Bryan's Fairchild 539A lathe project

[EpicenterBryan]

Amazing! So inspiring! Respect to the mad skills

Yes, you are Bryan, its all very EPIC!

This is awesome!

Thanks for the nice comments folks, but I've run into some snags that I'll go over in the next few days, and the fixes... For now, waiting for some parts.

B

[EpicenterBryan]

Hey guys,
Not much time tonight but here is one problem I figured out and fixed today.

This issue related to when the servo motor 36V power supply and the controller that has the frequency generator were both powered down at the same time. The 5V supply for the controller would die before the 36V supply for the servo. That would cause the AD9833 output frequency to stop and the servo would try to instantly stop as well. Not Good.

I hadn't implemented a servo enable control because the polarity of the Optocoupler input on the servo was inverted. If nothing was connected to that input, the drive was enabled. I guess it makes sense for most people so that the drive would work if that input wasn't connected.

Anyway, what I wanted to do was to have the servo enabled only if the output from the AD9833 was valid. The easy thing to do was to enable the servo if the 5V supply for the frequency generator was on.

I found by using their software and by connecting to the servo via RS232, that I could invert the polarity of the enable input in software.

servo enable polarity.jpg

The only problem - although making the change did actually work, it was volatile and was lost when power was cycled. I knew I could read all the settings from another software function, but nothing could be edited on that screen. There was a download (really upload to servo) function on the same page but since you couldn't edit anything on that page I didn't get it...

Servo EEPROM Write.jpg

... until I figured it out.
1) Make the change on the first image
2) Read settings from the second image
3) Download (to servo) from the second image.

Once I did that, the setting "stuck" and that problem was fixed.

Bryan

[EpicenterBryan]

This post got messed up... See the next one.

[EpicenterBryan]

Quick update:
Still working out a mechanical issue coupling the servo to the transmission.
More on that soon.

But the servo speed control has been incorporated on the touch screen.
The way it works is like this:
1). Power up - Servo is disabled, speed is zero. Speed is in Red (shutdown) with no background.
2). Pressing speed the first time starts the servo and runs up from zero to 1/2 speed for 33 1/3 (16.67) - but displays as 16 rpm for space issues. Notice the background color and text color change.
3). The next press initiates 22.5rpm (1/2 speed for 45) but displays as 22. Also notice background and text color change indicating 1/2 speed.
4). Pressing again go to 33 1/3 rpm (displayed as 33). Background goes away. Text color changes. Simple display.
5). The next press goes to 45 rpm, same simple back ground since 33 and 45 will be used the most. But 45 is highlighted with a new text color change.
6). The final press ramps speed back down to zero, changes text color to RED and disables the servo.

Here is what things look like as of tonight:

Speed Screens.jpg

I wanted the two 1/2 speed options to stand out and that's why there is a background color change using colors not used elsewhere.

At some point, I may make the first button press default to go to 33 rpm, and the rest of the pattern would continue from there.

That's it for tonight. Hopefully I will have more to report on motor coupling in the next few days...

B

[EpicenterBryan]

More details about the coupler issue to come.

But here is a teaser for a new post relating to an inexpensive Microscope I found and have mounted to the Fairchild.

Drool for now.... Wow! And inexpensive.

IMG_6180.JPG

B

[EpicenterBryan]

Hey guys,
Here are some details about that little microscope and how I set it up on the Fairchild.
First off, the part number is G600. I bought mine from Tmart, but you can find them on E-bay, Alibaba and even Amazon. I paid $40 for mine with shipping included!

It turns out there are a number of similar ones out there, and this may not be the best one. So be sure to watch some reviews on YouTube before you decide.

Here is what you get for $40:

600X-4-3-LCD-USB-Digital-Microscope-Portable-8-LED-3-6MP-VGA-Electronic-HD-Video.jpg

HTB1GXTTPpXXXXbHaXXXq6xXFXXXJ.jpg

There are several draw backs as listed below:
1). There is no standard threaded hole for mounting. So the options are to use their plastic stand that clips into the back, of look for a seller that includes a metal stand that clamps around the lens section.
2). It has an SD slot and takes video, but not still images. I guess that could be done via usb but no software is included.
3). The Video recording function is strange. It can be set to record when there is motion, or for a set period of time.
4). There is no remote. So even if you could click a button to take a still, the microscope jumps around.
5). There is no video output.
6). The focus knob / mechanism is on the cheesy side but functional.

Now for some pluses:
1) The light source can be dimmed, or even turned off.
2). Built in display!!!!
3). Good range of magnification.

As mounted on my lathe, her you can see the range of magnification...

IMG_6183.JPG

IMG_6182.JPG

As you can see when really zoomed out and my linear stage is near the top, the LED lighting is not usable. You see the light pattern from each single LED. An external wider pattern light is necessary.

Here is how I attached this to my setup:

scope parts.jpg

The linear slide I used is from Edmunds Optics and is crazy expensive ($290), but I bought this one last year on E-bay for around $20 used! You see I used some bar stock to adapt the stage to the scope arm. Then I made a bracket on my 3D printer so a section of the included plastic stand could attach. It is possible to raise and lower the scope by changing the angle of the plastic mount to some extent, so you really don't need a liner stage if you don't mind fussing around a bit. There is a ball at the microscope end that can also adjust the angle of the microscope.

So that's about it. I think this thing is an awesome addition for $40. Well worth it.

B

[SONARC]

Hola! I was still looking around for something like the Spencer microscope, and now you put this in front of me...
How did I not stumble across something like this in my own research, I wonder?
Always impressed by the overall scope of your projects, even though admittedly, much of the math and science tends to be way over my head...
No matter how many manuals and technical papers I read, it seems that the only way I learn any of this stuff is through one-on-one osmosis.

[EpicenterBryan]

Always impressed by the overall scope of your projects...

Thanks! Glad you enjoy!

------------------------------------------------------
Here is an update on July 4th:

I've been messing around for a month or more it seems with this coupler issue for the servo.
As I'm sure I've mentioned before, the servo motor is 8mm and the transmission is 9.525mm (3/8"). That is a coupler combination that is nearly impossible to come up with and I've needed to mess around with making my own. One coupler I was able to get with that bore combination had a hard insert between the two half's and the plan was to use the 1" ID latex tubing I had used before to connect the half's and provide flex. However, the diameter of this coupler pair was 1.2" so it was very hard to manually stretch the tubing over the couplers before gluing them. After doing this several times using my alignment shaft (8mm on one end, 3/8" on the other) it became clear that I just couldn't get the latex tension even. So once the completed coupler was taken off the jig the shaft holes were not aligned exactly and it caused vibration when installed.

There are several ways shafts can be misaligned and I've run into every combination by now.

fig1.jpg

Assume that I've had the transmission and servo shafts perfectly aligned on the lathe like shown in the upper left of figure 1. Since the servo is suspended, it can move around slightly when the coupler shaft screws are tightened and if the coupler is not perfect it's really impossible to know without running the motor or manually rotating the transmission and using a dial indicator on the servo suspension. So I started checking the couplers with two shaft segments installed and chucking the coupler in a drill and inspecting for shaft movements in the final shaft segment. Doing this, I was able to spot obvious issues before installing the coupler, and that has helped of late since every time I change couplers it changes how much counter weight is needed to keep the servo motor CG centered between the suspension springs.

Recently I tried a new way to install the 1" ID tubing over the 1.2" OD couplers. The idea was to lubricate the tubing by installing the latex under water. It worked like a dream! But then, how to dry the inside of the coupler so it could be glued?

The solution was pretty clever and it actually worked!
I put the coupler and alignment shaft in my HarvestRight Home Freeze Dryer and used it as a vacuum chamber.

The trick is that water boils at room temperature under a strong vacuum. In the second photo the water droplets you see are actually boiling! Within a few minutes all the water had boiled out of the coupler and it was ready to glue!

IMG_6167.JPG

IMG_6168.JPG

That worked pretty well, but after removing the alignment shaft and attaching two independent shaft segments and spinning it with a drill, it was clear the shafts were not aligned perfectly.

So I tried something else. The black coupler was made by installing the coupler half's on the alignment shaft and slipping some heat shrink tubing over it. Then it was spun with a drill while heating the tubing. After cooling, a second layer was added and the result was checked. Damn, that worked well. However, once installed and after running from stop to 16 2/3rpm, to 22.5rpm, to 33 1/3 rpm and on up to 45 rpm, the coupler on the servo side slipped / misaligned and all hell broke loose. It looks like my acceleration in speed change was a bit much. But it showed this could work.

IMG_6198.JPG

Since then, I tried using a more expensive stick of heat shrink that has a built in heat activated glue. This stick was $18 compared to the stuff I tried before (which cost under $5). It held together without changing the acceleration, and although I used a coupler end that I knew had a slight bore offset it did show that this tubing works and I will try this again with another 1/2 coupler that is centered better.

IMG_6199.JPG

But this brings up another coupler that I tried. Looking at the photo with the black coupler - in the upper left is a stainless steel flexible coupler that I found on Alibaba. That one seems to be only misaligned in an angular sense and I'm going to try to correct that. I'll chuck it into my mill with two shaft segments and will use a dial indicator to measure the far end shaft deflection. I'll spin it by hand and I'll try to over-extend it in the opposite direction of maximum deflection until I get it totally centered. Since it's stainless and has so much less flex than Latex tubing or an aluminum coupler, it wants to move the servo around rather than flex like it needs to.

Anyway, more on this stuff - other things I've tried - and what I have on order that may eliminate all this messing around in the near future.

For now, have a safe and enjoyable Independence Day everyone!
Got to head home and get those ribs on the grill!

Bryan

[EpicenterBryan]

Another update - This time unrelated to couplers for a change.

One thing I had been planning to do was to have a master power on switch for everything in the console. I also wanted to have a switch for powering the vacuum pump and the stylus heater. I had two existing holes to work with on the lathe that were left over from the old power switch and fuse, but really didn't want AC up there, so the plan was to wire up some relays and outlets in the console.

I ordered a bunch of modern looking switches with LED rings in various colors, and also some Capacitance Touch Switch Modules in a bunch of colors.

I ended up loving the Touch modules so let me tell you about them.
They are made by Heltec Automation and look like this:

41f4bk1ZizL.01_SL400_.jpg

51D4teq3NDL.01_SL500_.jpg

As shipped, they act as latching type switches but that can be changed if needed. All you do is give them 5V and ground, and you get a 5V signal out when the switch is on. The internal controller runs the LED and you have no control over that externally though. You don't really need to touch these switches, just get your finger close and they turn on.

I printed up a bezel that allowed me to mount two of them in only one hole. The print job was a bit sloppy but I can live with it for now.

IMG_6205.JPG

SEXY!

IMG_6206.JPG

Blue is the master power on, and Green is for vacuum and stylus heater. One thing I might do is add a stick on letter on each button - something like P (for power) and H (for heater). Not sure yet. Anyone have inputs? "O" for On? "S" for Stylus?

Anyway, I found an awesome relay box with electrical outlets that can be controlled with a wide range of signals. Check this out! It's called an IoT Power Relay - made by Digital Loggers. It can be controlled via 3-60VDC, or 12-120VAC on the green connector. It has one outlet that is always on, one outlet that is normally on, and two outlets that are normally off.

iot.jpg

Since my touch panel switches need 5 volts all the time to be able to turn one the relay box, I plugged a USB charger into the "always on" outlet and ran that power to the switches, and ran the output signal from the switch back to the control input on the relay box. Then connected a power strip for the entire console into one of the normally off outlets. Now, one button powers everything up, and the "Heater" button will run another one of these just boxes for the stylus heater power supply and vacuum. And no AC wiring to mess around with!

Anyway, thought you guys might find a use for one of these relay boxes. Not having to mess around wiring AC makes the $30 price tag so worth it!

Bryan

[EpicenterBryan]

Tonight, I can report that my last attempt at a DIY coupler just passed with flying colors (with some help).

IMG_6203.JPG

I ended up manually adding a little drag on the transmission flywheel as I ran through all the speed changes and by the time I let off at 45rpm, everything was smooth as a baby's bottom. Then I ran through all the speed changes up and down. Awesome.

Although all speeds (16 2/3, 22.5, 33 1/3, 45 and 78) are all on the money, and vibration is gone, I'm still not thrilled that it took goofy last minute "drag induced stress" to get the vibration to go away. What if I ever need to move this? Do I need to go through a crazy "drag the servo" routine to tweak the coupler into the correct zero shaft angular alignment? I'm going to live with this for now - but this is not right.

Fortunately, as I have mentioned, I have a bunch of things on order. And I still have not tried to straighten the stainless coupler and will still try that.

Of things I mentioned I have on order, I found a direct drive coupler that goes from 8MM to 9.525mm (3/8") like below:

41WF5LAsW4L.jpg

I'm not happy about direct driving the servo to the transmission but there is plenty of give in the servo mount so if this ends up working, so be it. I have also looked at some crazy magnetic coupling options but haven't been able to get any "ball park prices'... so I kind of gave up on those as viable options. They all seem to be custom made. Check out the three most common types on-line. Sorry - I don't have an unlimited budget.

The next option is to go with a belt drive system. Since the Fairchild has a 54:1 gear box, there is practically no torque needed between the servo and the transmission. However, any belt drive (and I plan on a 1:1) requires the belts to be exactly matched and the center Bore of both pulleys to be centered...

Here is what I have on order from the same supplier.

pully.jpg

Like usual, no one makes both metric and US shaft bores. So, I ordered three of these in 8MM, so my machinist has 2 chances to bore one out at 3/8", and I also ordered one at 5mm, and one at 6mm so I could use either for belt tensioners.

That's where I am tonight.
Direct drive or not? Not sure.
Will Belt Drive work? Not sure.

Other comments?

B

[EpicenterBryan]

Hey guys,
The pulleys arrived Saturday but sadly, neither of my machinist friends were at the shops - but I did get a bunch of stuff done.

First, I did try out the fixed shaft coupler, and also straightened up the heavy stainless steel flex coupler. I chucked that in my mill and manually rotated the mill with a dial indicator set up. Then found the spot where it was misaligned, and over extended it in the opposite direction. It only took a few minutes to get in right on the money.

IMG_6209.JPG

But since it's so heavy, it kept messing up my shaft alignment by moving the CG of the servo this changing my shaft adjustments. I ended up deciding to just wait until the pulleys arrived.

That brings me to to the next topic. I decided to make an isolation platform for the servo. After all this work, I don't want any chance of vibration getting into the lathe. So, after doing a bunch of research for something other than rubber issolators, I kept seeing wire rope issolators. It turns out a lot of people use them in camera mounts. Anyway, once I saw a few videos on how well they work I decided to give them a try, and make some myself.

First, here is a video that will give you a good idea of how effective they are. In this video two containers are on a vibration table. One has wire rope issolators, one doesn't...

I made the issolators with a 12 circuit terminal block I found in the electrical department at Lowe's, cutting them in 2 circuit segments. I tried 2 sizes of aircraft wire rope (0.063" and 0.094") and decided the thicker rope was a better fit for the weight of the servo.

The aluminum extrusion I used was OpenBeam, and the right angle corner brackets were also OpenBeam. The other plastic parts were 3D printed. I messed up the design on those, but trimming with a pair of dikes worked. The right angle bracket came from Ebay, and the top plate was from OpenBuilds.

IMG_6213.JPG

IMG_6214.JPG

IMG_6215.JPG

IMG_6216.JPG

IMG_6221.JPG

Now comes the funny part.
My 36 volt power supply has a fan that increases speed as more current is used, and the &#%$@ fan vibrates! So, I'll be upgrading the fan for sure. China!

B

[EpicenterBryan]

Hey guys, just about done!
Tonight I'll share some more trials and tribulations before the final solution!

When we left off, I was messing with the wire rope isolated platform.

I tried a round belt and found there was enough tension needed that the top platform would rotate since the pulley was not at the center of the platform. This caused the two pulleys to go out of alignment and caused the belt to wear. Then it started making a loud squeaking noise. Not shown, I also tried a synchronous belt which didn't need much tension but would have taken a much more refined speed up and down ramp, and at the highest RPM made too much belt noise.

You know, if I moved the motor so the belt was at the center, and if I added counter weight to compensate the other end of the top platform (because of moving the CG of the motor), perhaps it would have worked out better. Some times there are better solutions, so I pressed on.

IMG_6231.JPG

Next up, I tried a fixed mount (but with an isolator between the servo and mounting bracket). Remember, the transmission shaft on the Fairchild is offset by about 9 degrees, so this was configured at that angle. This was bolted down to a shelf below the lathe inside the console. By loosening 4 bolts, the motor could shift forward or back. By loosening 4 other bolts, the motor could move left or right. Or, by loosening all 8 bolts, the angle, forward/ backward, left/right could all be adjusted. This turned out to be too damn tricky to adjust, transferred too much vibration into the wood shelf, and the flexible motor isolation gizmo was again too flexible to tension the belt without changing belt alignment. These 8 bolts were accessible via the 4 black corner connectors shown... Look for them. Each has 2 bolts connecting one aluminum extrusion to the other.

IMG_6264.JPG

I don't have time tonight to complete the entire post, but let me show you the guts of the final version. The critical and clever final bits will have to wait until the weekend.

I went back to the original bracket that hung the old NYC-34 Bodine motor on springs, and re-purposed it. Way back, I drilled 4 holes to mount the servo bracket. I reused those. As it turned out, the bolt pattern was not exactly in the center and the last time I used this bracket I corrected for that with a sub plate... Details don't matter now.

servo left-right.jpg

What I did was to mount the servo bracket via rubber isolators to a segment of v-slot aluminum extrusion. Then the extrusion was mounted with T-slot bolts to the u-shaped motor support bracket. All that was aligned so the motor shaft was at 90 degrees to the bracket. Then I positioned two right angle brackets and drilled holes to mount the brackets to the u-shaped motor support bracket. Notice the right angle brackets and attached to the motor support but not the aluminum extrusion. That way, the brackets act as a guide so the motor can slide left and right and stay perpendicular to the u-shaped motor support bracket. That's really important.

That's all the time for tonight...

B

[EpicenterBryan]

Here we go with the good stuff.

As you saw, the U-shaped motor bracket has 4 big holes. Originally, 4 bolts and spacers went through there but did not touch the bracket. They were there to limit movement of the Bodine motor when transporting the suitcase (the lathe was originally mounted in one). So I used those, and the 4 holes in the brackets under the lathe.

This is where it gets good.
I made a set of eccentric nuts / spacers. A 22mm wrench fits these. The round section goes inside the big holes on the bracket. Notice there is a 5mm hole that is offset by 3mm from the center of the nut. It's hard to see in this picture, but there is a notch on the nut flat nearest the 5mm hole. That is a visual indicator for where the hole is closest to the edge of the nut.

IMG_6349.JPG

To explain how these nuts work, let me put up a batch of photos.
This is what I call the baseline. All four nuts are positioned in the same orientation as shown.
None of these are to scale or represent the actual shape of the u-shaped bracket.

eccentric baseline.jpg

If all 4 nuts are turned Clockwise by 90 degrees as shown, the motor bracket moves 3mm closer to the front of the lathe. As a side effect the motor also moves 3mm closer to the transmission. However, the motor can be moved away by adjusting the aluminum extrusion.

eccentric cw90.jpg

If the nuts are rotated counter clockwise by 90 degrees, the bracket shifts toward the back of the lathe by 3mm. Again the motor moves closer to the transmission as a byproduct and can be adjusted back with the v-slot.

eccentric ccw90.jpg

And the combination of adjusting two nuts differently than the other two can twist the bracket to correct angular error like in these pictures:

rotate left.jpg

rotate right.jpg

Any adjustment less that 90 degrees, results in less change but I wanted to show the extreme case.
The nuts also acted as spacers to set the motor shaft at the correct height. I printed 3 sets, where the top thickness changed in 0.5mm increments. The idea was to insert the set that was the closest, and add how many 0.1mm shims needed on top. In my situation one of my sets was correct so no shims needed.

Here you can see the 4 eccentric nuts, and 8 vibration isolators.

Eccentric and Isolators.jpg

I did a quick check using a direct 3/8" to 8mm coupler, some 8mm extension shaft and Chinese helical coupler.
Wow, the least vibration I've had to date. Still a small amount of vibration in the 45rpm setting but very good.

Then I tried this and ALL vibration is gone!!!!!!
It's basically a drive shaft like what's in your car expect this isn't telescopic.

Driveshaft parts.jpg

I've got a few other parts on order that I hope are a bit higher quality but this totally works!
And why?

The trick is that there are two places where angular misalignment is dealt with. All the couplers I've tried work with angular error, but can only deal with extremely small horizontal or vertical shaft misalignment even if the shafts are parallel. So, imagine if there is no angular misalignment, but there is vertical or horizontal error. Using 2 couplers like shown below fixes this. It converts the horizontal or vertical misalignment into two angular errors. In this example the vertical shaft misalignment is converted into a positive angular error by the first u-joint, and into a negative angular error by the second u-joint.

IMG_6350.JPG

So, if I had installed two helical couplers like shown below on day one, I probably wouldn't have had any issues at all!
Duh!
This has 3 shaft segments. This is not just one shaft!

IMG_6356.JPG

Other stuff I learned.
1) The quality of Chinese couplers and pulleys is all over the map. I've received some where the holes and not centered, and have received some where the hole diameter is a bit too big so there is no chance to get the shaft centered using the set screws (especially if they are not on opposite sides). I had similar problems with clamp style and set screw style.
2) I totally underestimated the amount of tension needed for a round belt where some torque is needed. I thought that configuration would be the most forgiving. It was not.
3) If at all possible, don't mix metric and English. Since the motor was only available in metric, the first thing I should have done was use an adapter on the transmission and convert it to mm. You can find anything you want in metric, not so much in inches and very few things with metric on one end and inches on the other.
4) Don't keep trying the same thing but don't give up. Let it sit until you come up with another idea.

Hope that helps people some day.

B

[EpicenterBryan]

Time for an interesting update via YouTube:

[jjwharris]

Amazing stuff again Bryan!