Building a Strasser Regulator Clock

It has been over a year since the pinwheel skeleton clock was completed and I have been itching to get started on another project. After spending many hours digging through books and searching the internet, I finally found something that looks interesting to build with plenty of challenges to conquer. Here are the basic design ideas I would like to incorporate.


1. Accurate - That table top skeleton clocks I built look nice, but I would not set my watch by them. This clock will have components to help achieve an accuracy of less than one minute per month. That is the goal anyway. We are not talking real astronomical regulator precision with exotic materials or massive structures, but I don't want to adjust the time between windings.


2. 31-day run time - I have enough clocks to wind every week, so this one will be designed to run for a month between windings. Gearing and size of the weight will be determined as the design progresses.


3. Laterndluhr Case - Just because I like the way they look. Very elegant.


While researching regulator clock escapements, I found an interesting design invented in 1899 by Ludwig Strasser. Refer to the attached photo. The suspension spring assembly is designed so the pendulum hangs from a lower block attached to two upper blocks. The inner upper block is fixed and supports the weight of the pendulum. The outer block is connected through a linkage assembly to the pallet arbor. This allows the pendulum to be impulsed from above the suspension rather than below it on the pendulum rod. Although this is not a completely free pendulum, it is very close. Since the pendulum is impulsed in both directions through the same spring, every impulse should be identical. In theory, variations from the escapement and wheel train should not affect the pendulum.




Here are a few links to give you a preview of how this escapement works. This is an animation that allows you to stop and advance step by step. For some reason, I cannot copy the direct link to the Strasser escapement, so after you get <here> go to the left side and navigate to Theory\Escapements\Detached\Strasser.


There are also several Youtube videos by Steffen Pahlow showing how he converted a Graham deadbeat escapement to a Strasser. Here is a link to the first video. There are three parts total. Subtitles in German and English.


Construction will begin in my next post. This will be a real-time build log and I will post updates as they occur. There may be long gaps where I am busy on other things and I may need to scrap a design and start over if something does not work out. This may take a while.
Enjoy,
Allan

 

Tok,
I believe I found a photo of the Sattler clock with the escape wheel near the bottom of the movement and the verge upside down. Very interesting design.

One of the nice things about the Strasser escapement is it uses the same escape wheel design as the Graham, so I was able to use a design spreadsheet developed for Graham escapements. The dimensions provided by the spreadsheet were then used to create a CAD drawing of the wheel and a fly cutter. The wheel is made from 1/16" thick brass and the fly cutter from a piece of drill rod ground to shape and hardened.


I am using DraftSight for all my drawings. It is only 2D, no 3D capabilities, (which I have not been able to comprehend anyway) but it is free and uses the same commands as Autocad. Sorry that some of the dimensions are crammed together. These drawings were made so I could build the parts and mot spend a lot of time making the drawings look pretty.

This is a photo of the wheel being cut on the Taig lathe. The layout die on the edge allows me to see when the cut is deep enough to leave just a slight edge on the tooth. Notice that the cutter is offcenter to the left where the front of the tooth is vertical. When the tooth is at the top of the wheel, it will be sloped to the right so only the tip of the tooth contacts the pallets.


OK, so I don't treat some of my tools very nice, but this is a cheap scriber and I am using it to layout the 5 spokes with the aid of the dividing plate. The spokes are cut out by hand with a jeweler's saw and filed to their final shape.


The wheel is almost finished, but I need to enlarge the center hole to fit the hub. In the photo below, I am using a technique called "turning in a box". I found this idea in JM Huckabee's book "Top 300 Trade Secrets of a Master Clockmaker". A block of wood is mounted on the lathe and a pocket is turned into the face. This pocket will run perfectly true as long as it is not removed from the lathe. The diameter of the pocket is just large enough so the wheel fits snugly inside. Now the center hole can be bored true with the teeth. This works well for small wheels and I will show a somewhat different method I like to use for larger wheels.


Here is the finished wheel and you can see screwup #1 (the first of many I'm sure). I should have left more material in the center to fasten the wheel to the collet. I don't want to rivet it, so I decided to split the difference and have the screw half on the wheel and half on the collet. It should work OK but it is more fiddling than just drilling the holes.

 

The collet for the escape wheel is made from 3/8" brass rod. I will use #36 music wire (.106") for the arbor since I have plenty of it on hand and anything bigger will not clear the screw heads used to fasten the wheel.


I did not take any photos of the collet being machined, but there are a few things worth mentioning. The 3 screw holes were drilled and tapped before the spigot was cut for the escape wheel. It would have been impossible to drill the holes halfway over the edge of the spigot. The diameter of the spigot was machined for a tight fit inside the escape wheel. Also, the center hole was drilled to about .010" under size and then bored to its final size with a very small boring bar. Boring the hole ensures it runs true with the outside and face of the collet so the escape wheel doe not wobble or run off center.


Here is the wheel mounted on the collet, minus the screws. The spigot sits about .010" lower than the wheel so the screw heads will contact the wheel and pull it firmly against the collet.

 

The next part to be made is the anchor and pallet assembly. The Strasser escapement uses a two-piece pallet which I will call lift and lock. The lift pallet has an slightly inclined impulse face that lifts the anchor arm and provides impulse to the pendulum. The lock pallet provides a detent to stop the escape wheel and hold it until the pendulum swing travels far enough to release it.
Without dimensioned drawings, I will need to make some assumptions in the design. Luckily, the pallets will be adjustable so the lock and drop can be tweaked to find the optimum settings. If anyone sees an issue with the methods I have used, please let me know and we can discuss alternatives.


The known design criteria I have to work with are:
1.2" escape wheel diameter
7.5 tooth span
2 degree pendulum swing


This drawing is a result of a bit of trial and error CAD work. I basically used a spacing between the escape and pallet arbor that looked about right. This determined the length of the anchor arms. A greater space between the arbors will require longer arms to reach the escape wheel. I hope this works because the anchor is not adjustable. The drawing shows the escape wheel locked on the left pallet and the pendulum hanging straight down. The blue line represents the pendulum and a small beat scale is shown in the center of the escape wheel for reference. The magenta line represents the outer suspension spring that connects the pallet arbor to the pendulum and the rectangular box at the top is the outer block of the suspension assembly.





As the pendulum swings to the left, the suspension spring drives the pallet assembly clockwise. When the pendulum swings 1 degree, the left pallet unlocks the escape wheel. With the length of pallet arm used, the locking pallet face needs to be .010". If this needs to be adjusted, the anchor arm will need to be longer for a larger face and shorter for a smaller face. The escape wheel tooth is almost touching the right lift pallet. I set the angle of the lift pallet at 10 degrees which gives .005" of clearance for drop, allowing the escape wheel to turn 1.7 degrees before contacting the impulse face. This was an arbitrary choice. To change it, I will need to change the angle of the impulse face on the lift pallets. A 45 degree angle on the exit side of the lock pallet is also arbitrary but not critical. It just neds to be adequate to clear the previous tooth.




The escape wheel is now locked on the right pallet and has pushed the right anchor arm up, driving the outer block of the suspension assembly to the left. This position corresponds to the pendulum position at zero. With the pendulum 1 degree to the left, the suspension spring is providing impulse to push the pendulum to the right, represented by the skewed magenta line.




The pendulum has now returned to zero and the escape wheel is still locked on the right pallet. The suspension spring is once again aligned with the pendulum and provides no more impulse. From this point, the pendulum will now drive the pallet arbor through the suspension spring to lift the pallets and unlock the escape wheel.




The pendulum has now swung 1 degree to the right, unlocking the escape wheel from the right pallet. This time, I have shown the escape wheel rotating 1.7 degrees until it contacts the left impulse face of the lift pallet. Note that the suspension spring is still aligned with the pendulum so it is not yet providing impulse.




Now the escape wheel has rotated an additional 4.3 degrees (6 degrees total), lifting the left anchor arm, driving the outer block of the suspension assembly to the right with the suspension spring providing impulse to the pendulum.



There is no way I could have figured this out without CAD. It has really helped to be able to visualize how the parts interact and to determine some of the dimensions through simulation. In case you are wondering, I did not animate the CAD drawings to make the parts move automatically. Instead, I rotated each piece using basic CAD commands.
Please comment if you see any mistakes or poor design practices. I feel like I am stumbling my way through the dark on this one.
Allan

 

I would like to thank John for sending me some additional information on the Strasser escapement which helped to confirm the design. Here is the CAD drawing. All that remains is to build it and see if it works.



The anchor is made from 3/16” brass. The outline is cut with a jeweler’s saw and filed to shape. Lathe work is done using a modern version of a shellac chuck. The chuck is a 2” brass disk with a steel arbor pressed into the center. Since I did not have any shellac on hand, I used a glue stick for a hot melt glue gun. The chuck is heated with a propane torch and then the glue stick is rubbed on the surface. The mass of the chuck keeps the glue fluid for about 5 minutes which is plenty of time to position the part. A small pilot hole is drilled for the arbor and is used to center and hold the part with the tailstock until the glue sets.

A small 1/8” boring bar is used to cut the slots for the pallets and thin the arms. In this photo, the lathe is run in reverse to cut the areas that cannot be reached when cutting from the normal side. The arbor hole is bored with a smaller bar using this setup to ensure the hole is exactly the same distance from each pallet opening. The center hub in this photo was removed and the mounting holes were spotted and drilled through on the drill press in a later step.



The chuck is heated to melt the glue and remove the part. Excess glue is easily wiped off while it is still liquid and any remaining glue will be removed during final finishing. Two pieces of 1/32” steel were filed to shape for the pallet retaining brackets. This photo shows the anchor and retaining brackets which still need to be finished to remove machining marks. I will make a collet for the anchor similar to the escape wheel.



The pallets will be made next. In my never-ending quest to make life difficult, I have decided to make the pallets out of tungsten carbide. Sapphire or ruby would look cool, but I have no experience working with it and it is not as readily available as carbide. I have started experimenting with working the carbide and I can say with a great deal of certainty that this is incredibly hard stuff. There may be a brief pause in this project while I “tool up”. If anyone has any experience in this area, I would love to hear from you. At one time I recall reading a BHI article on making pallets from carbide, but it seems to have disappeared from the internet. Any leads would be appreciated.
Allan

 

Jim,
Thank you for providing the photos of your pallet adjusting tool. It looks like a handy device. In this escapement, the junction of the lifting and resting pallets are located on the diameter of the escape wheel, so it should be possible to lock the verge and rotate the escape wheel to push the lifting pallets into position. I may be over simplifying the process as I have not given it a lot of thought yet. I may need to take you up on your offer if my method fails.

Regarding cemented carbide acting as a lap and wearing the escape wheel, I had not thought of that and will do some additional research in that area. You stated that you plan to use ruby for your pallets and I see that you have installed jeweled bearings in your regulator project. Can you share your source for the ruby material and jewel bearings? The suppliers I have found seem to have a hefty minimum order. I am thinking of using ball bearings on the rotating arbors but would like to use jewel bearings for the verge arbor.
Thank you for the helpful insights,
Allan

 

I appreciate all of the information regarding carbide vs. ruby or sapphire and the sources for the ruby material. I have decided to proceed with using tungsten carbide for the pallets and hope that Jim will keep us posted on his experience making pallets from ruby. We should get together in 50 years to see how they compare!


I suppose I am obligated to state the health risks of working with carbide, specifically the dust created from grinding. Use dust control measures and proper protective equipment. There is lots of information on the internet; here is a good, easy to understand explanation of all the hazards. Sounds nasty; you have been warned.


I am using SNG carbide inserts as my source of material. $3 each for Wholesale Tool's import brand. These are intended for lathe cutting tools and are 1/2" square and 1/8" thick. I need 4 pieces approximately .4" X .13" X .125" so I intended to cut these in half. Using some cheap diamond cutoff wheels (5 for $5) and water for coolant, I quickly found out just how tough carbide really is. The diamond grit (they said it was diamond anyway) was stripped off of the edge of the wheel in about 10 seconds and not a bit of the carbide was removed. Tried another wheel at slower speed, more water, less pressure. The same thing happened only the wheel lasted about 30 seconds. Thinking the high quality cutoff wheels might be to blame, I tried a small Dremel brand diamond grinding wheel and it did actually remove some of the metal, but it is too wide to saw with. Note to self; stop buying cheap tools.


Obviously, I am not going to be able to cut these in half without taking them to a shop equipped with EDM. Time to regroup. Back to Wholesale Tool. (I feel so lucky to live in one of only three locations in the entire US with an actual brick & mortar Wholesale Tool store!) Pick up a (brand name) silicon carbide grinding wheel for the bench grinder. These are the green wheels used to sharpen carbide tools. After a lot of grinding, I was able to reduce 4 perfectly good inserts like the one on the right, to 4 future pallets like the one on the left. The 6-inch grinding wheel did a great job cutting down the inserts, but it is now about 1/4" smaller. Did I say this carbide is tough stuff?



Now that the blanks are done, they need to be ground into curved shapes. I don't think this is going to be easy.

 

The pallet blanks need to be ground in a curved shape so the edges will stay the same distance from the anchor arbor as the depth is adjusted. The drawing in post #12 shows the outer pallets will have an outside radius of .644" and an inside radius of .594". A piece of brass is turned to form a jig to hold the pallets with an outer radius a few thousandths under size and the inner radius a few thousandths over size. This provides clearance so the diamond wheel does not contact the jig. The first photo shows the pallets hot glued to the jig and the second photo shows how the assembly is mounted in the lathe. A 7/8" diamond wheel is mounted in the Dremel flex shaft tool which is mounted to the cross slide. A tool post is mounted on each side of the handpiece for extra support.





The lathe headstock is turned by hand while the diamond wheel is fed into the pallets .001" each time. Water was applied with a spray bottle, which the diamond wheel promptly sprayed all over the shop, and me. Unfortunately, the hot glue was not strong enough to handle the pressure from grinding and the pallets came loose from the jig. As a second attempt, the pallets were fixed to the jig with super glue and the center of the jig was drilled and tapped so a cap could be fastened to hold the pallets in place. This worked much better and the outside of the pallets were ground in about 20 minutes.





But now I cap blocked access to grind the radius on the inside of the pallets. To fix this problem, the cap was replaced with a scrap of aluminum with a hole bored in it as shown in the next photo. Luckily, the super glue held the pallets in place while the cap was replaced. The next problem to arise was with the diamond wheel grinding inside the pallets. Although the wheel fit, the pallets began to act like a brake shoe on the wheel. With too much contact area between the pallet and wheel, the Dremel tool would bog down so much that material removal was extremely difficult. Replacing the diamond wheel with a small diameter diamond point solved this problem. I also added a shroud made from a plastic milk carton to keep water from spraying all over the shop, and me.




This photo shows the entire setup. A piece of wax paper was laid over the bed and directed the water and waste carbide into a bucket. Although it looks hokey, it actually worked quite well at dealing with the water. I did need to remove the chuck and assembly from the lathe to measure the inside diameter with a calipers, but that only took a few seconds.





After removing the outside pallets, the jig was then turned down to cut the inside pallets. Although setup times were faster for the second set, the diamond wheels had lost some of their edge and did not cut as fast. After about 5 hours of jig making and grinding for both sets of pallets, we have a set of curved pallet blanks. the next step will be to grind the lift and clearance angles. I should be able to do this with the diamond wheels I already have, but I will need to order some diamond compound to lap and polish the working faces.


After all of this grinding and exposure to water, I treated the lathe to a tear-down to clean, dry and oil the entire machine.

 

The diamond polishing compound arrived and a bit more progress was made on the pallets.
I ordered a kit from Gesswein that included 230, 600, 1800, 8000 grit compound and a bottle of lubricant thinner. I used the oil-based compound.
I started off by grinding the ends of each pallet to the approximate angle so I would have a reference point while polishing.


Next, I turned an inside pocket on a brass bar to the same radius as the center pallet line. This served as the lap to polish the face of the left locking pallet. Even though only .010 inches of the face will be exposed, I lapped the entire face to keep the curvature consistent. Starting with the 230 grit compound, I was surprised how fast it cut the carbide. After just a few minutes at slow speed, the grinding marks were removed. This photo shows how I hot glued the pallet to a smaller piece of brass for use as a handle while working inside the pocket.





The diamond in the compound had embedded into the brass as expected and formed a very effective abrasive surface. This also meant that I could not just apply the next finer grit compound for the next step. Instead, I had to machine off the pocket and cut a new one. Going through 3 grits (I skipped the 600 just to see if I could and it worked just fine) took longer to machine the pockets than it did to lap the pallets. After the 8000 grit compound (2 - 4 microns), the pallet had a nice mirror finish.





Next I lapped the face of the right locking pallet. Same technique, except this time I needed to lap an inside radius so I turned the outside of the brass bar to the same diameter. Here I was able to hold the pallet in place with my finger. This photo shows the second pass. The bright section of brass to the right of the pallet was where I did the first, coarsest lap and then turned down the diameter to remove the embedded compound so it would not scratch the finer finish. The dark band to the left of the pallet is just oil and carbide that was removed by the lapping that got smeared from my thumb. Lots of fun but a little messy.





Now I need to grind the lifting pallet faces to the correct angle and lap them.

 

Before lapping the faces of the pallets, I ground a rough approximation of the angles for each one on the bench grinder. This saved a considerable amount of time cutting the final angles. Each pallet blank was drawn inside a rectangle that approximated the general outline. A custom "protractor" was also drawn so each pallet could be laid on the drawing to check the angle as it was ground.





The clearance angles of the locking pallets where done first to practice on parts where the angles are not critical. This photo shows that the pallet blanks were mounted in the verge for grinding with the lifting pallet pulled back out of the way. The pallet being ground was extended about 5-10 thousandths more than its final length to leave enough material to square up the face from the rough grinding. A drop of super glue was placed where the pallet and verge meet for a little added protection to keep the pallet from slipping. The glue bond is easily broken with heat. The verge was squared with the diamond wheel and advanced about .001" for each pass.





The next photo shows the exit locking pallet being lapped. Three small brass disks were cut from 7/8" diameter rod to use as lapping wheels; one for each grit. This allowed each angle to be ground and lapped with without moving the verge. The diamond particles in the compound embedded nicely in the brass. Compound was only applied to the lap for the first use. After that, I only applied fresh cutting fluid. I made the laps small so they would fit inside the verge.





This photo shows one of the lifting pallets being lapped. It is a little blurry, but notice a button has been inserted into the verge arbor hole. The button is turned to .105" radius per the drawing in the first photo. A straight edge is placed against the button and the pallet and then the entire assembly is rotated so the straight edge is parallel to the lap. This sets the angle of the lifting pallet.





Here are two of the pallets after lapping is complete. The 8000 grit compound leaves a nice mirror finish.





The pallets are installed in the verge in their approximate position. The only work left on the pallets is to grind some of the excess material from the tail. That is a good thing because I am ready to wrap these things up and move on to another part.

 

I won't bore you with the final cleanup details of the pallet assembly, so let's move on with the suspension spring. Normally this is a fairly plain and simple part, but not in this case. The suspension spring is the heart of the Strasser escapement.




The drawing was made by estimating the approximate dimensions from photos and drawings. Fortunately, the assembly can be disassembled if adjustments are needed. I suspect if the dimensions I have chosen do not work as planned, the only components that will need to be changed are the springs themselves. Thinner or thicker, longer or shorter,wider or narrower springs should give a wide range of adjustment. On the unfortunate side, there are quite a few small components that make up the suspension assembly including 16, 00-90 screws. Life is too short for me to be making these screws and I will buy them instead.


Note that the inner springs are slightly lower than the outer springs. This is to align the bending moments of the inner and outer assembly as shown by the magenta line. According to an article found on a German website (http://www.info-uhren.de/) and translated by Google Translate (which does not do such a great job on technical terminology) the bending moment of the inner assembly is 1/3 of the distance from the top while the bending moment of the outer assembly is 1/3 of the distance from the bottom. If I understand the article correctly, Strasser initially had difficulty with his design with the springs in-line and later offset them to correct the problem.
Strasser used a ruby pointer and cup to couple the suspension to the verge, but I am going to try something a little simpler which will be described when we get a little further along.

 

There is not a lot of exciting machine work to making the suspension spring. The mill was used to square up a few edges and make sure things are parallel, but most of the work is done with a saw and files. Here are all of the pieces.






The spring material is .005" thick blue spring steel. It was rather hard to cut all of the pieces to the same width, so I ended up putting them in a machinist vise with the amount of material to be removed sticking up. The vise was then touched to a belt sander which removed the excess material in a flash and left a smooth, straight edge. The ends where sanded to length after assembly.
The jig on the far right with the hex screws was used to punch the holes in the spring steel. It is just two pieces of scrap steel screwed together and then drilled through with two different sizes. One hole is for a close clearance fit of a 00-90 screw, and the other is a little bigger in case extra clearance is needed. The pins above and below the jig are made of music wire ground flat on one end for cutting and rounded on the other where it is driven with a hammer. The spring steel is marked to locate the holes, slipped between the plates of the jig, screws tightened to clamp the spring and the pin is hammered through to punch the hole. It would have been nice to make the jig with a guide to help align the holes better, but they get covered up anyway.
Drilling and tapping the 00-90 machine screw holes was a bit teadious. I bought two import taps from a clock parts supplier. They were cheap at $6 each and that is what I got; cheap taps. The points of the taps were flat instead of pointed and they cut terrible threads; more like a spiral ridge, but the screws go in and seem to hold OK but I am careful not to overtighten them. I also broke a tap even though I was using a tapping block to keep the tap straight. I know I said earlier that I should quit buying cheap tools, but a name brand tap in this size was listed for $40.


Here is a front and side view of the assembled suspension spring. The long pin in the top is just stuck in there for now until the length can be determined. This is where the impulse from the verge will be applied to drive the pendulum. The center pin with the brass sleeve will support the suspension in the bracket. I may replace the lower pin and attach the pendulum rod with a screw instead of hooking it over the pin.




This is a drawing of the setup I plan to use to connect the verge to the suspension. It is from Charles Gros "Les Echappements d'Horloges et de Montres"(Paris 1913, 2nd Edition). It is simpler that Strasser's frame assembly and should provide the same function, although without a beat adjustment. I am going to borrow Jim's idea from his precision regulator and use drill blanks or music wire as wear pins where the arm contacts the suspension drive pin at point 'c'.

 

The old sketch in the previous post showed one version of the simplified linkage to connect the verge to the suspension spring and drive the pendulum. Here is my version.



I do not see any reason why the bridge is needed for the verge arbor unless there is some clearance issue that is not in view, so I will locate the crutch inside the back plate. I will also turn the verge collet into a spool (upper left) that will hold both the verge and the crutch with 0-80 screws. The length of the spool was determined so the escape wheel will run on the center of the arbor and the crutch is very close to the back plate.
This may be one of the shortest crutches ever at slightly over 1 inch. It is made from 1/8" brass rod bent at 90 degrees with the ends reduced to a 1/16" pin on each end. One pin goes into the crutch collet fastened to the spool while the other pin attaches to the fork. Solder would probably be the correct way to fasten the components together, but I am using Loctite 609 retaining compound. It is much faster and easier than solder and it has never failed on me. At a shear strength of 3000 PSI, I suspect the crutch will bend before the Loctite gives way.




Here is a partially completed spool and crutch. I still need to make the fork. Note that the screw holes on the spool are oriented differently on each end. I had already drilled the verge with one of the holes at the top, but this would interfere with the crutch pin on the other end. No big deal, but I had to make the spool piece twice because I overlooked this detail on the first part.

 

The fork was cut from a small scrap of 1/8" thick brass. All of the work was done on the mill to keep the edges square and also to drill the holes for the drill blank inserts to a tight tolerance. This photo shows how a small Jacobs chuck is inserted into the end mill holder for drilling.



The Benchmaster mill only has about 9 inches of vertical work space and this setup used all of it. The table is all the way down. There was not enough clearance to fit a 1/16" end mill into the chuck to cut the slot in the fork and I don't have a 3/16" end mill holder. The supply house did not have a 3/16" holder so this was a perfect excuse to buy that set of MT2 collets that I have been wanting for the mill. Worked great! Despite the Benchmaster's limited vertical clearance, I would not trade it for anything. It is perfect for the size of work I do and is a pleasure to use.




Here is the finished fork attached to the crutch. You can see the 1/16" drill blanks that will serve as the contact points on the suspension spring drive pin. There is currently about .005" clearance between the drive pin and the contact points. I will leave that clearance for now and adjust later when the entire escapement is tested.





I think I will make the suspension bracket next.

 

Here is the drawing for the suspension bracket. Nothing fancy. The only critical factors are that it be rigid and the suspension assembly fit between the forks without any slop. I don't think that the suspension will move around due to impulse from the escape wheel; the weight of the pendulum should prevent that. However it could move over time and affect the beat.




A few additional dimensions are provided to make it easier to cut the notch for the suspension pin. The notch was cut by tilting the bracket at 45 degrees and making a pass with the corner of an end mill. The .691" dimension provides the starting location and .083" diagonal measure is the amount the table was advanced in the X direction. The end result is the notch is centered .75" from the base and .079" deep so the center of the pin sits below the top of the bracket. CAD is a wonderful tool!





The rest of the mill work was pretty straight forward with two exceptions. I did not have any 3/4" square brass, so it was cut from a piece of 1" rod. The outer corners of the forks are the only clue that this came from something round. Cutting the countersink for the mounting screw was complicated by the fact that the head of a #10 screw spans the entire distance between the forks and a standard countersink would not fit. I had to make a countersink from a piece of drill rod. It wasn't the best cutting tool, but it did the job for this one part.





The drawing shows a small hole for a taper pin to keep the bracket from turning. Instead, I may cut a small pocket in the mounting plate.


It may seem like I am jumping around from part to part with no obvious plan. That is probably because I am jumping around. A little insight into where I am going may help. I am currently building up the parts needed to test the escapement since this is the big unkown right now. If I cannot get it to work, I don't want to scrap any more work than necessary. Here are the parts I still need for the test:
mounting plate for suspension bracket and plates
plates and pillars (may use a temporary assembly)
escape arbor
pendulum (final rod, temporary bob)


I am still pondering if the escape wheel should be driven with a small weight and cord wrapped around the arbor or if I need to make the second wheel and drive the escapement through a pinion. Will the dynamics be close enough driving the escape wheel directly? Since there is no recoil, I would think so. Anyone have any thoughts on this?

 

I was in junior high school when the great metric debate was going on. If we had only bit the bullet and made the switch, all these fractions and odd decimals would be a distant memory by now.


Spent some time on the plates, pillars and mounting plate.





The mounting plate is 1/4" aluminum. The suspension bracket will mount in a slight recess to keep it from turning. With the recess .022" deep, the distance between the back plate and mounting plate will be a nice even 1.25" (I hear you laughing tok-tokkie, it's OK.)
I decided to go ahead and make up the actual plates instead of a test assembly. Hopefully I will not mess up and have to redo them after the escapement test. The plates are 1/8" thick leaded brass. All of the edges and holes are positioned using the mill so they will be square with each other. I had an idea of installing standoffs that connect the movement to the mounting plate using the same screws that hold the pillars, so I drilled the pillar holes through the plates and partially through the mounting plate to mark the position. After some thought, I see a lot of work making these mounts so the movement can be installed in, or removed from the case with this method. So, I decided to offset the standoffs inside the pillar holes. Since the standoffs will only connect to the back plate, I positioned this plate on the mounting plate using a brass rod in the pillar holes for alignment as shown here. I guess the previous holes in the mounting plate will be extra.





I have slept on this and changed my mind again. Attaching the movement to the mounting plate with standoffs will secure the back plate, but the front plate will be suspended from the pillars. I fear that the amount of weight needed to make this clock run for 30 days will put a bind on the movement using this configuration. Instead, the movement should sit on brackets that support both plates. So now I have 8 extra holes marked on the mounting plate. At least 2 of these holes will be used to attach the new brackets. Unfortunately I still have 4 extra holes in the back plate; big 1/8" holes. Looks like I already messed up.


While looking at photos of several Vienna regulators, I notice that most mount the dial pan to the front plate with standoffs. Hey, I will just swap the front and back plates and use the extra holes to mount the dial. Yeah, I meant to do that all along .


I just need something to hold the plates for now, so only the ends of the pillars are machined at this time. I will turn the center sections later so I do not have to worry about scratching them up during construction. Turning 8 ends still takes a fair amount of time, so I try to make as few changes to the lathe tooling as possible. However, this is probably getting a little ridiculous!



You can probably figure out what each tool is for, but I might explain that the curved tool on the right is being used to undercut the end of the pillar so only the outside edge of the pillar contacts the plate. This keeps the pillar from rocking. Because of the large contact area on this tool, the lathe is turned by hand to prevent chatter.




Here are the assembled plates and pillars. I will make the screws and washers later. For now, store-bought ones will be used during development. I need to order some material for the brackets, so they have not yet been started.

 

I spent a lot of time getting the escapement components mounted in the plates. At first the pallets were too long and locked up the escape wheel, so I had to adjust them a little at a time until the action seems about right while moving the wheel and verge by hand. No doubt additional adjustments will be needed as I try to get it to run. For the test, I am using standard pivots and broached holes in the plates. These will be replaced later with jewel bearings on the verge and ball bearings on the rotating arbors.


I also got the brackets made and mounted to the backplate. This took lots of test drawings in CAD along with double and triple measuring to make sure the brackets position the movement at the correct height and distance from the suspension spring. If the height is off, the verge arbor will not line up with the flex point of the suspension spring, resulting in excessive sliding between the drive pin and the fork. Here is what I came up with. The movement fastens to the brackets with #6-32 screws into threaded holes in the bottom pillars.





After all of my careful planning, the height looks pretty good, but somewhere along the line I ended up with the fork too close to the suspension drive pin. It is at the very back of the fork instead of centered. Luckily there are lots of options to correct this and I am going to make a new verge spool that is about .15" shorter. This photo also shows the small brass bobbin mounted on the escape wheel arbor. A string will be wrapped around it with a small weight to drive the escapement for testing.

 

This will definitely be a learning project with this escapement. It should not be long before I can run it and determine the amount of weight needed to keep it going.

I think I have addressed the issue of the connection between the fork and spring pin, pretty much just as you described. Take a look at the last photo of post #33 to see that a pin has been installed on each side of the fork. This should allow the spring pin to have enough freedom of movement so that it does not bind. I left just a few thousandths clearance between the pins and spring pin. Moving the verge and escape wheel by hand shows that there is not that much travel when the escapement drives the pendulum, so I cannot afford to leave any more clearance than necessary.

 

The Escapement is WORKING !!!!

A temporary assembly was rigged up and the escapement runs great for about 4 minutes, which is when the string runs out. I am using the pendulum rod with no bob, so it runs a little faster than 1 second. With an initial swing that is just enough to unlock the pallet, the rod gains a little amplitude as it runs. This is a good sign, indicating that the escapement is providing enough impulse to keep the pendulum going. I am currently using a small piece of brass that weighs about 10 grams to drive the escape wheel, although my kitchen scale is not very accurate. The drum is about 1/2" in diameter.

Sorry, no photos. How about a video instead? Watch it here.

I feel like a major milestone has been reached.

 

Rob,
I agree, trimming material off of the back of the suspension mount would have accomplished the re-positioning needed. I also could have moved the escape wheel and verge forward on the arbors, moved the entire movement or shortened the crutch. It is good to have options. In the end, shortening the spool was the method I used.

Tok-tokkie,
Most of your calculations match up very closely with mine. Thank you for providing this information. The ratio of escape wheel to great wheel in my calculations is 2250:1. One difference is that this will be a 30-day wind (actually 32) and so the drum diameter is about 21 mm to reduce the drop per day to make it run longer. Obviously more weight will be needed for the smaller drum, so this is a concern.

The test run was mainly to confirm that the geometry of the escapement was correct before proceeding with the remainder of the movement. The 10 gram weight is highly suspect since the scale I used was a cheap kitchen spring type with the smallest increment being 10 grams. The escapement had a fairly healthy snap to it so the required weight could likely be less. I will also be changing out the standard pivots with miniature ball bearings, adding a bob to the pendulum and lots of other changes that will affect the drive weight.

My current plan is to build the remainder of the train and determine the drive weight through experimentation. Not very scientific.

I found some 12-leaf cycloidal pinion wire that I plan to use. If you read my thread on building the pinwheel clock, you know that I had a lot of problems cutting pinions in the past. Today I sliced off a section of the pinion wire, drilled and reamed it to fit the escape arbor in about 10 minutes! This is much easier and I wish they still made pinion wire for clocks. I will need to harden the pinions since the steel seemed to cut very easily. The leaves will also need to be polished to remove score marks from the drawing process used to manufacture the pinions.

Does anyone have a good method for polishing pinion leaves and wheel teeth? I recall seeing some information about using a wooden form and Simichrome but cannot locate it.
Thanks,
Allan