As noted in earlier episodes, your (ever learning) local horologist will often effect repairs even in the early phases of the examination of a clock. Pleasingly more of these have become regular parts of basic maintenance.
Amongst the many efforts made by this lover of brass in the early phases of examination of Charlene's Ingraham regulator clock was the quick identification and gentle straightening of an errant escape wheel tooth, hopefully preventing the problems that can result.
Subtle, yes, but that tooth is bent forward which can upset the beat of the clock.
While the condition was visually subtle YLH did spy that bent tooth right after disassembly of the movement and while evaluating each wheel.
YLH really enjoys no longer being a clock freshman.
However, the errant tooth was not the source of the double tick / noise that was vexing this horology hound.
No, it turns out the underlying issue of that condition would be somewhere in the alignment of the anchor to the escape wheel.
Alignment is almost a euphemism in this case.
In our last episode, this essayist took the plunge into some of the essential parts and actions of the escapement.
Summarily, the escapement is that part of the clock movement that transmits and regulates power from the wheel train to the pendulum.
And with that synopsis, a quick repost of a graphic from last time with two additions.
Observe now the pivot cock and the pivot cock rivet. The reader/viewer will recall that the anchor pivots on the pivot point highlighted above.
Newly highlighted and key to understanding the issue of this extra sound is to observe that the pivot itself rides on a piece of brass that serves as an arm (AKA a cock) and it is riveted to the larger brass plate below. Furthermore that rivet, while extremely tight, does facilitate (with careful but very firm pressure) the turning of the cock, AKA moving the pivot point towards or away from the escape wheel.
Why would anybody want to do that?
Good question Watson.
The fact is that the pallets and the escape wheel teeth have to mesh at specific depths and angles to each other.
If the playful visualizer were to imagine some fool wrenching the pivot cock sharply upward, the pallets would mesh so closely with the escape wheel teeth, that nothing would move. Reciprocally pulled downwards, and away from the escape wheel teeth, the pallets might not engage with the teeth at all.
So yes there is a Goldilocks just right depth.
And maybe there is some math that can figure it out but it was clear to YLH that Charlene's escapement was meshing too tightly AKA "locking" too deeply.
Deep locking can rob power from an escapement and sometimes stop a clock entirely.
YLH wrote about locks (and drops) for a different clock when he was... (what's before a freshman? a tween?)... a long time ago.
Let's understand locking in the escapement a bit more... deeply. Sorry.
Of the two principal components of Charlene's escapement, her anchor is a less common variety known as a half deadbeat, most often seen with Ingraham clocks.
And here is an Ingraham with a half deadbeat escapement that YLH worked on himself, in what feels like a lifetime ago.
And one of the NAWCC boys was struggling with a half deadbeat and asked about this depthing problem receiving... somewhat helpful replies.
Here's animation example of a half deadbeat.
One of the reasons that half deadbeats are less commonly seen is that they are finicky about this meshing issue. They don't like being deeply locked.
The "lock" is when the escape wheel tooth lands on the pallet. The depth of the lock is how far down along the pallet the tooth falls.
Here is the locking issue for Charlene's clock's exit wheel pallet.
When the lock is too deep the escape wheel tooth can slide off the face (beveled edge) at greater than intended velocity (because the lift angle gets very shallow when the wheel is too deeply locked).
The lift angle is the angle of bevel on the pallet impulse face that drives the power to (lifts) the pallet.
The diagram below show a model of the end of a pallet and the tooth hitting (locking with) it. In this example the pallets are above the escape wheel but the principles and action are the same.
The purple lines represent the impulse face of the pallet and its lift angle when the escape wheel tooth (blue triangle) slides along it.
See that as the pallet rises, the angle on the impulse face changes?
When locking is too deep that lift angle gets even more shallow before the tooth drops (whisks) off the end.
Anyway back to Charlene's escapement.
Locking that is too deep leads to...
Too a shallow lift angle...
And that can...
Wait for it...
Make a whooshing or snapping noise when the tooth is sliding across the impulse face.
Yep.
And with Charlene's clock the exit pallet releases when the pendulum is near the end of its swing from left to right.
Yes you recall correctly that it is near the end of the left to right swing where we hear that extra tick.
And that snapping sound would be followed almost immediately by sound of the entrance pallet locking.
Two beats.
OK it appears that we have an extremely strong candidate for this errant sound / beat / tick.
Problem solved!
Let's go get a beer!
...
...
Wait.
Wait!!
What?
Doctor clock?
Yes?
Why doesn't it make that sound going in both directions?
If the exit pallet lock is too deep, isn't the entrance pallet going to be the same?
Won't there be extra beats on both directions of swing of the pendulum?
Sigh. Not necessarily.
And such was the case with Charlene's clock.
This!
This is what really vexed YLH.
Charlene's exit pallet had teeth landing very deeply and the teeth hitting the entrance pallet landed normally... maybe a little shallow.
How is the depth of one side of the anchor adjusted / depth differently from the other?
And how to adjust one side and not screw up the other?
OK nerds. Tighten your belts.
Below is an animation of a deadbeat escapement with the anchor above the escape wheel instead of below. (A deadbeat is slightly different than a half deadbeat but for illustrative purposes they are effectively the same.)
What YLH is trying to show is how in this animation the anchor and the crutch (the rod swinging gently in the background with the anchor) are both connected precisely at the pivot point and centered directly in line with the escape wheel.
So crutch and anchor all connected in the middle at the pivot point.
With this escapement adjusting the depth of BOTHpallets would require simply raising or lowering the anchor. Both pallets would go up/down exactly the same amount, even to the escape wheel teeth.
Well...
It ain't like that with our half deadbeat friend (what a terrible thing to say) and most clocks for that matter.
First, this anchor pivot doesn't sit directly below and in-line with the center of the escape wheel.
So yes "adjusting" the height of the anchor means twisting/turning that brass pallet cock counterclockwise and that will raise the anchor...
BUT it will shift the anchor up AND to the left!
What does that do to each pallets depthing with the escape wheel teeth?
As the geometrically minded and astute reader will see from looking at the locations of the escape wheel, the anchor and the pivot relative to one another... turning that pallet pivot cock a bit counterclockwise will push the whole anchor such that the entrance pallet (right side) will be a little tighter to the escape wheel and the exit pallet (left side) a littler further from the escape wheel!
So doing that alone doesn't help.
And to further complicate things...
Charlene's crutch does not hang from the pivot point of the anchor like in the animation above. It is connected along the base of the anchor to the right of the pivot point towards the entrance pallet as do most crutches in clocks like these.
And what does THAT do to the meshing and depthing of each pallet?
Well, the good news is that YLH is not the first clock jock to run into this problem.
Doing the usual research this horological computer nerd discovered a fellow making an adjustment to a half deadbeat as he was similarly trying to address the disparity between the entrance and exit palette depths.
This is the crux of what YLH has been trying to solve.
The bad news is that lest the viewer want to mimic his solution, that videographic clock mechanic has made a risky move.
He adjusts the individual angles of the pallets by bending one pallet towards the other.
YLH mostly avoids YouTube for clock repair instruction. There are some excellent practitioners but there are a lot of hacks and troublesome practices on display.
This fellow makes certain (hamfisted) adjustments that create some of the desired changes but he unwittingly created new problems in the lift angle of those pallets.
Let's not go down that rabbit hole yet but suffice to say... right problem, wrong solution that comes with new problems.
Sometimes bending the pallet arms of the anchor is warranted, but it's an aggressive move (pallets can be ruined by bending them), and is unnecessary if one understands how the crutch can be adjusted to achieve the same thing.
The crutch!
Bending the crutch will effect the beat of a clock but with real judo moves it can also be warped such that it will change the way the two pallets engage with the escape wheel teeth.
And a combination of adjusting the pallet pivot cock and the crutch facilitates changes to each side of the anchor.
YLH will need to go back to school and get a PhD in Physical Engineering with a major in mechanical drawing to really show how all this works.
Sounds like fun (for the first week).
Meanwhile let's give it the old college try.
Curving the crutch in or out from the pendulum rod changes the location of the center of gravity of the entire crutch / anchor because they are directly connected to each other and act as a single clock part.
Here's a very typical crutch from Nana's clock a couple years back.
Quite similar to Charlene's version. Note the crutch is similarly connected to the right of the pivot.
The crutch above is not perfectly straight. Bent slightly outwards.
Now a republishing of an earlier image.
See how the crutch in Charlene's clock is not a straight line, parallel to the pendulum?
That bend is key in adjusting the center of gravity of the crutch/anchor such that when it hangs on the pivot the angle of the entire anchor changes subtly as the crutch is pulled away from the pendulum.
Any bends or changes to the crutch must be compensated such that the crutch loop will still stay wrapped around the pendulum rod and be effectively neutral in position (not pulling left or right) when at rest.
Tricky.
So the crutch must be bowed outwards and the more it thus bent the more the center of gravity shifts to the right and this results in pulling the anchor downward. This is because the pivot point of the anchor is to the left of the crutch attachment point so when the crutch center of gravity shifts right it pivots the crutch more to the left to rebalance itself and that has the effect of tilting the whole pallet so that the right side drops a bit and the left side rises a bit.
Oh and don’t forget that the crutch loop can’t shift left or right.
How to do this without tying the crutch into a Gordian knot?
Whew!
The effect is to shift the entrance pallet (right side) away from the escape wheel (decreasing its depth) and the entrance pallet (left side) toward the escape wheel (increasing its depth).
The opposite effect of adjusting the pallet pivot cock upwards.
And now coming full circle…
Our solution to this incredibly vexing problem is a very gentle dance of these two types of adjustments, a little pivot moving and a little crutch bending, facilitating an exquisitely artful way of effecting individual changes to depthing of each pallet.
Once the bulb went from dim to warm in YLH's head, he knew this would land the plane.
A lot of trial and error.
But it woiked!
Is it beer o'clock yet?
YLH needs whisky and apologies to any of you who somehow decided to stay through this insane post. Hopefully it didn't cause you to hit the unsubscribe button.
To date this was the most horologically brain-busting challenge YLH has taken on. While not exhausting every reference available, YLH researched hard and found plenty of clues but no final example of how to resolve this specific issue.
But once adjusted the clock ran like a champ.
No double ticking.
Happy, strong running clock.
Returned to Charlene.
She was thrilled.
Future posts need more sprinkles and less complex flavors.
In our last episode we discussed the roughly 100 year-old Sessions York banjo clock emblazoned with Betsy Ross on its glass tablet and an eagle sitting on top. And while in very good shape overall it was ticking and tocking in an uneven manner, what clockmakers call a "beat" problem.... What is a beat problem? Being in beat mean the clock is ticking and tocking very evenly, eg. tick-tock-tick-tock. That is a healthy beat for a clock. A tick or tock of a clock is the sound made when an escape wheel tooth hits a pallet... well... look at the gif above. You can see the escape wheel with its pointy teeth turning and gently colliding with the pallets of the pallet fork (that bent strip of grey metal). There are two pallets, one curved on top (the entrance pallet) and a second flat one on the bottom (the exit pallet). See each one hitting a tooth? That pallet fork, also called an anchor, is attached to a crutch (the brass rod that is bent and then descends down below the image)...
Charlene (name changed) called a couple weeks back. "Chip [neighbor who runs Wittpenn Antiques] gave me your name and number. I have a 'Regulator' clock that I think might be 'overwound'. Anyway it stops running after a few minutes." A few days later your local horologist headed over to Carmel for a look. Charlene was gracious and appreciative that this clock jock made house calls. And there it was in situ. A large and fine old American clock. An Ingraham timepiece (which means time only, no strike or chime; see the single winding hole on the dial just left of center). Charlene called it a Regulator, per the glass decal, and that is mostly correct as larger timepieces like this one were considered to be pretty accurate time keepers and commonly named as such. However the term Regulator ended up getting applied to a lot of types of clocks as more of a marketing gimmick. Spitballing YLH guessed it was made somewhere between 1890 and 1915. A couple things also...
Comments
Post a Comment