My current project

JimDawson

Global Moderator
Staff member
H-M Lifetime Diamond Member
Joined
Feb 8, 2014
Messages
11,144
I can only show bits & pieces of this one. I can't show the proprietary parts or discuss the purpose. But there will be some interesting machining and work holding solutions for the parts I can show.

The pallet full of stuff that is going into this project
1643174926278.png


This is the part I've been working on for the last couple of days. A spacer to mount a hollow shaft gearbox to a machine frame. OK, maybe more like a day in real time, but I'm old & worn out and can only work in the shop 4 or 5 hours a day, and then not too hard. I'm glad I have a CNC machine, it does most of the work. If I had to turn cranks for 8 hours, I couldn't do it. :faint:

1643170776538.png

It goes in this assembly. The motor mount and motor are a bit unconventional in this application. I'm using a 750W servo motor (basically a NEMA 34 size, tiny little thing) to drive the gearbox rather than a conventional NEMA 56C frame motor that it is designed for. The little servo motor and drive was actually less expensive than the 56C wash down motor and VFD that I would have used, and has much better torque characteristics than a standard 3 phase motor, and it's wash down rated. Using a hydraulic pump/motor adapter, installed backwards, to connect the motor to the gearbox. Going to need to modify the adapter a bit also. But for the $30 the adapter cost, I couldn't build one. Just the material for it would have cost that much or more.

1643173205985.png


1643170867486.png


The work piece started out life as a 6.5 x 6.5 x 12'' chunk of aluminum, then I sawed that down to 5.5 x 5.5 x 3.125''. The local metal supplier had a bunch of these drops on the shelf.

Then the question of how to hang onto it for efficient machining. I need to machine all six sides of the part as well as the ID and I don't want to do multiple setups. So one possibility is to cut the raw stock a bit thick and machine dovetails in both the part and the vice jaws. The dove tails would allow holding the part with minimum grip height and allow work on 5 sides of the part in one setup. Then for the last operation just flip it over and face off to the proper thickness.

A note on vice jaws, I don't even own a set of hard jaws for my Kurt vises, we just make mild steel or aluminum soft jaws as needed. And if they need machining for fixturing a job then you can always make new ones when needed. We usually make a half dozen or so sets at a time. Also much easier on tools than hard jaws when you run that endmill or drill bit into them.

The alternative is to use a fixture plate that you could bolt to the table or hold in the vise, which is what I did in this case. But I still cut a dovetail on the vice jaws, the jaws in this case are mild steel.

1643172400376.png

Using this 14° dovetail router bit. I wanted the sharp edge to catch the fixture plate so it could not pull out of the vise jaws. I could have cut a mating dovetail on the plate also, but I really didn't think I would need it. The fixture plate is a piece of 3/8 thick aluminum.
1643172498758.png

First I drilled & tapped the back side of the work piece and also drilled a couple of dowel pin holes, I reamed these at 0.251'' for a slip fit on the dowel pins.
1643172900975.png

Then put the fixture plate in the vise and drilled the mounting holes, and reamed the dowel pin holes at 0.2495'' for a tight fit on the dowel pins. The work X0, Y0 was also established at the upper right corner in this operation and will remain there until the job is complete. I was a bit worried about the 3/8 plate bowing from vice pressure, then I realized it is not going anywhere when I screw that 3 inch thick chunk to it.
1643173120563.png

Then just mount the work onto the fixture plate, locating on the dowel pins and bolt in from the bottom. Then drill & tap the mounting holes on the top side.

1643174112322.png

That end mill is a bit long, I don't like that much stick out, but I need to reach 3 inches deep, and it's the only long end mill around that size that I had. A 4 flute is not the best for aluminum, a 3 flute aluminum cutting carbide would have been better. Ya gotta do what ya gotta do. o_O

1643174266238.png

And making chips
1643174560603.png

And finished, well almost.......
1643175721575.png

I measured the bores and was way undersize.... HUH? :confused 3: Like about 50 thou undersize, and the outside is oversize by the same amount. OK, go back and check the drawing..... Nope, drawing is correct. Look at the Stock to Leave in the CAM setup.... Nope, it's correctly set at zero. The only thing left is the 1/2'' end mill is undersize. YUP, only it's not a 1/2'' end mill, it's actually 12mm (~0.475''). I had no idea I had one of those. :rolleyes: I normally check the endmill for actual size when setting up the CAM, but this one was new out of the box and looked to be high quality so I just expected it to be on size. No big problem, back into the CAM setup and plug in the actual size and run the finishing again passes tomorrow. At least they are under size, I can fix that pretty easy. Had they been over size I would have been building a new one.
 
Last edited:
Delta servo?

I recently also used AC servos in an application that would be considered "overkill." Because I needed a lot of torque in a small package. Basically a 1kW deer feeder. Price of servos has come down to where this isn't completely stupid. I used DMM servos and I don't recommend them; at least not for analog input speed mode applications. They seem to work just fine in the configuration everyone typically uses them in (pulse/dir, position mode) but I can't weigh in on that.
 
Jim,

Thanks for sharing whatever you can on this one.

I always pickup a few nuggets of wisdom from your posts, and this one is no different..... I have already seen a couple new tricks.

Watching.

Thanks again,
Brian
 
Nice work as always Jim . :encourage: I have a small similar project I need to make if I ever get motivated enough to start it . :big grin:
 
Delta servo?

I recently also used AC servos in an application that would be considered "overkill." Because I needed a lot of torque in a small package. Basically a 1kW deer feeder. Price of servos has come down to where this isn't completely stupid. I used DMM servos and I don't recommend them; at least not for analog input speed mode applications. They seem to work just fine in the configuration everyone typically uses them in (pulse/dir, position mode) but I can't weigh in on that.

In this case these are LS servos from Automation Direct https://www.automationdirect.com/selectors/ls-servo. A new product line for them so we'll see how they work out. In this case, I will be speed controlling these with a pot on the electrical panel, no positioning required, much like your deer feeder. Automation Direct also sells Delta servos, the SureServo2 product line. https://www.automationdirect.com/selectors/sureservo2

I have both Delta (7.5kW) and DMM (1.8kW) servos on my lathe. I do have to say that the Delta servo seems to be much more responsive than the DMMs. I'm operating the Delta in analog speed mode with positioning (C axis) for my spindle, quite easy to set up and get it to do what I wanted. The DMMs are operating in analog torque mode, and I did have a bit of trouble getting them set up tight enough to suit me. But I really can't complain, the machine holds +/- 0.0002'' all day long. I have used DMM servos on other projects and had the same problem in torque mode, and they are a little fiddley in pulse/direction mode also. I need to get a DMM on the bench and have the time to play with it and understand how they really work and how to properly set them up.
 
In this case these are LS servos from Automation Direct https://www.automationdirect.com/selectors/ls-servo. A new product line for them so we'll see how they work out. In this case, I will be speed controlling these with a pot on the electrical panel, no positioning required, much like your deer feeder. Automation Direct also sells Delta servos, the SureServo2 product line. https://www.automationdirect.com/selectors/sureservo2

I have both Delta (7.5kW) and DMM (1.8kW) servos on my lathe. I do have to say that the Delta servo seems to be much more responsive than the DMMs. I'm operating the Delta in analog speed mode with positioning (C axis) for my spindle, quite easy to set up and get it to do what I wanted. The DMMs are operating in analog torque mode, and I did have a bit of trouble getting them set up tight enough to suit me. But I really can't complain, the machine holds +/- 0.0002'' all day long. I have used DMM servos on other projects and had the same problem in torque mode, and they are a little fiddley in pulse/direction mode also. I need to get a DMM on the bench and have the time to play with it and understand how they really work and how to properly set them up.
IMO the main weak link for the DMM is the PC software. I could expound but I'm sure I don't need to. Reading through the RS485 section of the manual makes me think the drives are capable of a lot more than the software makes available to you.

Please let me/us know how the commissioning of the new LS servos goes compared to what's been discussed so far. I have been eyeballing those pretty hard but I don't have a good (enough) excuse to buy one yet.
 
OK moving right along...

I corrected my little oops above and finished the bores to the correct size. I had to deliver the gearbox and spacer to my customer today to facilitate ongoing assembly on their end. Everything fit as planned. They don't need the motor yet, which is good since I won't have a control panel built to run it for a week anyway.

But I still need to finish the motor mount. So the project for morning, or when ever I get out to the shop..... I need to make an alignment adapter (the red washer looking thing), modify the motor adapter (green thing), and make the 5/8'' stub shaft (blue thing) to mate with the input socket in the gearbox.

So why am I going through all of this just to hook the motor to the gear box? Glad you asked..... The gearbox is designed to mate up to a standard 56C frame motor, 5/8'' shaft size and has the proper mounting alignment flange protrusion that mates with the machine relief in the gearbox flange. The hydraulic pump motor adaptor that I am using is also designed to mate with a 56C frame motor, so it also has the machined alignment relief. The servo motor is being mounted where the hydraulic pump would normally go. I need something to go between the gear box and the motor adapter for alignment, thus the alignment washer.

But that still doesn't explain why all the hardware just to get the motor hooked to the gear box. The 80mm frame servo motor has a whole different mounting flange, much smaller than the 56C flange. It also has a 19mm (~3/4'') shaft, and this is the reason I couldn't just build a simple adapter plate to bolt the motor right up to the gear box. There is no practical way of turning the motor shaft down to 5/8'' and boring the gear box input socket is out of the question. So I need a coupling in the middle, one end 5/8'' and the other 19mm, an off the shelf item. And to have room to stuff that hardware in there I need that big hollow motor adapter. And I also need to maintain more or less off the shelf hardware for easy replacement in the future. If the motor or gearbox fails in the future, stock replacements are available overnight, and no modification required.

1643271909603.png

I'll be using the same fixture that I used to machine the spacer, I just need to drill & tap a few more holes in it. And the X0, Y0 location does not move. So the machine is already zeroed and ready to run the program.
1643274532877.png

Once I get the alignment washer chewed out, then I'll mount the motor adapter and make the needed modifications to it, using the alignment washer to locate the motor mount. Only the Z0 location has been moved, X, and Y are still in the same spot.

1643274727218.png

Pictures later.
 
Well it's been an interesting day, what should have taken about 2 hours turned into a all day ordeal. More on this later.

Here is the redrilled & tapped fixture plate from above. I just moved the drawing over 0.6'' to clear the original holes.

1643336827296.png

Then clamp the workpiece to the fixture plate and drill the bolt holes. I just noticed that the lower center hole is not on the center line, weird. :confused 3: I didn't catch that one in the drawing, or when I set up. Not a big problem, this setup is the only place it is used. But kind of indicative of my day so far.
1643336973061.png

I hope there is enough bolts in here to hold the workpiece down :grin::rolleyes: This will make sense later.
1643337352207.png

And completed, about 2 hours later, should have taken about 12 minutes. As you can see the bolts hold the otherwise loose pieces in place after they are cut loose from the actual part. Keeps things from jamming up. One of the reasons it took so long is that about halfway through the outer arc the endmill broke. My mistake, I had the feed and speed set for a 4 flute endmill rather than the 2 flute I was using, so the chip load was twice what it should have been for that particular endmill, and of course it was a bit dull to begin with, a bad combination.

Normally not a big deal, I would normally just install a new endmill, reset Z0, and keep going. But that would be too simple, and not in line with the cascading events of the day. One of the things I needed to do when installing the new end mill was adjust the size in the CAM setup. The original endmill was a 1/4'' carbide router bit, and actually measured 0.246'', pretty common for those. The new endmill was a high quality HHS, and measured 0.250'', thus necessitating the size adjustment in CAM. This is where Fusion360 started acting up. :faint: More on this in a bit.

1643337584870.png

I finally got the piece done and ready for the next operation.
1643338842170.png

The workpiece on the fixture now becomes the alignment fixture for the next operations on the motor mount. The motor mount bolts right down to the drilled & tapped holes that held the scrap piece on the fixture. Preplanning how you are going to hold, locate, and machine the parts saves a lot of time when you are ready to make chips.

So simply by bolting the workpiece down it is aligned and ready to make chips. I used flat head screws in two of the bolt holes so the cone of the flathead would align the workpiece because the holes are drilled as a loose clearance hole, about 0.405'', for a 3/8 bolt.

So load up the program and press Go...... Spindle starts, coolant comes on, table moves to start position..... And I hit the E-stop :eek: The table is about 3/4'' out of the expected start position...... After some investigating I found the X0, Y0 position had changed in the Fusion 360 CAM software. This was again my fault for not checking it properly during setup. No problem, that's a quick fix. Move it back to where it's supposed to be and click OK.... And crash, Fusion 360 kernel error..... o_O So...

1643338985458.png

I have Fusion 360 on my machine computers as well as my desktop. This allows quick CAD/CAM changes at the machine and saves walking the 150 ft. or so to my desk to make changes from my office computer, and also allows me to post the file directly to the working folder with no USB sticks or network drives involved.

So into the house to see if this is a computer issue or what is going on. No problem on my office computer, everything worked properly, but I also tried it on my lathe computer and it crashed doing the same thing. What is common to the lathe and mill computer is that both are running Win7 Pro, and my desktop is running Win 10 Pro. The mill computer is upgrading to Win 10 as I type this. I hope it works OK after the upgrade. I have run my CNC software on Win 10 with no problems, so I hope the antique computer in the mill won't argue with Win 10 too much.

Fusion 360 did an update a couple days ago, and this seems to be when the problems started, they made quit a few changes. There has been a banner on Fusion 360 for about a year that says Win 7 is no longer supported, and Win 10 is recommended. Now it seems that Win10 is required, at least I hope this fixes the problem. I guess it had to happen some time.

I may finish this job tonight or may wait until tomorrow. I may be worn out by the time I get Win 10 running properly.

Just a bit of fun, it's hard to believe that this is a 1 HP (750W) motor, and can out put about 2HP intermittently. The max torque is about double that of a standard 1 HP, 3 phase motor, and this one is rated at 3000 RPM (5000 max RPM)

A tiny little thing.

1643342001541.png
 
Last edited:
Jim
Thanks for sharing. Seeing how your setup and workholding is done is teally helpful.
 
I got Win 10 installed on the computer and Fusion 360 is happy now. Was able to complete the job with no further issues. I'll get some pictures of the finished bits tomorrow, I forgot to stick my camera in my pocket and was too lazy to walk back into the house to get it.

My software ran fine on Win 10, but I did notice a problem. I use the arrow keys on the keyboard to jog the table around, and they quit working. I still have the buttons on the screen that I use with the mouse, but I use the keyboard arrow keys a lot, so I need to dive deep into the code to figure out why they don't work, they seemed to be locked to the main screen rather than the active window. Windows is strange sometimes.
 
Back
Top