The saw, is a Rockwell Unisaw. Today that would be Delta (following an acquisition and rebranding). From some casting marks, I conclude the saw is from 1979. The saw has seen a number of significant changes in the past 4-5 years but prior to that, it saw almost no mechanical changes since it debuted in 1944! This saw spent it’s first life in a vocational school and it was a tough life.

The saw was equipped with a three-phase motor. I could care less if it was any good since my shop does not have 3-phase power. From the start, I knew the saw was going to need and get a new motor. I eventually settled on a Leeson 3HP motor. The biggest advantage to this unit, in addition to its power, was that it already was configured with the necessary pivot bracket used by the Unisaw. With the motor on order, I set to working on the saw.

The plan was to refinish the top, check the bearings (and pray they were all right), replace the belts, figure out the electrical system, create some option for dust collection, and build a motor cover.

The frame #1 in filmstrip shows the saw with the fence removed and my first attempts at cleaning the top. Eventually, I went for a substantial tear-down. You can see the yoke and other castings in frame #3. After more effort than I care to acknowledge – and more than a few recitations of "wax on / wax off" – I finished stripping the coating on the top. The mess I was dealing with was painted on urethane. This was most likely applied because a vocation school could not be bothered with enforcing proper tool maintenance by the students. Sadly, it was applied after some amount of abuse had already been inflicted. I used a combination of solvent and variously grades of sandpaper and scotch pads to get down to metal. Then a wipe down with Naval Jelly (now just call "rust remover jelly") to get at the rust in some of the scratches and tiny pits that I could not remove. After washing and neutralizing the (phosphoric acid) jelly, I did a little more sanding with 320 wet/dry. The remaining scratches and machine marks were too significant to remove without rendering the top unusable. I finished with three coats of good old fashion paste wax (frame #7).

I had to remove the fence and top along the way so I decided I’d clean up the fence rails even though I plan to eventually replace the fence with a Biesemeyer. I also cleaned the various screws and bolts (a habit I picked up on the clock restoration).

I had to do a bunch of digging but thanks to the Old Wood Working Machines wiki – yes, there is a web page for everything – I found I could use most of the existing electrical system. The control box houses the "Magnetic Motor Starter". I had to switch the step down transformer (which feeds the remote starter buttons and back to the relay) from 200v (part of the 3-phase wiring) to 230v and replace the main power leads and plug. I then used just the first two legs of the system for the new power and motor. Frame #4 shows the electric box with the  motor, power, and remote switch wires left-to-right at the bottom; the 24v relay switch for 3-phase power in the upper left and the step-down transformer in the upper right; the block that looks like it has springs is actually the thermal breaker and those are heavy gauge heater coils that will trip the breaker if they got too hot from too much load.

The new motor arrived via UPS and was install that same night. Frame #5 shows the new motor already installed and the old motor and 3-phase power on the floor. The local industrial /farm supply only had 2 of the belts I needed. The motor drives the blade using three redundant belts. The last had to wait a couple of days.

For dust collection to work, I needed to create a filler panel that is behind the front louvered access door (frame #8) and build a motor cover (frame #6). The cover has a frame contoured and attached to the side of the base cabinet and then, using a piano hinge and latches, the box cover enclosed the motor. Remember, the motor pivots and tilts with the blade so it move quite a bit inside the cabinet – hence the motor cover needs to be spacious.

The finished assemble is seen in frame #8 and you can see the clean and slippery smooth top in frame #7. The blade insert is in bad shape but at $40 each, I’ll make a set out of hardwood. I need to make my zero-clearance and dado inserts anyway so another couple won’t take any extra time. The last step is to align the top to the blade before tightening it all down. Frame #9 shows the dial indicator gauge in position. The first test measures the distance from the track to the blade at both the leading and trailing edges. The second test measures any wobble in the blade. Wobble could be from a bad blade or from misalignment in the mounting arbor disks. Final calibration is not finished yet and I have 0.008 front to back and 0.006 of wobble.

It’s been a fun project. I am really looking forward to setting the Unisaw in place. It will go at one end of the long outfeed table and my JET Contractor Saw will say at the other. The nice part is that the JET Dust Collector has two inlets so I won’t need to switch when I go back and forth between saws (though a remote might be handy <grin>).

The restoration had a number of interesting twists …

I ended up buffing out the damascened that had been applied to the arms of the escapement. The original looked like an after-thought and rushed. Re-applying a circular one would not have fit with the  “simplistic” design of this clock (this clock was never of the design quality of a  Seth Thomas).

The pilot face and the manufacture’s name plate would have originally been cast brass. I designed the pilot face and nameplate on the computer and then we transferred the data to a CNC. The finish was done with a 30/1000ths tool to give it the minor radius that would have occurred from casting. Lastly, it was media blasted to “age it”. I created a custom font for the job to replicate an old / simple casting. You will notice the numbers do not have perfect symmetry. This was a challenge to get something that did not look too perfect while also not looking like a child’s drawing.

Throughout the project, I got the feel this clock was somewhat of a “trial” for the creators. As such, it was simplistic in many places. I tried to keep that feel – while at the same time trying to refine it a bit. One example was the counter-weight to the Harrison maintaining power. You’ll recall it only had a cut piece of steel. This was replaced with a machined “barrel” of brass – made in two parts that thread together. The barrel duplicates the needed weight but looks more appropriate.

Eventually I’ll add the motion works to a 24″ face above and to one side. I picked up a small motion works rather than consider the monstrous one that belongs to the clock.

Right now it is keeping exceptional time – easily within 5 seconds per week. It has about 40″ of fall and I get 32hrs of run. Winding once a day is all it needs.

I definitely want to do this again. I’ll take my time to find the right project but I enjoyed the process and the results so much it will be hard to be patient !

The Arnold & Lewis clock had its ‘minutes’ clock face (called the pilot face) replaced at some point or it never got its final face. The clock had a simple hand stamped brass blank. This was probably not the original plan. Based on seeing an old photo of a similar clock and referencing other clocks, the pilot face should have been cast brass piece.

These days, finding a foundry to make a single piece is costly. But the clock really needed it. Fortunately there are talented and creative craftsmen (gender neutral usage here folks). The machinist I hired thought he could build the piece using a CNC* machine. I created a custom font to resemble cast iron and he programmed the face. The last step was using a thirty-thousandths cutter (0.0030″) to emulate the variances that casting would have produced. It took some trial and error *but* he was ultimately successful.

I liked the resulting part so much, I had a duplicate made as a desk display. The machinist did the same <grin>.piece.

*CNC = Computerized Numerical Control

I suspected friction so I went through it three or four times – cleaning and adjusting the exact seating of the arbors, stands, etc.

When I picked up the clock in Chicago, Mark had a different weight (approximately 80 pounds) and suggested it needed 15-20 additional pounds. I installed the original weight which is 218 pounds on a pulley and provides about 109 pounds of uncompounded force.

I could add another 20-25 lbs but that just doesn’t seem right.

Clearly the escapement is either under-powered or you have some binding somewhere in the train. This is a very high-ratio train – it has five wheels, so it very susceptible to any points where friction might come into play. The way I attack this is to tear it down, starting from the lowest point in the train (the winding barrel) put it together and check for free spin. Go on to the next wheel down the train and so on. With no weight on it check for proper ‘end shake’. Each arbor must freely slide horizontally between the bushings at all rotational angles. The old saying, ‘if it don’t rattle it won’t run’ is true. Another trick is to run the clock just as in the video, start loosening bushing bolts by 1/2 to 1 turn from the upper train down, and see if a surge of strength occurs. If so you have either a miss-aligned, damaged bush, or a poorly finished, bent pivot, (or more rarely, a bush out of rotation but this is hardly ever possible as they are not easily interchangeable).

Source: Mark at “Magnificent Time Machines”

I went through the entire trail again. I was pretty sure the resistance was in the last gear to the escapement but given all the gears in the train, I followed Mark’s advice and went back to the beginning.

I believe I have found the problem. I kept looking at the gears for the resistance but it turns out most of the resistance is in one of the escapement “arms”. There photo depicts the two long arms which hang from their attached long thin arbors.

The escapement arbor turns more slowly when the pendulum swings to the right. During this phase, the escapement arbor is trying to “cock” the left arm. This arm is providing too much resistance.

Part of the problem is resistance in the arm. I’ve eliminated some of that. The other problem is the clock is not perfectly balanced and level. I thought the adjustable ends on the two arms (seen at the bottom of the photo) would solve that but I forgot that the slight off balance nature of my installation means the two arms will want to lean a bit to the right. This means the right arm applies a bit less resistance than normal and the left arm applies a bit more resistance than normal. This is exactly what I hear, when listening to the clock – a solid clank in one direction and a less solid clank in the other.

Balancing the fully installed clock will be a bit tricky but I think that should resolve the issue. (Fortunately, a small 4 ton hydraulic jack makes pitching my clock tower is a lot easier than righting the tower of Pisa <grin>)

Addendum: Turns out the clock was pretty close to perfect level and true. I disassembled the entire clock and the escapement and adjusted every bushing, every arbor, every bolt, and every screw. I found some additional resistance in the left arm of the escapement but not enough. So, in the end, I added 20 pounds … oh well !

The machinist fabricated a vertical stand for the escapement arbor. If you look at the two versions of the clock in “Is it lefhanded or righthanded” you will see the right clock has a small vertical stand holding the escapement arbor in the middle of the A-frame whereas my clock has a cross brace. There were holes drilled and tapped in the A-frame of my clock so I assume it was designed for the vertical stand but was changed. I opted to change it back. In the picture below, the stand is raw steel. It will get the same satin black finish as the A-frame.

I still have to install the reverse preventer on the far left and the third gear in the train. I have some touch-up paint before these can go into place.

I am also waiting on the “minute” pilot face. It goes around the end of the arbor of the second gear in the train (from left to right). I have already installed the gears in the back that would start the lead-off to the large clock faces on the exterior of the building.

You may also notice the small semi-circle to the left of the escarpment stand. This is the old “seconds” pilot face. The new one has not yet been drilled for the mounting screws. However, the face also serves as a bushing for the forth gear in the train so I needed to temporarily place the old one in position to check alignment.

ttd = 86

There?s a fan-looking like assembly to the left of the escapement assembly.. Is that an air governor to slow the speed of the strike train or is its purpose to cushion the action of the escapement itself?

Well, it is an air governor of sorts but the clock is a “single train” or “time only” so it’s not for a strike.

There is a drawing of the double three-legged escapement on page 91 of “A Rudimentary Treatise on Clocks, Watches and Bells” by Edmund Beckett. (You can download the eBook and the link is elsewhere in this blog.) The description starts on page 85, with the four-legged escapement and by the time you get to page 87, the “fly” is finally referenced.

The great feature of them is the regulation of the velocity and the avoidance of the banging on the pallets and the risk of tripping, either actual or approximate, by putting a common fan-fly on the scapewheel arbor.

So, it is, for all intents and purposes, an air brake. It is help in place by friction so when the escapement arbor is arrested on a stop, the fan continues to rotate forward by inertia and there by preventing a bounce. Further, it applies a small amount of resistance, at the start of the release, to prevent a trip.

The picture is an enlargement of the center of the fly. The fly is loose on the horizontal arbor. If you click on the picture and look closely, you will see there is a second “plate” that is slightly bowed. The second plate is trapped along with the fly between a fixed brass disk (on the right) and a pinned brass disk (on the left). The distance between the two disks is enough to allow the fly to rotate but close enough to compress the spring plate slightly – creating the necessary friction.

Early in the clock project I tried to explain the escapement. The description leaves a bit to be desired but I make reference to a video clip (from Mark Frank’s website) that does a great job of showing the escapement and the fly in action.