Building a new Do-All band saw variable drive

[A_Pmech]

I'm starting this thread to document a job I have ongoing here at the shop: Building a new variable drive assembly for a Do-All saw. I received this job from a customer here on GJ so that he can get his Do-All up and running.

Do-All no longer supplies parts for these assemblies at a reasonable cost, and many of the parts are no longer available. So, in order to get my customer back up and running, I offered to build a new variable drive assembly from scratch.

On the left in this photo you can see the customer's original assembly. This die cast Zamak pulley set was used on Do-All saws between about 1946 and 1953.

On the right in the photo below you can see the pulley set out of my Do-All V-36. This Bakelite pulley was used between 1939 and 1946 and again after 1953. Presumably Do-All reverted back to the Bakelite design due to failures with the Zamak pulley set. While superior to the Zamak pulley in all respects it still has problems of it's own, mainly the Bakelite cracks with age and the outer pulley halves get loose on the hub as the pulley assembly wears.

I'll be taking the best features of my Bakelite pulley set and making a new replacement pulley set out of aluminum and steel in the next few posts.

Here's the failure which rendered the customer's pulley set inoperative:

To begin, I had to reverse-engineer the Bakelite pulley set out of my saw. Determining the angle of the sheave faces is one of many measurements that had to be taken:

After making up a set of sketches for the various parts, I was ready to begin making chips. The pulley sections will be machined from a 7" diameter bar of 6061 aluminum. Here it is being center drilled in the radial drill:

Machining commences by turning what will become the OD of the pulley sections:

After facing the end, a section of the bar is cut off and mounted in the chuck so the back side features can be turned:

Then the front taper is turned with the lathe running in reverse:

The final turning operation for each outboard pulley half involves boring the center out. The outboard pulley halves will be a press fit on the hub, ensuring a more permanent assembly:

Here's the finished turned pulley outer half blank next to an original bakelite pulley half:

 

[CecilTheTurtle]

Subscribed! Man, I want to learn how to use a metal lathe...

 

[Stuart in MN]

Will the replacement pulley need the radial segments removed so it looks like the Bakelite pulley, or were those done on the original just to save material?

 

[A_Pmech]

PART 2: MAKING THE CENTER SHEAVE SECTION

Here's the center section after roughing one side of the face and boring the hub bearing to final size. The original center section used a bronze bushing with a felt wiper to hold lubricant. Bronze is obsolete in such an application. The new center bearing will use MDS Filled Nylon, a self-lubricating plastic bearing material designed specifically for this type of bearing. The groove you can see in the bore is the retaining groove for the nylon bushing:

To ensure that the two drive faces of the center section are co-planar and concentric to the hub bore, they will be turned in one setup on a shop-built lathe arbor. In this photo I'm turning down a section of 3" steel bar stock on a shop-built lathe arbor to make an expanding collet to hold the center section.

The lathe arbor is simply a shaft tapered at .050" on diameter per inch, roughly the same taper as a Morse Taper. The blank is bored with the same taper and forced onto the arbor where it is self-holding:

After extracting the collet blank from the tapered arbor a number of saw kerfs and relief holes are laid out around the periphery. These will allow the collet to expand as it is forced up the tapered lathe arbor:

After drilling the relief holes the relief grooves are sawed on the Do-All. Then the collet is deburred and the ID honed to remove any upset metal left by the drilling and sawing operations. Then the OD of the lathe arbor is blued up and the collet checked to ensure it is making complete contact with the arbor. Finally, the collet is installed on the lathe arbor and a final .0025" finishing cut is taken to clean up the OD. The OD is lightly deburred with 150 grit sandpaper:

A close-up of the completed collet:

The center section blank is mounted on the arbor by first sliding the collet onto the arbor. Then, the blank is slid over the collet. A forcing tube is then installed on the small tapered end of the arbor and the forcing nut is tightened, forcing the collet up the taper where it expands and tightens against the ID of the blank and locks into place:

Once the collet is forced, the forcing tube can be removed as it distorts the arbor slightly. The taper is self-holding. Indicating the blank showed face runout to be about .0005". Perfectly respectable results.

Machining then commenced. Here's the center section after completing all required turning operations:

The collet is then extracted from the arbor in the same way it was installed. A threaded rod is installed in the back side of the arbor and the forcing tube used to push the collet off the taper.

The completed sections ready for cutting the meshing grooves:

 

[bggrnchvy]

Will the replacement pulley need the radial segments removed so it looks like the Bakelite pulley, or were those done on the original just to save material?

I believe those slots allow the two sides to mesh so you can change the diameter the belt is working on and in effect change the speed.

In short, yes I believe he will be adding those.

 

[A_Pmech]

Will the replacement pulley need the radial segments removed so it looks like the Bakelite pulley, or were those done on the original just to save material?

I believe those slots allow the two sides to mesh so you can change the diameter the belt is working on and in effect change the speed.

Correct! The drive works by altering the ratio between the left and right sides of the pulley. As the center section slides one side of the pulley develops a larger datum diameter and the other section develops an equally smaller datum diameter. The result is a drive ratio change though the pulley.

If the sections can't mesh, they can't produce as large a ratio change. This pulley set produces a maximum speed change of about 8:1.

Machining the timed grooves which allow the three parts to mesh will be the subject of my next post.

 

[Stuart in MN]

I see what you mean now - without the slots, the two halves couldn't come close enough together to drive the belt out (unless you had a really wide V-belt. )

 

[A_Pmech]

I see what you mean now - without the slots, the two halves couldn't come close enough together to drive the belt out (unless you had a really wide V-belt. )

Yep!

Later Reeves drives use a much wider belt. In the case of my Bridgeport it's about an inch and a quarter wide. This pulley set is designed to use standard B-section V-belts.

 

[mjozefow]

I spy Lewis Produce Market bag in the intake of the Quincy... Is that to keep the locusts out?

 

[tk2014]

Awesome!!!!!

I wish you would post more often!

 

[A_Pmech]

I spy Lewis Produce Market bag in the intake of the Quincy... Is that to keep the locusts out?

Something like that...

 

[930dreamer]

Nice work.

 

[mtnkrake]

Subscribed again

 

[BigRed390]

I also spy a certain 2.5 ton drill press that we got teaser pictures of but never saw any "it's done" shots of like the DoAll.

I keed, I keed. But I'd love some info and follow up on that monster. I've been trying to convince myself that I need to wait a while before jumping ship from Delta stuff into bigger industrial grade 'arn. Your posts usually end up pushing me more towards cranking the truck and hitting some estate/going out of business sales.

 

[CarterKraft]

awesome!

 

[Ign]

Awesome. We used to have a large horizontal auto Do All that wouldn't hold blade speed either. I think it was circa '70s tho. There was a hydraulic cylinder that held a pulley or something (memory fading) which kept leaking down. Pulled it apart, cut some gaskets, fixed the leak, worked great.

The same saw would blow apart the bow lift cylinder if you held the joystick too long, sending hydraulic fluid all over you. But otherwise a workhorse of a machine.

 

[OccupantRJ]

Nice work, AP. Love turning aluminum!

 

[NASTYZEN]

That's looking real nice. Are you planning on mounting your vari drive with brass as pictured in the beginning? I got fed up with the rattling in my mill and used Nylon for bushings. What a difference. Lasts longer as well.

 

[A_Pmech]

I also spy a certain 2.5 ton drill press that we got teaser pictures of but never saw any "it's done" shots of like the DoAll.

I keed, I keed. But I'd love some info and follow up on that monster. I've been trying to convince myself that I need to wait a while before jumping ship from Delta stuff into bigger industrial grade 'arn. Your posts usually end up pushing me more towards cranking the truck and hitting some estate/going out of business sales.

Oh, it's not done by any means. It is still very much on the project list. I did fix it up enough to get it running for a few jobs.

If you want big stuff, don't bother with estate sales. Hope you have a 3-ton truck.

The same saw would blow apart the bow lift cylinder if you held the joystick too long, sending hydraulic fluid all over you. But otherwise a workhorse of a machine.

Bwaaahahahah!

Nice work, AP. Love turning aluminum!

It's good fun.

Are you planning on mounting your vari drive with brass as pictured in the beginning?

The axle has an oil reservoir which does an excellent job of keeping the hub oiled but a very poor job of keeping the center sheave oiled. The main hub bearings will be bronze. The center sheave will run on MDS Nylon.

 

[jdcompman]

AP have you finished this guy up yet? I keep checking back every day to see if it's done yet! I'm dying to see your progress and finished product!

 

[A_Pmech]

Good timing! I'm a ways further along with it this evening, so I'll post up some photos shortly!

 

[A_Pmech]

The next step in making the variable drive is to slot the outer pulley halves. The purpose of the slots, as previously mentioned, is to allow the outer halves and center section to mesh with each other to create a variable width space where the belt can ride up and down the pulley cone.

The geometry of the slots is somewhat complicated. First, the walls of the slot are machined on radials emanating from the center of rotation of the pulley. Second, due to the conical section of the center pulley section, the floor of slot must also be an approximation of a conical section.

A couple hours of notebook scribbling and geometry ensued to plan out the cuts, but I won't bore anyone with that! Instead, I'll jump to the machining.

To locate the part I cut out a plug of 3/4" aluminum and drilled a center hole in it for a clamp bolt to secure it to the rotary table. I also placed a piece of paper between the pulley and the table surface. This increases friction slightly which is important, as there is only one bolt holding the pulley half to the table. I used the cross-slide feature of the rotary table to bring the part concentric to the axis of rotation of the rotary table.

Here it is set up on the rotary table, which is set up on a pair of sine rolls and shop-built jacks:

Cutting proceeded by first hogging out the majority of the material from the center of each slot:

Then, I moved the rotary table to locate the radial of the first slot wall. I also made a move in Y to determine the width of the slot relative to it's central radial. Here's the result after machining one half of the slot walls:

Another indexing move was made along with another move in Y to locate the other slot walls. Here's one outer half after completing the slotting operations:

 

[OccupantRJ]

Now that's the kind of project I like to get my hands on, AP. Looking good. I'm machining a special clutch collet at work to hold a cylindrical part at 45 degrees to the collet face, to hold a part in the CNC lathe to drill and tap an angular hole. It's been interesting today.

 

[A_Pmech]

Now that's the kind of project I like to get my hands on, AP. Looking good. I'm machining a special clutch collet at work to hold a cylindrical part at 45 degrees to the collet face, to hold a part in the CNC lathe to drill and tap an angular hole. It's been interesting today.

Sounds like fun! Please post some photos if you can, I'd like to see how you approached it. I'm always looking for new ideas!

 

[gorilla]

Was the second indexing move to put the taper in the slots?

 

[A_Pmech]

Moving forward, the next stage is to machine the grooves in the center section. This has a couple challenges. First, the grooves on either side must be timed to each other. Second, there's no convenient way to hold the section on the rotary table.

My solution to the first problem will involve some careful setup when I start cutting the opposite side of the pulley. To solve the second problem, I had to make some more tooling.

Here's a special expanding collet I made to grip the ID of the center section and hold it to the rotary table. Unlike the expanding mandrel, this collet uses a self-releasing taper as opposed to a self-locking taper:

Here's the setup on the rotary table:

 

[A_Pmech]

Was the second indexing move to put the taper in the slots?

Hi Gorilla,

Yep! After plowing out the middle of the groove I indexed forward 8 49' 30" to get on the radial of the slot wall and moved moved Y-.163" to determine the slot width. The Y move was calculated by taking the tangential movement of the slot exit point during the indexing move and subtracting half the endmill cutting width and half the desired slot width.

After machining one side of every slot I made indexed back 17 39' 00" to get on the radial of the other half of the slot and moved Y+.326 to set the final width.

(It's funny how some numbers can stick in my head days after the job and others I forget while I'm still doing the job.)

 

[A_Pmech]

Slotting the center section proceeded the same way as the outer sections on side one.

But things get a little more complicated once I flip the part over to the second side. The slots on this side must be timed in the center of the lands of the opposite side. To do that I centered the part on the rotary table and centered the rotary table relative to the spindle in X. Then I used a long stem indicator to indicate a random wall of a first side slot relative to the X axis of the machine:

Knowing the dimensions of the slot and the dimensions of the land and knowing I now had a slot wall parallel to the X axis, I made a move on the rotary table equal to the angular width of the slot plus half the angular width of the land. This placed the spindle centerline in the middle of the land on the reverse side of the part. Then I could proceed with with the slots on this side.

Here's the finished result. You can see how the slots on one side are timed to the lands on the other side by looking though the bore:

Next up is making the hub and fitting the center section bearing.

 

[1320stang]

My question is.... are you just making one set? Or are you making a set for yourself as well?

 

[Dirk Hollis]

Will you be applying any surface treatment to the belt running areas? Maybe a hard anodizing?

 

[mjozefow]

It would seem to me that if the Bakelite held up for 50 years, then aluminum would be more than durable enough.

 

[Amitygravel]

John

Boy , those sure are purdy! Neat stuff.
Thanks for the as always awesome photos and explanation of what you're doing.

Craig

 

[A_Pmech]

My question is.... are you just making one set? Or are you making a set for yourself as well?

Two complete drives. One for me and one for the customer.

Will you be applying any surface treatment to the belt running areas? Maybe a hard anodizing?

The pulleys will be as-machined with no surface treatments applied. I thought about anodizing mine a bright red or blue, but just can't bring myself to do that.

As Mitch says, considering the minimal amount of wear in my bakelite pulley set after 60 years, I think aluminum will probably outlast me. It can always be skimmed 50 years from now when the bronze bearings in the new hub wear out.

Boy , those sure are purdy! Neat stuff.
Thanks for the as always awesome photos and explanation of what you're doing.

Thanks!

 

[bggrnchvy]

Two complete drives. One for me and one for the customer.





The pulleys will be as-machined with no surface treatments applied. I thought about anodizing mine a bright red or blue, but just can't bring myself to do that.

As Mitch says, considering the minimal amount of wear in my bakelite pulley set after 60 years, I think aluminum will probably outlast me. It can always be skimmed 50 years from now when the bronze bearings in the new hub wear out.



Thanks!

Will the timed slots begin to gall over time considering the bare aluminum to aluminum mate?

 

[A_Pmech]

Will the timed slots begin to gall over time considering the bare aluminum to aluminum mate?

I don't expect it to be a major issue in light of the wear in both the Bakelite and Zamak drive assemblies I have. They will begin to gall and wear eventually, but we're talking a LOT of hours from now.

The slots have a lot of clearance to each other so, within reason, wear can be tolerated without affecting operation of the drive.

 

[GarageEnvy]

And you're going to bury this in the bowels of a machine where it can't be seen. That's just not right. I think you really should consider a plexiglass door on that saw.

 

[daveroy]

And you're going to bury this in the bowels of a machine where it can't be seen. That's just not right. I think you really should consider a plexiglass door on that saw.

+1
WE have a winner!

 

[1320stang]

You can clear anodize as well.

 

[Nealcrenshaw]

Good work John as usual!

 

[Hephaestus29]

You should have just went ahead & made a pattern to have them cast & then you could have sold some cast pieces that should hold up.