Sunday, April 23, 2017

Rosewood Seth Thomas (Thomaston) Ogee Clock - Part 2

Apparently I've forgotten all about this series (there will be at least 3, maybe 4 parts to this clock restoration), so here is an update on this clock (which is currently finished and running quite nicely).

The previous part 1 is here:

The only part of this clock that I had posted about so far was the restoration of the movement. This is a somewhat in-between to late model Seth Thomas ogee. It has a Thomaston label, but a Plymouth Hollow movement. I would guess that it's as late as the 1870s. This would mean that it would have a decal style glass (fairly detailed/printed). Probably flowers or a Victorian style scene. However, I wanted to install a somewhat earlier stenciled tablet. They are nicer to look at, and they look more convincing that buying a reproduction tablet from a supplier.

For this tablet I dug into my Fenn Stencil books and picked out a design that I've seen used on Seth Thomas clocks. I've seen this pattern on column and cornice clocks. It is also similar in style to a few other patterns of the period. The nice part of this specific pattern is that I had photos of a popular colour scheme for it which looked quite nice. Largely dusty grey and red.


A similar tablet in black and gold:

Ives St HillsG

Two ogee clocks with the blue/gold/black theme:

Ogees 01

This is more or less the exact tablet I wanted to copy (as far as colours go):

ST Column1

Different pattern, same colours:

Terry Andrews Four-Finial Steeple (Sold 1100 USD) 01

I cut all my stencils from chemical-proof plastic. Anything that won't melt or dissolve in lacquer thinner. I have used thin plastic file folders as well as Mylar for this. This was actually a stencil that I had already cut a few years ago. Initially the size of the pattern was a bit smaller than I wanted, but I assumed it would look alright. I went ahead and stenciled the glass with bronze powders using my usual method (see Mirror Clock Project).


The finished glass didn't look TOO BAD, but I hated how the blue-grey turned out. It was far too blue, and too bright.


I decided to scrap it and restart. I took the glass out, stripped all the paint off, and I decided to enlarge the pattern also. These patterns are incredibly tedious and difficult to cut, so on something like this I actually only cut HALF the pattern, and I stencil it in 2 parts. All tiny dots are painted in freehand.


Stripping off all my hard work:


Here's the new stencil before cleanup. You can see some of the thin streaks of varnish across the glass. The arrows point to all the goof areas that need to be cleaned up. I use a toothpick to just scrub the glass clean in all these spots and to clean up any rough lines.


Before cleaning up (note centre of flower):


New bouquet stencil. You can probably see where a lot of the dots are missing (centre of flowers and ends of stems etc).


Bouquet and red border done. At this point I really didn't want to mess up the grey so I did over a dozen colour samples.


Here's the finished tablet. The gold pattern sort of disappears depending on the viewing angle, but it turned out beautifully and I'm very happy with it.




Month-Going Longcase Clock Project - Updated Dial Photos

Here are a few up to date photos showing details of the dial. In the last post the dial and hands were largely finished, but I had not yet done the chamfer on the winding square hole, or polished the dial plate.


Here you can clearly see the detail of the winding hole as well as the end of the brass screw fixing the spandrel in place.


Here is what the back of the dial looks like. I plan to completely blacken and antique it at some point. The backs of dials are never normally cleaned, so they are usually completely dark and brown with oxidation. Alternatively they are painted brown or black on painted dials. In either case the reverse is usually very dark, and the clean(ish) brass doesn't look right.


Wednesday, March 15, 2017

Month-Going Longcase Clock Project Part 5 - Custom Longcase Hands

Cutting longcase clock hands always seems to be a popular subject. Everyone loves to see raw steel turned into beautifully cut curlicues and then turned a beautiful blue-black.

This is one of those "essential skills" that most clockmakers should learn, but that very few do. There are a few reasons for this. Normally it's a time/money issue. Hand cutting and filing steel longcase hands is a time consuming project, and most clockmakers would prefer to buy laser cut or mass produced replacements rather than spend the time and effort on custom hands. It's also expensive for the clients. Who wants to spend 100$ or more on such small parts? Then again, depending on the clock, you wouldn't want precision-cut copies. When you get into rare early pieces, the value of the clock justifies the price for custom work.

For someone like me, I do this work for fun, so I spend whatever amount of time is needed for the work I'm doing (because it's for me). That said, I will happily cut hands for a client so long as I'm adequately compensated for the work.

Anyhow, I think I'm starting to ramble, so on with it.

Materials Needed for Longcase Hands:

- 1/16" thick steel with a carbon content for bluing (test the steel if needed as not all steel will blue nicely) steel must also be annealed for easy cutting
- Paper pattern
- Saw blades*
- Sand paper and polishes


- Jeweler's saw
- Files
- Alcohol burner or small torch
- A bluing tray filled with brass shavings, OR a cast iron skilled filled with fine sand, OR a thick brass plate
- Small vise
- Saw support (a wooden plank with a V sawn on the end)

Step 1: Pattern

Choosing the correct pattern is probably going to be the most important, and the most difficult part of the entire process. If you already know the exact pattern that you need, then give yourself a little pat on the back, and count yourself lucky. If you DON'T know which pattern to use, this is where you will have to do your research.

When I last talked at length about clock hands, it was on the mirror clock project, here: All the same points still apply when dealing with longcase clocks. The wrong hands can seriously affect the overall look and feel of the dial and of the clock. Hands for the period I'm copying (1680s, but with more of a 1740s chapter ring) vary WIDELY in style, shape, and size. It's also important to note that certain style hands look better on certain style dials. There is a huge variety of early square brass dials. Some have large ringed winding holes, wide or narrow chapter rings, centre engraving, seconds bits, etc. Everything needs to be taken into account.

It's also partially left to personal taste. I find that a fair number of clock hands are "ugly" or disproportioned in one way or another. The hour hand may be far too large compared to the hour hand, or I find a certain shape too pointy, too ornate, too plain, etc. Anyhow, after probably a month of debating, I narrowed it down to a particular pair of hands that I found on three similar clocks.

John Knibb London, Circa 1685
John Knibb London Ca 1685 8 Day (ht 6ft 2in, 10in dial) 06

Joseph Knibb Month Duration Longcase with Roman Striking, Circa 1685
Knibb Month Roman Striking Longcase ca. 1685 03

The third clock (not pictured) is another Joseph Knibb ebony longcase, month duration with Roman striking, and also circa 1685. This one was sold in the "Masterpieces From The Time Museum" sale back in 2004. The case is similar to what I want to build (caddy top, 3 finials, ebony case with no trunk door lenticle, but this one has flat Doric hood columns and I want Barley Twist ones).

The pattern was traced freehand on paper based on the dimensions I needed for my dial.


Step 2: Sawing

The rest isn't rocket science. You saw the hands with a jeweler's saw. All the interior openings have to be cut with a pilot hole to start. I find it easiest to start with the openings. Try to stay outside the lines.


Here you can see my cutting support block. The end has a sort of keyhole shaped opening.



I have yet to find a great way to avoid breaking lots of blades. The only advice I can give is to never push against the blade. Let the blade do the work, and turn corners slowly. You can lubricate the blades with wax or oil, but this may ruin your paper template. The blades also become dull from use as you're cutting. *As a general rule, you need 3 teeth for the thickness of material you're cutting. The thinner the material, the more teeth you need. I believe it's 48 teeth per inch for 1/16 thick stock, but I found that slightly finer blades cut a bit more easily. Experiment and see what works. Luckily jeweler's saw blades are very inexpensive. All these broken blades probably cost about 1-2$ for the lot.


Now, no matter how careful you think you're cutting, the results will probably look something like this:


It's not great, but everything else can be cleaned up with files.

Here is the minute hand. Note that I'm using the other end of my saw support block (which has a narrower key slot).


Here are both rough cut hand blanks:


Step 3: Filing

It's hard to explain filing. I think it's pretty self explanatory. Remove bumps, clean up the lines and shapes, thin areas as needed, and then add any necessary "carved" areas such as leaf grooves, decorative ribbing, etc. Use any file size or shape that's convenient. I used maybe a dozen different ones.



I gave them a fairly smooth polish, but I also didn't go too crazy trying to get beautiful perfect edges. I also didn't completely remove some of the steel's texture from the annealing process. I want them to look 200 years old.


Step 4: Bluing

Bluing hands (or other small steel parts) can be a tricky process. The important part is even heat distribution. There are several ways you can blue hands, and some are easier than others. The traditional method is to heat them in a small metal tray filled with brass shavings. This is done over a flame (usually a blue alcohol flame). Alternatively, you can also use a thick piece of brass, and heat the hands slowly (do one at a time). You can also blue hands very slowly in a bed of hot sand. This is how I blued these hands. I used an old cast iron skillet with fine white sand. This method is very slow, but you have the best chance of success.

If you overheat the steel, you will pass blue and go back to straw. If you overheat the tips of the hands, the colour won't be even. If you screw up, just re-polish the whole hand back to bright steel and try again.

The hands are then quenched in oil or water.

If you want more of a black-blue, and you don't want to mess with heat, you can use gun blue. If you have a particularly odd blend of steel that won't blue correctly or evenly, gun blue can also save you. This was the case for a hand I made several years ago.




Apparently I still had not polished the dial plate at this point. I will have to take more photos later.

Month-Going Longcase Clock Project Part 4 - Dial Plate

I have had the dial and hands for the clock finished for quite a while, and I even have the photos ready to post, but I've been procrastinating. Since I had the afternoon free, I thought today would be a good day to tackle these two posts.

As you may know, this is an ongoing series on my clock blog, where I am attempting to put together a Knibb style ebony longcase clock (reproduction) using spare and orphaned parts, along with new/replacement ones.

You can view the previous parts here:

Before getting into the details about making the dial, there's an additional photo that I didn't know where to include. When I was restoring this movement, it was evident that it had gone through MANY repairs over the past century or so. The movement had about a half dozen or more added holes in the plates. Some are a complete mystery. On the backplate photo below, you can see what appears to be a pair of holes for a bell stand on the left. I have no idea if the original bell stand was on the left, or if the original location was on the right, and a different bell stand was installed in new holes later (or vice versa). There was also another hole behind the current (new) bell stand. Regardless, the larger hole to the upper left had been plugged previously, but not the smaller pin locator hole below. Additionally, the bridge has been moved, and the old holes had been left (on the bridge and in the plate), so I plugged those as well. There are 4 small holes (two per plate) on the top and bottom centre edge, and these are normal to find; they are alignment pins from the original construction of the movement. I left those alone.


I will also need to address all the mismatched screws on this movement. It would seem that most of the screws (including the motion works bridge screws on the front) are mis-matched, and no two seem the same. A few also have different threading, which makes it a pain to keep track. I don't love the idea of having to hand-make an entirely new set of screws, but I would prefer that they all match, especially since this will ultimately (and hopefully) be a rather nice clock with a somewhat rare movement.

Fitting a Replacement Dial

Fitting a new dial might seem both easy, and complicated. In truth, it's a mix of both. It's easy because you get to custom fit the dial exactly how you like, and it's difficult because you have to be quite precise and careful.

The materials you will need for a traditional brass dial are as follows:

- A set of pillars (or alternatively brass, bronze, or steel rod stock)
- A set of 4 cast brass spandrels
- An engraved, etched, or CNC carved chapter ring
- A brass sheet*
- Taper pins to attach various elements
- Brass or steel machine screws (as needed), or specialty spandrel bolts
- Lacquer and black sealing wax
- Silvering powder

Tools needed:

- Lathe (to turn/trim pillars)
- Drill press and drill bits (or a hand drill)
- Countersink or chamfer bit to deburr holes
- Files
- Hacksaw
- Sandpaper, buffing compounds and/or steel wool
- Jeweler's saw

Step 1: Dial Plate

Cut a square sheet of brass. This is quite simple and straightforward to do, and I simply used a hacksaw and files. The size will depend on the type of clock or the type of dial you need, OR on the chapter ring you are using. *The thickness of the dial sheet you want to use is largely left to preference. Some old dials had paper thin brass sheets, while some are nearly 1/16" thick and weigh several pounds. My dial plate was cut from 0.035" sheet (0.95mm).

The Knibb clock I'd like to copy is from around 1680-1690. Most clocks from this early period had 10 inch dials. One of the chapter rings I had on hand (a gift from my friend and mentor Jim Dubois) is from an early American longcase clock, probably from the mid-to-late 1700s (maybe 1760s). It does NOT match the style or layout of a 1680s clock because of the large Arabic numerals on the rim, but it is just over 9 inches wide, and perfectly suited for a 10 inch dial plate. It is very unusual for such a late clock to have such a small dial. Normally dials increased to about 11 inches as early as 1700-1705, with most dials growing to 12 inches or more by the mid 1700s.

Another nice feature of the salvaged chapter ring is that it is umarked (no maker's signature). This is especially nice for pairing with an anonymous movement. Early dials typically had the signature on the dial plate edge, or on a circular boss in an arch above the dial.


In the photo above, you can see the original dial plate for the engraved chapter ring. It was a pretty crude "cartwheel" style dial with a calendar opening. Someone had painted the sheet white at one point.

The hardest thing to find for a replacement dial will be the engraved chapter ring. As noted in the parts list, there are other options available. You can create one with the use of a CNC, or by electrolytic engraving. I have also occasionally found spare (orphaned) chapter rings and dial parts on eBay. I have a very beautiful one likely from a single hand country longcase, also for a 10 inch dial. Unfortunately that one has a signature, but the price was right.

If you're looking for a hand-engraved custom dial, you will need to do some searching. There are a lot of engravers still around, but the ones specializing in clock dials have largely disappeared. Be prepared to pay a premium for hand engraved work. A dial ring (about 10 years ago) was priced at at least 300$.

Step 2: Dial Feet

Dial feet can be purchased, but the size and quality may not be as great as what you expect. I made simple rod feet for this dial. Nicer feet would have a nice cup flange against the back of the dial plate, but I don't have the lathe tooling for this kind of work. Since each dial foot hole on my clock plate was a slightly different size, I had to make each post a custom fit. This was a bit annoying, and the fit doesn't necessarily need to be that precise.


You will want the end of the foot to have a rounded profile (or a triangular point), and you want to make sure that the end is long enough to pass through the plate with a minimum of about 1/8" protruding inside the movement (for the taper pin). You also want to have a decent amount of brass stem protruding from the top for riveting to the dial plate.


Locating the position of the holes is a tricky operation. You need to ensure that the hands will be perfectly centered on the dial plate. One way to mark the pillar positions is to mark the dial centre on the back of the dial plate, then use the front plate of the movement (with all the posts and sundries removed) and transfer the locations to the back. I believe this is how I did mine (it's been months so I don't remember exactly).

Once ALL the holes are drilled (including the holes for the chapter ring and spandrels, give the pillar holes a light chamfer on the front (top), and rivet them in place.


This shows one of the riveted pillars up close.



Everything fitted temporarily.


To attach the spandrels (which I glossed-over) they will need to be drilled and tapped for a machine screw or a spandrel bolt. Spandrel bolts are simply machine screws with a squared head. The typical location for the hole in the casting is below the cherub's heads (on this pattern). A lot of old longcase clocks use steel screws, but I chose brass so that the screws don't show as much in the spandrels. The spandrels are then screwed in place from the back. In the photo above, the screws are not in the spandrels, and you can see them on the table.

In many longcase dials, some of the dial feet will be within the centre field of the dial. This is usually due to clearance issues, but in this particular case, there is no reason why the two lower dial feet couldn't have been positioned lower on the front plate. If you go back and look above, you will also see that the dial pillars are all randomly spaced on the front plate. The upper right one is near the edge of the plate, and the lower left one is higher and farther to the left than the other pillar. I have no idea why the pillars weren't equally spaced. This is a time only movement with no obstructions anywhere near the edges of the plates. What likely happened here is that the maker had a pattern already in use for a standard 8 day movement, and the pillars normally had to be in those locations.

Anyhow, because some of the pillars land within the centre, they need to be filed and smoothed flat. A simple way to get the bulk down flat is to punch a hole in plastic sheet, and rub a file over the rivet until you start to file through the plastic. This will protect most of the plate (which you don't want to scratch, especially along the edges).


Finish the filing as carefully as you can. You don't need to be extremely precise because the centre will be matted.


Step 3: Matting

Traditional dials are frosted in the centre with specialty tools such as a "frosting punch" or roller. This is a tool that is not easy to find or replicate. I had made a previous dial using a stack of hacksaw blades:


This is used in combination with a hammer. The matting must be done over a hard surface with the dial clamped flat. This is to avoid deforming the brass (as much as possible). If you're too aggressive, the dial will dish. You will also note that I started by going over the dial centre first with sand paper to get a scratchy "matte" base.

I had used a spruce 2x8 here, but this was NOT adequate (too soft) and I got some dishing on my dial.


Step 4: Waxing and Silvering

I didn't take any photos of this step. I thought I had already posted a tutorial on the subject, but apparently not. Instead I will differ you to this tutorial:


For the centre hole, you will want to cut it undersized and file it to enlarge it. You want about 1/16" or less around the perimeter. Same for the winding square hole(s).

At this point I wasn't too happy with the matting. I had done this same matting on another dial and liked the look, but it's possible this brass was a bit harder. I decided to try an electric engraving tool to see what it might look like. Here you can see a small spot done with the electric engraver tool.



I decided to do the whole dial with the engraver. I did not, however, go about it in the best way. I intended to do it in sections, and go over it in a criss-cross fashion. The effect left way too much of a striped effect.


I can't even tell you how much time I spent on this. In the end, I had to go over the entire thing in random squiggles in all directions. I'm still not entirely thrilled with it, but it looks better than the frosting punch (hacksaw) method above. It is VERY time consuming either way. Another option is sand blasting, which gives a pretty good effect depending what blasting material is used.


Step 5: Cleaning and Polishing

Once you're all done, a light chamfer can be added around the winding square hole, and the entire dial plate can be polished (including over the matting). I had not done this yet in the photo above. You can still see a spot at the 6 o'clock position, and below the upper right spandrel.

Cutting the longcase hands will be in its own separate entry (coming up next).