommend it, even though it added $8 to the build. I used standard USB A connectors for the scales' interface since both connectors and cables are much easier to find. This forced me to change the cables on both scales, but that didn't cost much. The tachometer's plug is also USB to avoid the issue I had using a 3.5mm headphone jack for the tach on the mill. Everything was mounted in an old Dell laptop power supply brick's case I had on hand. Neodymium magnets were glued to the case's top for mounting on the back of the lathe.

The tach, like the mill's, uses a Hall effect sensor since they're much easier to set up than an optical sensor and are just as accurate in this application. The lathe previously had a spindle extension installed, and for the tach's magnet I drilled a hole on the extension and used JB Weld to mount a small neodymium magnet in it. 

Spindle extension with magnet mounted.

The tach's sensor was mounted to the outside of the lathe's gear cover. I considered mounting on the inside but space would have been an issue and it works perfectly well on the outside. I covered the top of the sensor with epoxy putty to protect it and keep any swarf from shorting it out. If you look closely you can see I've bent the sensor itself up and away from the spindle; that's to provide a more optimal orientation to the magnet. The sensor's USB cable is run down the back of the lathe to the Arduino's case.

Hall effect sensor mounted on gear cover.

I'm using a HTC Incredible 2 phone as the Android device running the TouchDRO application. Since the lathe only has three readouts (X,Z, and tach) the phone is adequate. It's currently mounted to the top of the headstock using a bracket I fabricated. 

Mill: Arduino DRO Tachometer

Please read the complete article:

The Arduino sketch for the Android TouchDRO has supported a tachometer for a while now. In fact, it supports two tachometers so TouchDRO can tell which way the spindle is turning. Between not realizing it had this capability and reluctance to try and build an optical sensor had delayed this project for a while.

At first I had tried using a prepackaged optical sensor for Arduino (available on Amazon), but switched to a Hall effect magnetic sensor since they're easier to use and for this application tend to be just as accurate. I was able to buy them already mounted on a psb with a LS393 comparator. The comparator allows you to have an essentially digital signal with it either on or off. 

I already have a belt drive installed, so I drilled two holes at opposite ends of the top pulley and JB Welded in small neodymium magnets. 

Magnets on the pulley. The black sections were for the optical sensor.

The top of the Hall effect sensor psb was covered with epoxy putty to protect it from swarf. Then it was attached to the pulley cover using foam tape and a mounting screw. The sensor itself hangs over the back of the cover and directly over the path of the magnets.

The Hall effect sensor covered with epoxy putty and mounted on the pulley cover.

I added a 3.5mm stereo headphone jack to the Arduino's case for the tachometer interface. If I get around to repackaging the Arduino I'll change it to another interface since the 3.5mm jack will short power to ground as the connector is inserted or removed, so you need to power down the Arduino before doing so. It's not a show stopper, but it is annoying. I've hot glued the connector so it can't accidentally pull out while in use.

Be sure to add a 10K pull down resistor to the sensor output.

There were issues with TouchDRO reading 20-40 times too low on the tachometer. After some time spent on the TouchDRO Google+ development forum I changed the Arduino sketch to one being developed by Ryszhard and the tach immediately started working. I checked its readings against my laser tachometer and they match to within 20 RPM. 

 Ryszhard's sketch:

Thursday, June 19, 2014

Mill: Large Table Upgrade

I installed the large table from Little Machine Shop on my mill. It's the larger table with LMS' own mills use. Let me say I can't think of anything I can do on the larger table that I couldn't do on the stock table, however, the large table is a lot nicer to work with and makes setup easier. It also has a much more robust Y axis gib which is a lot longer than stock and has two additional set screws to adjust it.

Close up of the Y axis gib set screws.

Installation is pretty easy, and I took the time to drill all the holes for my DRO setup before removing the old table from the mill. I also made my standard modifications to the gibs, as described in a previous post.

All in all, I really like this upgrade and it makes the mill nicer to work with. It surprising how much of a difference a couple extra inches makes. My biggest issue with the upgrade was the poor quality of the Y axis dovetails; there is enough variance that you can feel the handwheel get significantly tighter as you move the table toward the column. Fortunately I don't usually move the table that far toward the column, but it's still something I may try to address later.

The larger table installed. The power feed is visible on its left side.

Mill: New Mill Base

My X2 mini mill sits on a wooden work bench, and I noticed the mill's tram would change from day to day. Finally I figured out the wood bench was flexing and changing from temperature and humidity enough to effect the mill sitting on top of it.

To test my theory I took a big piece of 1/8" steel and put the mill on top of it, helping to decouple the mill from the bench. That fixed the problem quite nicely.

To make the fix more permanent I took a 1/2" steel plate and, with the help of another machinist with a bigger mill, milled both sides of the plate flat.

1/2" steel plate milled flat, with locations for posts marked.

I then drilled four countersunk holes in the plate for the mounting holes in the mill's base and JB Welded M6 studs into the holes. The mill was then placed on the plate and nylock nuts were loosely tightened down. They're left a little loose because I want the mill to be able to move slightly on the plate in order to keep it decoupled from the bench.

The new base in place and bolted down.

Tuesday, June 17, 2014

Mill: Gib Set Screws

Please read the complete article:

I don't like the how the stock set screws use a locking nut. Tightening the nut loosens the set screw against the gib. Instead I threw away the lock nuts and use green Loctite. Green is specifically designed for set screws and works very well at maintaining consistent locking even as you turn the set screw. You also apply it after installing the set screw as it's drawn in by capillary action.

UPDATE: This technique worked ok, but the set screws could still loosen over time, so I ended up going back to locking nuts.

Tuesday, June 10, 2014

Mill: Power Feed

I've wanted a power feed for the X axis for a while now, not only because it gets pretty tiring cranking away, but also because a power feed is much smoother than I could ever be and that will produce a better finish. I decided to wait until I installed the large table upgrade since that would effect how the motor mounted to the table.

I had been looking at using a car's power seat servo as motor for a while, and I finally committed when I saw apointofview's post at Home Model Engine Machinist use the exact same motor to create a very elegent and simple power feed:

The new large table's lead screw extends through a bushing at the left end of the table and has a slot cut in it. I pinned a steel key in place in the slot. I then took a 14" flanged nut and milled a slot in the middle of it, just large enough to fit onto the key.

Key fitted to the slot in the X leadscrew.
14mm nut with slot milled into it.

For the motor itself I used a widely available servo for a power seat in a car. You can get them from several places including Amazon: The motor is reversible from left hand to right hand orientation. I used this for the speed controller: since it was cheap and had good reviews. If I were to do it again, I probably would have bought this to save a couple of dollars:

For the direction switch I bought a DPDT DC switch that's center off. That way I can set the switch to center to stop the feed. I found a DPDT switch on eBay specifically design to be used as a direction control for DC motors, so that saved me a couple of wires and a bit of soldering. For the table jog I decided to just use the speed control and wired in a switch set the control's speed adjustment pot to maximum.

Wiring diagram.

For the power I used an old 15v laptop power supply since they tend to be pretty good quality. I wasn't worried about using 15v since most cars run between 13-15v anyway.

The whole mess of wiring was then packaged into a project box from Radio Shack and attached to the side of the lathe. The extra switch on the top right of the control box is the on/off for the spindle light.

Upper left is power, lower left is jog, center is direction.

A piece of aluminum plate was machined to bolt to the end of the table, and to accept the bolt pattern of the motor's housing. The back of the motor mounting holes in the plate were countersunk to provide clearance for the end of the table. The motor's leadscrew was removed and four sides milled flat so that a 3/8" drive socket can easily slide onto it. The leadscrew was then cut down so it would clear the end of the table's leadscrew's key when mounted in place.

Leadscrew machined down and motor bolted to mounting plate.

A 12 point 14mm socket was slid onto the leadscrew and then everything was mounted on the table. To engage the power feed I slide the socket forward until it engaged the 14mm nut. To disengage I simply slide the socket back. I purposefully made everything with very loose tolerances to allow easy engagement and to prevent any binding from misalignment. I cut some grooves into the socket on the lathe for a little more traction.

Power feed assembly bolted to the table. In the picture the power feed is engaged.

The latest beta version of Android DRO (now called Touch DRO) will display the feed rate for a selected axis. When using that with the power feed I can set the feed rate very exactly. As the power feed is configured now it will hold a speed to within 1/2" per minute, which I'm quite happy with. I think if I remove the X hand wheel's handle before using the power feed that variation will go away, as I suspect the weight of the handle is speeding up the feed for half a revolution and slowing it down for the other half.

FOLLOW UP: After a couple of days I started getting some vibration in the motor when running in a certain direction. I was able to fix the issue by taking out the plastic screw which supports one end of the motor's shaft and pressed a small ball bearing into its bore. This gave the motor's shaft a solid and smooth surface to push on and rotate against.

UPDATE: I've since relocated the control box to the side of the top of the column. It's a lot easier to maintain the mill and access the controls.

UPDATE: I added a small cover to the leadscrew engagement to keep chips off it. It's mounted via magnets on the underside of the cover.

Cover in place over leadscrew.

Cover upside down to show the mounting magnets.

Thursday, June 5, 2014

Mill: Spindle Light

I wanted a spindle light to try and reduce shades on the work as much as possible. So far I'm a pretty big believer in you can have enough light on the work.

People apparently install LED rings in their car's headlights to try and make them look like BMW headlights. So on eBay I found a LED ring with an 80mm diameter, which is about the largest the bottom of the mill's head can accommodate. It cost $22. There were cheaper ones, but I bought this because the LEDs and electronic were sealed, while on others the LEDs were exposed, and I was worried about swarf shorting it out.

I used 5 minute epoxy to glue 4 neodymium magnets to the back of the ring to mount it to the mill. It was then that I discovered the cover surrounding the spindle on the X2 is made from plastic. However, with a bit of maneuvering I managed to get three of the magnets on metal and get the light mostly centered.

Neodymium magnets epoxied to the back of the LED ring.

The wire from the light runs up the left side of the head, to the top of the column, and then down to the control box which houses a simple switch to turn it on and off.

Spindle light mounted and turned on.

Thursday, April 24, 2014

Lathe: Tormach 0XA QCTP

I've had a A2Z CNC type QCTP on the lathe for about a year now and it's been a huge improvement over the stock tool post. However, I didn't like that it was made from aluminum nor that locking the tool holder in place pushed it away from the tool post itself, so only the two dovetails were in contact.

I happened across the Tormach 0XA QCTP on Little Machine Shop and immediately ordered one. Not only is it made of steel, but it's also an AXA or sliding wedge style tool post. In this design as the tool holder is locked to the post, it's pulled in against the post, so you have a lot more contact area between the two than with the A2Z CNC style; the 0XA is just a better design. I also really like that it comes with a knurling tool, which saves me an extra $25. Also, tool holders for it are cheaper than the A2Z CNC as well, which is nice since you'll want a lot of tool holders.

After installing it last night, I'm very impressed with it. The quality is excellent, and the added weight of the post and the holders is very evident. The back of the holders have a ground and polished finish. There's a very nice, tight fit between the post and holders.

Brand new the A2Z CNC QCTP costs $99 while the 0XA costs $129. However, in my opinion the 0XA is absolutely worth the extra $30, especially since you also get a knurling tool with it.

Tuesday, April 15, 2014

Lathe: Bike Computer Tachometer

Please read the complete article:

After testing, it appears that most bike computers lose a lot of accuracy over 1,000 RPM. I suspect this is probably because the reed switch used to sense the magnet simply can't operate that fast.

I had been planning on installing bike computers on both the mill and lathe, but the 1,000 RPM limitation doesn't make it worth my time since a lot of my cutting it done on aluminum which uses speeds often well over that.

As of now the plan is to continue using my non-contact laser tachometer and I'll mark graduations around the speed control knob. Crude, but it should be effective enough.

Sunday, April 13, 2014

Lathe: Leadscrew Play

My lathe had play in the leadscrew. I've heard of people shimming the leadscrew at the pillow block, but to make is easier I just removed the pillow block, threw it in the mill, and made the bolt holes into ovals so I could adjust the clearance between the pillow block and leadscrew by just loosening the bolts and sliding the block. Using this I was able to remove all the play from the leadscrew.

Saturday, April 12, 2014

Lathe: Spindle Extension

I wanted to put a spindle extension on the back of the lathe for a couple of reasons. First I wanted to make sure any swarf which made its way into the spindle bore fell out of the machine instead of into it. Second, I wanted a convenient place to mount a magnet for my bicycle computer tachometer. The spindle thread is M27x1.5 and I was able to find a corresponding nut on Amazon. It's actually designed for compression fittings, but it works great. It's available in two sizes, 18mm and 35mm. Being cheap I opted for the less expensive 18mm. Once I received it I turned down 9mm of its end until the wrench flats disappeared, and then a little more. I then had to enlarge the opening in the gear cover which was easily done with a Dremel and a sanding drum. Finally I threaded the nut onto the end of the spindle and snugged it down. Neither lock ring on the spindle needs to be removed to install it.



Lathe: Gear Cover Hand Screws

I decided to replace the gear cover screws with thumbscrews to make its removal easier.

I bought a section of M5 threaded rod and two 25mm M5 threaded standoffs from McMaster. 9mm of one end of the standoffs were turned down to 8.5mm diameter, and then were red Loctited onto a 60mm long section of threaded rod.

Tuesday, April 8, 2014

Lathe: Carriage DRO

Please read the complete article:

I was finally able to install the DRO on my lathe's carriage. I used a 12" iGaging digital scale with a remote display, the same type I used on my mill.

When I do threading on the lathe I'll just back out my bit and run the lathe in reverse to rest the carriage, so I have no need for a threading dial and had removed it a while ago. That left a great threaded hole to mount the DRO bracket to. I suppose if you still have the threading deal there you could always just sandwich the bracket between the apron and dial. The bracket was easily made from 1/8" 1" aluminum angle. It's a bit overkill and I'll probably trim it down a little more later, but it's not really in the way as is.

The design I ended up with places the DRO below the lead screw, so it's fairly well protected from swarf. The mounting for it is quite stiff, so I only have the scale attached at one end. That was particularly helpful since I didn't want to try drilling into the lathe' body at the head since it houses the motor right there.

I still haven't found a good place to mount the remote display though, and I may end up doing an Arduino DRO display like I did for the mill.

Both the carriage and cross slide DROs are visible.

Carriage travel at the extreme right of the bed is limited slightly.

Monday, April 7, 2014

Lathe: Changing Gears

It took me a minute to figure out the proper way to adjust the lathe's gears when converting to and from threading operations. There are two adjustment, one locked in place by a 14mm bolt, and one by a 10mm nut. Use the 14mm bolt adjustment to set the clearance between the idler gear and the lead screw, and the 10mm to adjust clearance between the idler and spindle. Proper clearance should allow just the slightest amount of play between the teeth of the gears.

Lathe: Shimming Carriage Gibs

The gibs on the carriage went out of adjustment again, bringing the project I was working on to a temporary end. Highly annoying. Instead of just adjusting them, I took the time to shim them.

The mini lathe's carriage gibs are held in place and adjusted by opposing screws. Two of them push the gib away from the carriage, while three of them tighten the gib to the carriage. In theory, by balancing the screws against each other you can adjust the gib/carriage/bed clearance. In reality it's a crappy system. It's hard to adjust, puts a lot of stress on the gib, and bends the two ends of the big toward to the bed, making it even hard to adjust properly.

To switch to shims remove the small locking screws and leave the socket head cap screws. Starting with the back of the carriage, slide a shim between the gib and carriage, tighten down the three screws and see what it does. If the carriage won't move it's too tight, if there is any movement it's too loose. On mine the back gib ended up with a 0.035" shim. Once that's done, repeat on the front. When testing for movement, make sure there is there isn't any play between the carriage and the bed's prism.

Once the shims are in place, apply blue tread locker to the screws, tighten everything down, and lube the ways.

Mill: Mill Column Tuning Device

I jokingly call this my Mill Column Tuning Device, or MCTD.

It's very difficult to get the column exactly perpendicular to the base using just shims. Additionally, temperature can make the column lean more one direction or another. What I did was fabricate a bracket which is bolted to the base using the column's mounting bolts (longer bolts were required), and then connect the bracket with the top of the column using turnbuckles. Both the side and back of the column received a turnbuckle. Now, as I tighten or loosen the turnbuckle, it will push or pull the column slightly to account for any slight misalignment.

The turnbuckles actually bend the column, so the amount of movement becomes greater the closer to the top of the column you get. Therefore you can't use this system to get the column exactly perpendicular to the base at every position on the column. However, once you've done the best you can with shims, this will let you get it a little closer still, and I'm happy with an incremental improvement.

The rod ends were stock items purchased from McMaster. The turnbuckles themselves were created from 1/2" mild steel rod. The ended were turned down and then threaded on my mini lathe. One end is right hand thread, the other is left hand thread. A M14 nut was slide down the turnbuckle and then welded in place, giving a convenient way to turn them.

The power box was spaced away from the column using simple standoffs to make room for the turnbuckle.

As an added bonus, the bracket was is an ideal location to put my mill's DRO's control box and let me clean up the wires.

The system installed on the mill.

The bracket the bottom of the turnbuckles attach to.

Top of the turn buckles, and the power box spacers.

The DRO control box on the bracket and the wires corralled.

UPDATE: I ended up taking it back off the mill. In practice it proved pretty hard to use, and my column is already close to perfect that it just didn't make sense to keep it. Removed it also allowed me to move the spindle light and power feed control back to the top left of the column.