First Archtop Mandolin Build Thread

[Mike Conner]

After building 6 archtop guitars and 2 GOM (guitar octave mandolin), it is time to take the ultimate luthier challenge and attempt to build an F-style mandolin. Humbly submitted here will be all the gory and sad details documenting my experiences...

Background:
The MC is a wonderful resource for ideas, information and encouragement. My goal is to complete a F-style mandolin shaped object for my brother Marty Conner and share the results here.

Research:
There are several resources that have been helpful beyond all the great build threads, YouTube videos and builder web sites:

The MacRrostie F5 mandolin plan from StewMac:
http://www.stewmac.com/How-To/Bluepr...olin_Plan.html

"The Ultimate Bluegrass Mandolin Construction Manual" by Roger Siminoff is a classic reference and lots of detailed drawings:
http://www.musiciansfriend.com/books...5460x151905500

Peter Coombe has the fantastic Dudenbostel photo journal on his web site. (I've captured and saved all this content and started compiling into a Word document. Long way to go on that and wouldn't share it without permission from Lynn Dudenbostal and Peter):
http://www.petercoombe.com/Dude_cons...n/top_page.htm

Andrew Mowry has some great information on his website, including these topographical arch drawings:
http://mowrystrings.com/Mowry_F5_Contours_9_05.pdf

My Archtop Build Journal is posted here on MC and I have used or adapted many of the methods documented there in the mandolin build process. Some of the information below will therefore replicate information from my Journal, but I have tried to take new photos of the mandolin process:
https://www.mandolincafe.com/forum/s...+build+journal
I've started on a Mandolin Build Journal but that will take a while to compile and edit, so I will publish it here after the mandolin project is totally completed.

It helps to have some actual F mandos in hand to study and I was able to do detailed inspections of several instruments from friends and my brother.

The setup details were posted on MC:
https://www.mandolincafe.com/forum/s...etailed-Photos

The part that intimidates me the most is that pesky scroll!

[Bill McCall]

Very cool. I'll follow this with great interest.

And you're just down the road from Salem, my Dad's home town.

[Mike Conner]

Design and Preparation:

I am making some design changes to the neck joint especially, with a desire to use my existing dovetail jigs and hopefully get a more reliable result from a fit and finish perspective. We'll see how that works out

A detailed drawing took some time, but it is essential to get a full reference and understanding of all the features. I drew my own guided by the resources listed above, and especially focused on the head block and dovetail neck joint. I am using Weber specs as a foundation since that's what my friend Dennis James plays, and my brother likes the feel of his mandolins: 14" scale, 1-1/8" nut and 10" radius for the fretboard. All the F style mandolins seem nearly identical in the body dimensions and shape.

From the drawing I can create a template on 1/4" plywood, and I also incorporated the Mowry drawing for this. From the template I created the outside mold from 3/4" plywood and faced both sides with 1/4" plywood, so about 1-1/4" thick overall.

The leftover frame from the template is handy also.

The mold is designed to split in half at the body centerline if needed (like I used for the guitar and GOM). It got at least 6 coats of waterborne poly to keep it glue proof.

Preparing the wood:
An 8' long by 11-1/2" wide curly red maple board shall be sacrificed for this build. The template and frame make it easier to visualize and layout the features on the board. I will get enough maple to make 4 one-piece back plates and the rims and necks also. This is really great, since there will be backup pieces to cover any unrecoverable errors!


Here's the layout to get the rims and necks out of the board:


I need some thin walnut for the contrasting trim and especially the center racing stripe for the necks. I attached the donor board to a length of framing lumber using carpet tape, then guide it through the table saw. The thin section drops free on the "waste" side of the blade. I can index over about 3/16" for each pass and create strips a little under 1/8" thick:


The stack up for the neck billet looks like this:


The curly maple rims can be ripped the same way:


The 3/4" thick board can yield 4 matching rim sections:

[Mike Conner]

Head Block:
The head block design is a departure from tradition. Mahogany is the traditional choice for blocks, but I am using plain red maple from a tree from my front yard - I've got lots of it in 2.5" thick boards.

I started with 1-3/8" thick blanks, 2" wide and 7" long. The milling and assembly sequence looks like this:


The neck will be locked in with a screw tightened from the inside of the body (accessed via the endpin hole), so a through hole is drilled 9/16" down from the front face side. Then a 5/8" blind hole for a trim bushing to support the screw head.

Then the block is mounted in my dovetail jig modified with an extra fence and held in the jig with screws that drive into sections that will be cut away or hidden later:


Once the dovetail is cut, a 1" thick block is added to complete the scroll support. Drilled holes made the bandsaw cutting a little easier.

Lots of rasp and sandpaper work to clean it up.

I made about 6 practice blocks, first out of pine and then red maple, before getting a couple I deemed as keepers.

[Mike Conner]

Preparing the rims:
A cheap measuring tape makes quick work of determining how long each rim section needs to be. The long piece that wraps from the scroll to the treble side is about 27":


The rims are 1-3/8" wide. A length of framing lumber and carpet tape made it fairly easy to thin them down on the sanding station:


The beam is a little wider than the rims to make it easier to measure the thickness as we sand. Target is 0.065" to 0.070", a little thinner for the scroll section:


Bending:
My bending iron is homemade - section of 2" steel pipe with a electric fire starter as the heating element, controlled with a light dimmer and temperature measured with a 12" thermometer probe that came with a turkey fryer. I ended up making a wood support caul that matched well with the pit to help keep curly maple bending smooth:

(not sure how to have these photos come out sideways).

I tried bending the long scroll section first, lightly wetting the curly maple as I have in the past, with the temperature about 350 deg. Yep, didn't take long to break one! The wetted curly maple wants to "facet" and pop.

Revised tactics:
- 400 deg pipe
- very little water on the pipe side surface
and eventually..
- Super Soft 2 wood treatment


I also made a interior caul for the scroll bend. Got the second long piece to come out ok with the hotter temp and little water, but you can see some faceting in the photo.

Went back to the pipe a couple of more times to tighten up the bend, etc. (looks like these facets will sand out OK later).

The Super Soft 2 made the other sections come out way better:

[Mike Conner]

Blocks:
I made the tail block out of curly maple, with a pre-drilled 1/2" hole for the endpin jack since this build will have JJB piezo transducers.

We also need blocks for the points, and I am using red maple left over from making the neckblock:

It was pretty straight forward to shape and fit these blocks using the Ridgid sanding station. Don't know what I would do without it.

Rim Assembly:
I took some time to fit the neck block, and that the dovetail joint surface tightened snuggly to the mold to ensure a joint square to the centerline of the body:


Once everything seemed to fit OK, time to glue it up with Titebond original:


After the glue cured, the excess material on the neck block was removed with a jigsaw, and then sanding to smooth the blocks to the rim profile, etc.


more to come...

[Mike Conner]

Linings:
Interior "jacks" will help keep the rims tight to the mold walls as we continue. Left over plywood from making the mold and small turnbuckles went together quick:


Reverse kerf linings make the rim assembly stiffer, and I have been using them on the guitars. Dampening the "web" help them bend easier, and small spring clamps with Titebond get them on easy:

And there I have made my first major error - I didn't cut the kerfing down in width for the mandolin geometry and had used them as is. Didn't realize it until the next day. I shall accept the embarrassment and keep on going...

The point tips are traditionally bone, or ivoroid. The Weber and Breedlove has curly maple matching the rims and I like the look. Save those offcuts from the donor board! Glue oversize blocks on, then trim and sand them to shape:


Completing the Rims:
A plywood plate attached to the mold is combined with interior jacks to sand the rim edges flush. The sandpaper is 12x18" sheets from installing hardwood floors. Sequence is 36, 60, 80 and then 100 grits:

We will use this plywood again soon to hold the front and back plates for carving.

Locating pins will help keep the front and back plates aligned during later assembly. 1/8" diameter by 3/16" deep holes in the neck and tail blocks are drilled both front and back. Sections of a nail cut down with a metal cutting disk in a Dremel are sharpened to a chisel point and slipped in the holes.


Align the rim on the plates, then push firmly to mark the hole locations. Drill shallow 1/8" holes in the plates to match the rims. The nails are replace with 1/8" dowels glued in the rim holes and chamfered to smoothly mate with the plate holes:

The traditional pins are clipped off metal brads (Siminoff), but why leave metal in there to frustrate some future repairpersons by damaging their tools.

Now we can start on the front and back plates...

[Reinhardt]

Hi Mike, best of luck with your new project. That's some challenge you've taken on. I'd say F style Mandolins must be one of the hardest instruments to make. You sure look like you know what you're doing though. looking forward to following your progress.

John

[Mike Conner]

... You sure look like you know what you're doing though...
John

Thanks John, but I think you are being generous . The experience from building the guitars and GOMs helps, but I am no where near the level of the real builders here at MC. However, I really enjoy build threads, and I hope that especially newbie threads like mine help encourage others to give it a try.

We have a saying here in the US: "even a blind squirrel finds an acorn sometimes" !
//mike

[Mike Conner]

Front Plate:
I purchased several Englemann Spruce mandolin sets from Simeon Chambers (Rocky Mountain Tonewoods) as part of his going out of business sale. Really a good price, and 6 of the sets are likely from the same log.

My brother and I sorted through the sets and he picked one that is pretty fine grained (about 20 grains per inch) and nearly perfectly on quarter. Not as "creme white" as some of the other Engelmann I used for the guitars, and this has more distinct winter/summer grain lines and some darker coloration in what will become the treble side of the front. Nice tone when you tap it.

I don't think that finer grain structure, perfect quartering or even color are that critical to the final tone of the instrument. Certainly there are esthetic reasons to expect premium wood if you are building (or purchasing) a mandolin that sells for several thousand USD, but that's not where I am at as a builder and not what my brother expects. The guitars have sounded good to me even though they had wider grain, etc. Would they be even better with "premium" wood? I really couldn't say. Certainly some well respected luthiers have built great sounding instruments out of pallet wood. Anyway, I not building to market and we can just go for it.

The front plate comes as a bookmatched wedge:


The wedge opens to join at the thicker edge:


I use an "old school" joining method. I first rip the edges on the tablesaw so the width of the halves match, then rip standing up on the wide edge to reduce the thickness to about 3/4". Then, the wedges are clamped together with the outside faces out. I plane both edges together using a Stanley #6 bench plane that was my grandfathers (and about 100 years old!), set to a very fine cut:


Even if you are slightly off of square when planing, the error cancels when you re-join the plate halves. The joint line can nearly disappear, and I "candle" the joint against a bright light to ensure that there are no gaps:


I glued the front plate halves with Titebond using simple 2x4 lumber frame I made years ago, and wax paper to keep from gluing it all together. Don't clamp too tight and squeeze all the glue out!


After the glue cures I can plane both surfaces to true them up:


Back Plate:
I got 4 one piece back plates from the curly red maple board. Marty picked out the one he liked. Since the rims, neck and back are all from the same board, we should have a pretty good match for color and figure:

[Mike Conner]

Plate Arch Layout:
The Mowry drawings linked above are an excellent reference for laying out the plate arches. There are some templates in the MacCrostie drawing and the Siminoff book, but those would be useful for reference during carving.

What I do is find the center point of the plates - The widest part of the "lower bout" on a guitar. On the mandolin this ends up about 3/4" or so below the bridge line. The Mowry drawings have topographical lines in mm. In my process I work in 1/8" increments (more below), so after sizing on the copier to match the plate dimensions, I darken the lines on the drawings matching 1/4", 3/8", 1/2", etc. I then add radial lines from the center point, each 45 degrees apart (like compass points).


Measuring from the center point, in each "axis" direction, I record the distance of the intersection with each topographical line. I created a table for the front and back plates, and I'll include it in the Journal to be published later:


After measuring and marking these points on the outer surface of the plates, I can connect the dots with curves to show the topography:


I drilled 1/8" through holes on the bridge line corresponding to the inner notch of the f holes. This helps locate the bridge line later - an important reference feature.

[Mike Conner]

Plate Arch Guide Holes and Edges:
I learned from my first builds that it is really important to maintain a 1/4" plate thickness at the outside edge. For the red maple back plate, since the wood is pretty hard the easiest way to do this is to create a ledge using a rabbeting bit in a router (sorry about the fuzzy photo):

(Winston the shop dog was not impressed).

Holes drilled down from the outside surface of the plates provide a reference point for removing the material to match the arch. I have a simple gage to help set the drill press height:


Using a 1/4" bit, drill outside the lines, indexing the hieght from the drill press table an 1/8" at a time to match the topography:


The spruce front plate is more delicate (and wedge shaped), so I created the 1/4" edge reference using gouges and chisel. This is pretty easy work and goes quick:


Compared to a guitar front plate, the mandolin is rather small, and things were going along so nicely with the gouge I swtiched to an Ibex plane, then more gouge work. In about 25 minutes I had the arch mostly roughed in:


A 5" ROS with vacuum attachment smooths out all the plane marks and is pretty easy to control. Keep sanding until the dots are gone, and view from different angles to keep the arch symetrical:


Just few more minutes work with the ROS and the front plate arch is mostly complete. Lots to do on the scroll, but I figure that will come later:

[Mike Conner]

Plate Thickess:
Continuing with the front plate...

I have a simple jig to drill guide holes for the plate thickness. A chair leg tip on some plywood scrap supports the outside of the plate arch. The 1/4" drill bit is set to leave a 1/4" gap using the gage shown above:


I drill a series of holes in the inner surface of the plate. Like with the outside arch, the bottom of the holes guide the carving, and we just remove material until the dots are gone. This leaves a healthy 1/4" plate thickness. I used a Lancelot carving disc in an inexpensive side grinder to remove the overburden - a dangerous tool but not too hard to use while respecting the tool and keeping fully aware of where hands and disc are at all times!


Just 5 minutes or so with the Lancelot and the inside is roughed in, then some quick swipes with a gouge and we are ready to smooth things out. You can still see the dots at this point:


I did some work with 60 grit on the ROS to blend things together. However, I have discovered that my Ibex palm plane is too big to do the fine tuning and graduations. I tried a curved scraper but the spruce likes to tear out too much. So, the front plate will be on hold for a few days while I await delivery of a Ibex thumb plane from StewMac. About $75 with free StewMax shipping - not cheap but it will be a good tool to have for many years to come.

While I await delivery of the thumb plane, I'll probably take a whack at the red maple back plate and get the arch carved and thickness roughed in. Compared to the spruce, the red maple is hard work and my least favorite step in the build process.

[Mike Conner]

Front Plate Graduations:

I have a homemade thickness caliper (gage) assembled from a couple of inexpensive aluminum levels and a Harbor Freight 1" height gage, tied together with some scrap wood. Works fine and can reach enough to do an upright bass some day. I attached a small plaque with a fractional inch - decimal inch - mm table for easy reference:


After some more sanding to smooth the inner surface of the front plate, we can work around the plate to measure the thicknesses:


In some areas the thickness is basically where it needs to be, like where the arch decends down from the center peak. I am trying to average out the different available thickness graduations, including the StewMac drawing, the Siminoff book, various maps posted to MC and also the Arches graduation plans. Overall, I am shooting for 0.180" along the center, dipping to 0.150 roughly along the f-hole longitude and back up to about 0.200 just inside the edge.

The 10 mm size thumb plane is really good for this. I can measure the thickness and then fairly precisely shave small areas down to the target:


Once roughly at the graduation thickness, more sanding with the ROS and then by hand completes this step.

I held the plate up and tapped it as I progressed through the above steps, listening for how the tone changed. No real frame of reference here, but I do get a nice "bong' sound.

[Mike Conner]

Tone Bars:
The Englemann spruce sets came with tone bar stock, 5/16" thick x 7/8" tall x 12" long. I hand planed the stock down to 1/4" thick.

I lay out the tone bar locations, referencing the StewMac MacRostie drawing for the locations, and measuring the distance from the centerline. I marked the length and bridge line locations on the tone bar stock and then cut to length:


Holding the tone bar blank in position, the top profile is scribed to the bar using a compass set to the largest gap between the bar and front plate:

A chisel removes most of the excess material up to scribed lines.

I have attempted chalk fitting of the tone bars in guitars and found it frustrating. Instead, I created a fixture by ripped a 1/4" wide and 3/8" deep groove in some scrap wood. Removed wood at the ends, and drilled for two small brass screws to lock the tone bar blank in the fixture:


Scrap blocks align the fixture with the tone bar location and clamps can lock it in place:


A clothbacked ROS sanding disc is trapped between the bar and the front plate. Applying light pressure on the fixture and keeping the disc tight to the front plate, I can rub along the length of the tone bar. This sands the bar to conform to the top profile, even with the complex curved surface of the front plate.:


Working through 60 and then 80 grit discs, it doesn't take too long to get a pretty good fit. The bridge line marked on the blank helps to keep the tone bar on location as we sand:


Progressing to 100 grit, and then touching up the bar and the front plate with 220 grit we get a nearly perfect fit.


A nice bead of Titebond and we can use the fixture to clamp the bar in place. After the glue cures we can relocated the fixture to do the other tone bar:

[Mike Conner]

Completing Tone Bars:
Now we are ready to carve the final profile of the tone bars. Tapping the plate now, the "bong" is gone and we get more of a "plonk" sound. Painters tape at the ends of the bars help protect the inner surface of the plate:


For some unexplainable reason, carving the tone bars is my favorite part of building. It's very sculptural and looks really cool, but is relatively easy to do.

The target is a 0.300" height at the bridge line, tapering to almost nothing at the ends. The chisel makes the work go quick but controlled. The bass side tone bar is nearly ready:


After carving the height on both bars, the center section is carved to a triangular profile, flaring to flat to the ends:


Tapped, the "bong" is back and maybe there is a little more sustain. Sounds really good to my curious ears.

Soon I can cut out the f-holes and then bring the details in on the front plate. Need to start on the back plate after that...

[mikesbro]

Hey Mike, I've been watching your progress and documentation. Awesome. I would still be staring at a chunk of wood afraid to get a chisel near it. I didn't realize the tone bars were so thick, my guess would have been half that height at bridge line.

[Mike Conner]

... I didn't realize the tone bars were so thick, my guess would have been half that height at bridge line.

Hej Bro,
I am using this Dudenbostel information as a reference:
http://www.petercoombe.com/Dude_construction/page4.htm

No experience with this, and I have read that the Englemann may not be as stiff. The Dudenbostel thread is showing Englemann for the Chris Thile mandolin. The 0.300" height seem consistent with how the Weber and Breedlove tone bars look. We'll know when the strings go on ;-)

[Mike Conner]

(Finally back to the mandolin build after a break to do some real life - travel, played at a really sweet wedding, etc.)

Finishing Up the Front Plate:

The scroll is roughed cut using a Ryobi scroll saw:


Holes drilled in the front plate allow the scroll saw blade to pass through for cutting out the F holes. Going slow and careful, the holes provide safe places to turn or adjust. I can get really pretty close to the final shape:


Needle files clean up the cuts:


Surgical gauze glued with Titebond to reinforce the F hole perimeter. Not very happy with how this came out. Is it really worth it?


While the glue cures some more I trimmed the front plate edge to the rims:


The linen was trimmed from the F holes using a very sharp Exacto knife. A very tedious and meticulous process:


Starting the Back Plate:
The red maple back plate arch is roughed in using the Lancelot:

[Marty Jacobson]

I don't think the surgical gauze is worth it. I sanded it off on my first one after I tried it. If I feel like I want to add reinforcement to keep splits from occurring, such as with my pointy f-holes, I put a cross-grain strip of spruce about 15mm long, 3mm wide, and 2mm thick at that area.