half a millimeter... give or take

In our last episode your local horologist was finally able to get the Vienna regulator's movement to run on a test stand in the shop. The contraption to support the pendulum was simple in design but tricky in the precision of alignment.

What this enabled was the first really clear view of the escapement of the clock in action.

Now this view alone didn't reveal all the mysteries to your clock detective but did lead to the culprit.

Wait... culprit?

Ah yes. 

Why did this horological grease monkey attempt to build the contraption to get us this view in the first place?

It was that the clock was running (good) but with very low amplitude (not good) to the point of stopping (worse) -- and your clock detective suspected an escapement issue.

The only way to watch the escapement in action was to build the contraption.

And your clock doc saw the problem pretty quickly with these enhanced views.

Here's the issue in a nutshell:

The locking on the anchor's entrance pallet was landing so short that sometimes the escape wheel teeth were hitting the impulse face and not the locking face.

I know. 

We need to crack that nutshell.

A. What does all that mean?

B. How does this cause a power problem?

C. Why is it happening?

D. What are you going to do about it smart guy?

Thoughts...

Pondering...

Some summary history of this clock in particular and explication of underlying horological functionality is necessary to explain what is happening. 

Furthermore your essayist wishes to avoid boring drowning the kind reader with endless details (as has been done before) that may not be utterly fascinating to the un-obsessed.

So your editorially-minded scribbler will take advantage of earlier essays as much as possible and presume that brief summaries here and links to those essays will facilitate brevity and suffice for our needs.

Here's hoping.

Breaking the issue down:

Our clock had a locking problem. The "lock" of an escapement is when an escape wheel tooth lands on face of an anchor pallet. The meaning and function of locks and drops were shared in a deep dive of this escapement here. 

The "entrance" pallet is the first of the two pallets that engage with an escape wheel tooth as it rotates through an escapement (also detailed in the prior link).

Very important to note that the locking problem with our Vienna was limited to the entrance pallet. Locking on the exit pallet was fine. 

Here's a diagram of a deadbeat "Entry" (AKA entrance) pallet locking that goes way back to one of YLH's earliest posts on this blog. The escape wheel tooth has landed on the dead/locking face of the entrance pallet.

Now what does it mean when locking is "short"?

Here is another example from that very early post of an entrance pallet drop and an exit pallet lock. The escape wheel turns clockwise here.

The escape wheel tooth lands on what is called the "dead" or locking face of the pallet. It lands somewhat deeply (as they should in this particular type of clock). In this case deeply means the EW tooth lands relatively far from the impulse face. 

For another example, going back in the Clock Fool archive, Charlene's Ingraham regulator had EW teeth that were locking too deeply. 

Above are photos of the exit pallet locking of Charlene's clock. See how the EW tooth lands more deeply on the left than the right? The tooth landing is "shorter" on the right.

The diagram here shows a very short lock on the entrance pallet.

Courtesy of R. Croswell

Well what if gets too short?

Things can go badly.

That's what was happening with our Viennese friend.

The EW teeth were landing right on the edge between the dead/locking face and the impulse face of the pallet -- sometimes they were landing directly on the impulse face.

That is bad.

Why?

Here's another diagram related to the one above showing the entry / entrance lock. This one shows the impulse. After the pendulum starts to swing in the opposite direction the EW tooth slides off the locking face and onto the impulse face. The angle of that face with the pressure from the power behind the EW tooth provides the impulse to the anchor.

This part is critical to understanding the problem: 

In the normal motion of the escapement an EW tooth lands on the lock face, slides forward a bit, then the anchor starts its return swing, the tooth slides back off the lock face and briefly over the impulse face then drops off that.

Watch this with the slo-mo of Charlene's regulator. Observe the direction of the motion of the anchor and pallets as the escape wheel teeth land and slide.

Lock. Slide. Reverse. Slide. Impulse. Drop.

Both entrance and exit pallets. The same.

Now imagine as that anchor is rocking and the EW tooth is about to lock but it doesn't land on the pallet's locking/dead face. It lands on the impulse face. 

If the EW lands on the impulse face it is like slamming the brakes on the anchor. It robs a ton of the energy coming from the escape wheel.

Sometimes that alone will stop a clock.

Sometimes, after an EW tooth hits the impulse face, there is still enough energy (from the momentum of the pendulum) such that the tooth will ride over the edge of the impulse face onto the locking face and then reverse as usual.

This is what was happening with the Vienna.

Still robs a ton of energy from the movement and why we saw such low amplitude.

So to our list above we now understand A. What this means and B. How is causes a power problem.

Now we need to understand C. Why it was happening and D. What to do about it.

As to the C. Why...

There are few reasons that such a problem can occur. 

The first thing to check was the depth of the pallets to the escape wheel -- how the pallets and the escape wheel mesh together (e.g. too close, too far, just right). Technically our Vienna's entrance pallet landing too short means that it does have a depthing problem but the exit pallet locking and dropping was fine. 

Your aspiring engineering detective spent a long time looking at a similar kind of depthing problem with Charlene's regulator -- but -- it was not the same underlying problem. During that diagnostic effort your essayist produced a long exposition on the definition and details of depthing here.

Actually this horologist had a sneaking suspicion about one possible culprit in our current situation.

One more bit of history on this clock. The loyal reader (thank you) might recall that our Viennese friend utilizes a variety of deadbeat escapement known as a Vulliamy. 

Here's the whole anchor and crutch assembly laying on YLH's bench, propped on a pin clamp.

So what is important to remember here? Vulliamy escapements have adjustable and reversible pallets.

Why reversible? Well sometimes after many decades of use the pallet faces can be worn from those escape wheel teeth hitting and sliding on them.

Here's an example of pallet face wear from the Sessions York banjo clock. This is a strip type pallet but the groove from EW wear is obvious.

With Vulliamy escapements the pallets are separate steel rods that have been made, fresh from the factory, with identical pallet faces on both ends. 

One end gets worn? Flip the rod front to back. You have a brand new pallet face.

And when your horological photographer was examining the anchor on our Vienna he happened to capture evidence of exactly that in this super close up.

Those damaged screw heads were the telltale sign that the pallet had been um... adjusted perhaps many times. 

And you can see the grooved gouged in the old pallet face that had been flipped to the back.

Moreover the "new" pallet faces that had been flipped forward were in very good shape. 

Now with a Vulliamy, a clockmaker is seating the pallets on the anchor arms and screwing them down tightly.

But here's the thing with pallets: the length and alignment of the pallets has to be juuuuussstt right. 

So... YLH guessed that the pallet rod had been reseated maybe just a bit off perfect when it had been flipped.

Perhaps reseated just a tad short.

Now we had a strong candidate to question C. Why.

It is likely pretty obvious to the patient reader that the answer to D. What to do about it, is...

Yes, adjust that pallet outwards/longer.

And an important point: a basic tenet of pallet adjustment is... a little goes a long way.

So your micrometrist used a fine tipped sharpie to create a red dot (as seen above and below) that was about 1mm wide, with the intent that it would serve as a measurement aid to eyeball the 0.5mm adjustment YLH reckoned would be a good first adjustment to try.

Again, a little goes a long way.

And with some very gentle tweaking...

Looks ok to YLH.

Now the anchor/crutch was reinserted in the movement up on the test stand to see how we did.

First try!

The entrance locking was perfect, landing nicely along the entrance pallet locking face. Oh and we did not effect the already normal locking of the exit pallet.

Everything seemed right.

Also YLH had neglected to adequately photograph how he assessed a key measure of the health of the clock - the amplitude of the beat of the pendulum.

A metal ruler was set at the base/tip of the pendulum so the length of the swing could be measured. In clock terminology that's called a beat scale.

Like this:

(Clock parks not seen here but you get the point.)

The amplitude (length of the swing) went from a previous 0.75 inches to just about 1.5 inches.

Doubled!

Moreover YLH could assess that there was strong overswing (the distance the pendulum swings after a EW tooth drops).

Yes!

So the movement was left to run overnight.

Happily ticking away the next morning.

Re-lubricated the movement.

Installed in the case.

Runs with a strong, steady beat.