top of page

Making Low Amplitude Rosettes

Ever since I saw a rose engine in the 1980s I wanted one. With it I wanted to do the work I had seen from the 16th & 17th centuries. Years later I came to own a mid 19th c. French rose engine that had come out of Cartier’s workshop in Paris early in the early 1900s. Someday I will post the story of the adventure of how I came to own it. It was made by Antoine Duguet, Mécanicien, about 1835.

Rose engine

The style of pieces I wanted to make had very few “bumps” around their outside. A rose engine is a type of lathe that a cam (rosette) controls the movement of the work mimicking the shape of the particular rosette the machine is set for. Kind of like the Spirograph, a popular kid’s toy. The machine I owned had many bumps on each rosette, sometimes over a 100. The work I wanted to do needed rosettes with as few as 8 bumps. If you take a circle, which is the shape of a rosette, and divide it by 8 flat spots, the height between the high and low spots is called amplitude. For over a 150 years the problem of how to adjust amplitude on a rose engine had troubled many. In the ‘90s my friend Fred Armbruster solved the problem followed by a few others. Their solution would not work on the style of French machine I owned without major modifications, something that was not an option on a historic machine.

Rose engine apart

I had to take apart the rose engine to replace the rosettes. This would be totally reversible and do no harm to the machine.  What a treat that was, the craftsmanship and skill Duguet made the machine with is amazing.


This was made at a time before standard size screw threads. Each screw is numbered for the hole it is made to fit.


To do miniature work I needed rosettes with .010” to .030” amplitude. On a 4” diameter rosette that small amount would barely be visible. I’m old school, I do not use any computer graphics, CNC or any other modern ways. One morning I came up with the idea for my “Amplitude Predictor” which started out as a few scraps of wood with a pen drawing on a MFD test rosette.


By the time I started building it I realized I need to control the advancement of the pen, so I added some aluminum angle. Then a fine adjuster, then an adjustment to simulate the cutter above or below center, then different sizes of rubbers. And finally a dial indicator to measure the amplitude.


With this I could work out the design on a MDF disc. Laying it out with a protractor and dividers just as Charles Plumier (1648-1704) wrote about in his 1701 book, L’Art du Tourner. I just love to use period tools and books, I seem to get some kind of energy from their soul. Then shaping it with a sander or file, I could tweak the shapes of the curves and experiment. With double stick tape I could change the paper easily that the test design was drawn on.


For the brass rosettes I machined some large, at least for me, chunks of brass. This was from a box of scrap I bought over 30 years ago for $5, a tiny fraction of what this much brass would cost today. The pieces, about 4 1/2” x 1 1/2” with a hole to one side so they had to be turned true in a 4 jaw chuck. Then bored to a .001” accuracy.

milling 2

The brass rosette blanks were then mounted to holding fixture, that I also had to make, mounted on the rotary table. This was set on my Aciera F-3 milling machine. Accurate degree readings for each “bump” were calculated and duplicated with the rotary table. The radius of a concave “bump” was set by a fly cutter. Flats were milled with an end mill.


Some of the large radius were laid out with dividers and the line darkened. Some of these were only a few thousandths of an inch from the edge.


Using a WWII vintage die filing machine the rosettes were filed to shape.


The rosettes were put back in the “Amplitude Predictor” for final tweaking.


The finished low amplitude rosettes with test design drawings.

boring alum

I machined some spacers to perfect size from aluminum. Funny thing, when I bored the hole the spiral curly cue cutoff went forward of the boring bar and through the hollow spindle, spilling out the left end of the lathe.


The rosettes assembled on the barrel of the rose engine.


Before I took my first test cut I poured a glass of wine to toast Monsieur Antoine Duguet and thank him for creating such a wonderful machine. I wonder if in his wildest dreams he ever thought the rose engine he made would be used 185 years later in a far off land?


The first set of test samples along with the drawings. The working range for these low amplitude rosettes will be in the 3/16” to 5/8” diameter range. After over 20 years of thinking about this I have finally done it, now I just hope it doesn’t take 20 more years to make the miniature turnings I have dreamed about so long.

bottom of page