A couple of years of ago I finally added hard lines to my air compressor throughout the shop. What a nice upgrade! No more hoses all over the floor. However, it seems my copper sweating skills are maybe not what they should be as it seems to leak a fair amount of air. After I switch off the compressor it takes something like 4 hours for the system to depressurize.
I felt I could do better.
But how to isolate the problem? And how to make it a complex project?
Well, that's easy. Just build a wireless network connected microcontroller with a pressure sensor, an SD card for logging, and a graphical display. Have all your data get collected in two ways - 1) local to the SD Card and 2) upload to ThingSpeak.com.
My thinking is to attach the logger to different segments of the air line. As I move away from the compressor - therefore adding more pipe joints, valves, equipment - I should be able to deduce from the pressure decay rates what the relative leak rates are per section. That should give me an idea of what to work on. Additionally, I can verify my repairs in quantitative terms.
So after a few days of cobbling things together
Here you can see it's attached to my air lines - running about 115psi.
Close up of the mess. The screen is an older Arduino TFT (now discontinued) with an SD card socket. The smaller chip above that is an ESP8266 WiFi chip. For $5 you can't beat it!
The ESP8266 is a 3.3v part which needs some juice when transmitting. I happen to have had lying around an old XBee shield that I had previously added a 2 channel level shifter to. It took care of both issues as it also has a robust 3.3v power supply.
Here's a link to the ThingSpeak page where I'm uploading the data : https://thingspeak.com/channels/6008
I'll have to update this as start logging the system and try to figure it out.
After getting the 4th axis up and functional, I thought I might as well try a project utilizing it. My concept was to create a dining table using the rotary as well as some old barn wood we had left over from building our house. Here's the CAD version of the idea. The original concept had this with two legs. After calculating how much aluminum that would take and how long it would take to machine, I scaled it back to a single leg.
I had wanted to use the rotary as a true 4th axis (simultaneous movement in all axis), but ran into trouble creating the gcode. My skills in NX 7.5 just weren't up to the task. So I back-pedaled and went with more of an indexer mode - basically turn 90 degrees between each operation. Although it's not what I started out wanting - it still allowed me to machine all four sides without removing it from the fixture.
Which brings me to the fixture. I've never really put much time into setups as I usually just want to get the part done. This time, I realized the number of parts I wanted to make plus using the 4th axis required one. I'd also never used dowel pins for location devices, so thought that would be fun too! Although it's a bit cumbersome to mount and dismount the part, it worked as designed. When I originally made the fixture, I was a bit careless about maintaining parallelism between the top and bottom. I realized this at about the 3rd part when I was having trouble with things matching up when the rotary went all the way around. Once I fixed that - it turned out to work very well.
Once I got that sorted, the first operation was to cut a chunk off the raw 5"x2.5" stock.
Next up was to add a set of holes - 3 threaded and 2 reamed (for the dowel pins to locate) on both sides to use for mounting in the fixture.
Then I sawed this at a 30 degree angle to make 2 blanks. Here it is mounted up on the fixture in the rotary about to do the first position.
I had never used dowel pins to align a jig and got to use over/under reamers to make it work. It was actually pretty easy and it really made things very predictable and repeatable.
To complete 1 part took around 55 minutes of machine time. Most of this was trying to get a very fine finish on the contoured areas which took a lot of passes.
I was originally planning to do enough for 2 legs, which would have required 48 sections. At an hour each, that really added up.
Once I had the buttress pieces done, I moved on to the leg. I mounted the lumber in my lathe and turned the ends. What a mess! Turning wood in a metal lathe gets sawdust everywhere!
After turning that down, I slipped on some aluminum collars to each end and put together what I had for a bit of a mockup.
Now that I had the leg mostly there - I started in on the top. I had wanted to make it up out of a bunch of strips and glue/screw them together. I really ran into issues with warped and twisted strips. I ended up wrestling it around to something I considered acceptable, but it was a lot harder than I thought it would be. It ended up being 48x36".
One feature I attempted was to use 5 threaded rods to hold the top together. Although I got it done, it was probably the worst part of this whole job. I really underestimated how much trouble it would be to drill straight holes though 3 feet of 3/4" strips of wood of varying species! I thought it would be fun to make a set of custom nuts to hold it together. I wanted to recess them into the side of the top. I based the design off the 12 lobed base.
You may be wondering how I got those nuts tightened up. Well you make a tool of course!
I wanted to have a plate mounted on the underside of the top to use to attach the leg and also to help stabilize the wood. I found a galvanized round cover of sort make out of diamond plate. I was going to expose the diamond plate, but ended up liking the other side more.
Next was the base. I found a 20" diameter metal plate that was apparently used as a pipe cap. It probably weighs about 80 lbs and was a bear getting into the mill by myself! I milled a pocket in the top for the leg and on the underside I added pockets for 6 feet, plus thread milled the outer holes to accept an adapter I needed for the feet.
I realized I needed a way to securely mount the aluminum collar on the ends of the leg so they wouldn't spin. I decided to glue it on, but then added a wedge - just to make sure!
And here's the final result!
Some additional notes:
I think the overall design was OK, but in looking at the finished product I think it didn't come together as well as it could have. Not bad for my first attempt at furniture, but there's room for improvement. My brother pointed out there should be some metal form from the base/leg coming up through the top. I think he's right as that aspect is sort of the point of the whole thing. My wife thinks there are too many materials involved and it doesn't really come together - she's right too. Additionally I think proportions of the top to the base are off, I probably should have gone with a circular top. Although I largely managed to solve it, there is tendency for the top to be able to spin on the base as the compression collars just don't provide enough force to resist that.
After rebuilding the VH-65 and mounting it, I kept my eye on the oil level. After a few weeks it fell below the sight glass and I knew I had a leak. I had really fought with the large seal in front and immediately came to the conclusion I had messed it up. Careful observation seemed to show that it was leaking as there was evidence on the front of the rotary.
It's kinda hard to see, but there is some faint staining right below the seal area as shown here.
First up - remove the old seal. Easy - just drill a small hole in the seal and use a slide puller!
My problem the first time was I had no way to evenly pull in the seal. I had tried to gently hammer it in, but I knew at the time I had deformed it. I needed a tool.
When I was scrounging around my local surplus yard, I found a scrap ring of 6.5" ID aluminum. Perfect. For $10 it was mine. All it needed was to a bit of machining to open up the ID a bit and face it. I used the T slots in the face and a couple of scrap plates to rig up a way to pull it down. I carefully tightened each nut so as to install the seal square and even. It went in pretty easy.
I topped it up with oil and I'll keep any eye on it for the next few weeks a see what happens.
When I had purchased the rotary, the Fadal logo plate was missing. One happened to come up on eBay, so I decided - what the hell.
Since I got my '95 Fadal VMC20 a few years ago, I've always wanted a probing system. Not only does it help you stay accurate and reduce work, but you can insert probing routines in the programming. I wanted it all, a tool setter and an optical probe to use while machining. I hunted on eBay for a couple of years (it takes a while to find good/inexpensive examples) to finally assemble the following Renishaw Parts:
Next I had to figure out the wiring and how I was going physically mount everything. The first step was to create a diagram of how everything needs to be connected.
The above layout should work like this:
1) If you want to use the Tool Setter, you send an M64. This will cause the MI8-4 to select the Tool Setter as an input and output it's status to the 1060 board. If you send an M65, it will select the Inspection probe (OMM) for the input.
2) If you select the OMM, then you need to also send an M66 to actually start the probe. In all of the Fadal wiring docs I could find, they never connect the Error and Battery status to the control. If you use them, you can verify the probe started OK is working before you initiate probing. I've written the gcode to test those after a start is issued to verify the probe stated (i.e. go from error to OK) and to make sure the battery isn't failing. If there's a problem the program will loop to try starting again or allow you to abort.
In order to enable the M64 and M66 gcodes you'll need to populate the 1100 board with 2 SSR's (Solid State Relays) and 2 fuses. I used a Grayhill 70S2-01-A-03-A and a 1A fuse. Populate K31/F40 and K16/F10 on the 1100 board. These SSR's will be switching power supplied by the 24VDC power supply.
Here's a shot of the SSR's mounted and wired. The red wires are from the 24V PSU. The Green wire is for the M64 signal and the White wire is for the M66 signal.
In order to detect the Error and Battery status you'll need to create and adaptor for the 1040 board. This has a 26 pin edge connector that you need to interface with. By referencing the Fadal User Manual -> Macros -> Layout of I Macro, you can see that I(3) is pin 19 on J2 and I(4) is pin 20 on J2. You can test is the pin is logic high or low and create logic from there. I assume this card requires TTL voltage levels, so use 5V only!
I decided to mount the MI8-4, the 24VDC PSU, and the connector block to the inside of the electrical cabinet. Luckily there were two 1/4x20 studs sticking out that I could conveniently mount a DIN rail to. I also made up a small connector block out of some perf board to handle wiring up the various components. I 3D printed a couple of DIN rail adaptors for the backside. The 1040 board uses approx. 10K pullup resistors on the input pins. I set things up so an Error or Low Battery are high (active - i.e. the default state) and you need to pull them to GND to indicate a no Error condition. I added some caps to help with AC ripple, as it seemed the power from the Fadal was kinda noisy.
And here's the PSU, Connector Block and MI8-4 mounted and connected. All in all I'm happy with the layout.
I mounted the TS27R on the back left side of my table and ran the wiring up and over to the cabinet. Initially I was going to mount the OMM on the top left of the cabinet as indicated in the Fadal Maintenance manual, but looking at a bunch of pictures on the internet of HAAS and Fadal OMM installs - it looked like the back of the cabinet would be better. I mounted the MI12 on top of the pendent and ran the wiring under the cover that runs along the top of the cabinet. I had to create a hole to run the wires through on the pendent end. (The black box to the right of the MI12 is my interface for a programmable coolant nozzle I'm working on - more on that in a future post.)
I calibrated the TS27R (basically getting the stylus flat) and then determining the fixture offset. After that I tried a tool setting cycle and low and behold - it worked! Can't wait to machine something with my new capabilities.
Today I started in on the wheels. They're pretty ugly.
I need to determine hub offset. The front was easy as the hub is symmetrical and it's centered on the rim so I didn't need to measure. The rear is also apparently centered, but the hub is asymmetrical, so I decided to measure where it sits. I just laid a straight edge across the hub flange and measured the distance to the rim. Looks like about 3.5/64 or 1.4mm.
The nipples were pretty badly corroded, so I just used the Sawzall to liberate the hubs.
Ever growing pile of cleaned up parts.
I discovered the rims are nice aluminum ones. They look like they'll clean up nicely, so I might just polish them instead of painting. I also found the u-joint was a bit notchy - but I've decided to throw it back in as there is no easy way to replace it, BMW wants you to just get a buy a new drive shaft.
If you recall, when I drained the oil out of the final drive I found a bits on the magnetic drain and a chunk of metal that was a wee bit concerning:
After reading up on airhead final drives, I realized I could remove the inner cover without disturbing the gear setup. (That anti-seize lube crap is a bitch to clean out of the splines!)
Well I found the rest of it and I was able to put Humpty Dumpty back together again - sorta. I couldn't find a parts diagram for the final drive, so I have no idea what this is, but it looks like it's supposed be a trust washer that is located at the end of the needle bearing in the center of the drive.
You can see here there is metal smeared on the shoulder of the flange and there is some blueing from overheating visible too.
So this drive is toast. Too bad as the bevel gears actually look very good. This was a 32/11 drive with a ratio of 2.91. It's one of the tallest rears BMW made (I guess the R100S was supposed to be a powerful long range highway cruiser.) I've decided to replace it with a used 32/10 (3.20 ratio) unit off a '73 I found on eBay. This will screw up the speedo drive ratio, but who cares? Bonus is the brake pads are included - I hope they're good enough to use.
My sister-in-law found a rolling pin at a Salvation Army about 30 years ago - spending the outrageous sum of 50 cents on it. It's grown to be a coveted possession and is critically integral to our Thanksgiving Day feast. It's a key tool for the creation of her magic Pecan Pies. This year, right in the middle of major heavy pie madness, the ol' girl gave out. The handle pin broke at the at the roller. We managed to get through the rest of the pie build, but a fix was needed.
This thing was old when Howie found it, so I'm sure it's seen it's share of battle pie. From the looks of the remaining pieces, the pin was actually pressed into the end where there are 4 pieces of wood set like wedges to hold the pin. Although that's what it looked like, I wasn't sure and I was unable to do much with it anyway.
I decided to try to get the broken part out first. It measured about 0.5", so I rigged up a way to hold it on my drill press and went at a bit undersize with a 29/64. Just as I was getting to the final depth, VOILA! the broken piece just came out!
- I tried to get the 4 "wedges" out, but got no where. I could have simply made a new wooden pin, but I knew I couldn't match the end of thing. So I decided to create a hidden fix that would be strong enough to stand up to another 30 years of pie making.
I decided to use a couple of pieces of stainless, one - a shaft to go in the end of the rolling pin and stick out. Then another, thin walled sleeve piece to slip over that and provide the a place to insert the end cap.
I ordered a couple of bits of 304 Stainless from McMaster-Carr that was close to the sizes I needed. I had to turn a bit off the OD of the sleeve and bore it to 0.510. After a bit of finishing and cutting to size, I had the pieces.
I decided to simply glue the thing together. I used some JB Weld to join the sleeve to the shaft. I then used some epoxy to glue the shaft into the rolling pin.
Once the glue had set up I cut the wooden end cap down to about an 1" and then pressed it into the end.
Looks to me like it's ready for battle!
I've been wanting a tool setting probe for my mill for a while now. It makes changing tooling a no brainer and also allows for high accuracy. These things are expensive however, so I watch eBay. After over a year of waiting, one came up for sale for $80. I went for it in spite of 3 strikes against the eBay add -
It was covered in a rusty dust and had a very funky protective flexible conduit attached. I removed that and washed the whole unit. I inspected the flex seal in the nose and everything looked really good.
The next thing to do was to test the contacts - I hooked up the DVM and tested the Red/Blue pair. Everything looks fine.
Looks like I got myself a probe.
Now all I need to do is source a waterproof conduit and the fittings to seal to the probe, mount the probe, run the conduit, mount the interface card, figure out where to add the probe signals to the machine I/O and calibrate and test it. It'll take weeks, but I'm pretty interested in getting it in.
I've been going through the rest of the bike now that the engine is sorted. Making some progress in cleaning stuff up.
Not so great news in the form of lots of metal fragments stuck to the magnet when I drained the transmission oil. I didn't take a picture, but it was more crap than I'd like to see. Nothing big or or chunky, mostly dust like with thin flakes. Probably the gear faces.
The final drive was worse. Way worse. Big chunks of metal came out, including something rather significant. This really should be rebuilt, but not by me.
It appears Uncle Ben was not too up on maintenance. The engine, trans,final drive and tires all showed signs of lack of attention. I expect this bike was just never serviced and ridden hard.
Got the new rod bearings. When I was cleaning the rods, I noticed some bluing on the big ends at roughly 12 and 6 o'clock. Musta been bad in there at one point. I'm completely ignoring any of this sort of thing...
Pretty close to being complete!
In keeping with my strategy of getting the engine squared away before any other work, I finished the engine and rigged up a test stand.
So it seems to run well enough, no weird noises, leaks or other red flags. That gives me the go ahead to start working on the rest of the bike.