Trackerbar tubing

After getting the trackerbar stained with a "golden pecan" color and 3 coats of shellac, I decided I'd try to put some more brass into the build.

Since I'm planning on having a hinged top panel between the pressure box and the pipes, I thought I'd tidy up the wad of tubing that connects the trackerbar to the pipe nipples. I gathered up the 1/4" rubber tubing I had, and after seeing what 15 pieces of this looked like, and how much space it consumed, I wanted to go for a more streamlined arrangement. I searched the web and YouTube and found all kinds of info on bending brass tubing without flattening it or giving it an oval profile in the bends. After trying a couple of test pieces, I came up with a method that seemed to work fine with the tools that I had.

First off, after calculating and cutting the pieces to length, I annealed the brass - first with my propane torch, then after that ran empty, I used the gas stove.

 The idea is to get the brass tubing just to the point where it's faintly glowing red near the area to be bent. It helps to do this at night, with the kitchen light turned off, so the faint glow is easier to detect. After getting to the desired temperature, I then ran the tubing under the faucet to cool it off so I could handle it and anneal the other end.

Of course this leaves the brass discolored, so after the annealing process, I gave all the tubes a good sanding with some 1000 grit sandpaper. It's much easier to polish them up before they have bends in them.

Next I prepared the tube for bending by crimping one end closed with a pair of vise-grips

Then I filled the tube with table salt, making sure it was well tamped in and filled the entire tube. It's important that there is no moisture in the tube, so it's good to let them dry after their post-annealing quenching.

Once the tube was full of salt, I pinched closed the open end with the vise-grips and wound up with this:

Having calculated the measurements for the length and the reference points for the bend radius, I put the tube in my tubing bender, lined up the mark and put a 90 degree bend in it. The tubing bender I bought was a cheap $9 model from Harbor Freight, and I found that giving the tube and the tool a shot of WD40 helped the bender from dragging on the brass too much.

This left the bends with a nice circular cross section all the way around the bend. I wanted to make sure the air path was not obstructed by kinks or sharp turns.( I'm guessing that the travel through the brass tubing will actually prove a little more efficient than the airflow through the rubber tubing, since the brass is smooth inside and should cause less turbulent interference along the way.)

Here's a view of what it looks like from the bottom of the tracker bar. The 5 larger pipes with their 5/16" nipples and tubing will remain as in the plans.

And here's a view from the top. the array of brass tubes is about 2 1/2" wide at the point where it will exit the pressure box side. At the back of the pipe board, the black rubber tubing will vertically connect the the brass tubes with the pipe nipples.

Tracker bar serial number graphic

One more addition. Name, rank and serial number.

Tracker bar notes

Never wanting to leave well enough alone, I got to thinking of ways I could "fancy up" the tracker bar before I stained and finished it.

I thought it would be nice to have little note letters next to their corresponding holes, both as a useful index and as a way to see if I could accomplish "printing" something onto a wood surface.

I worked up the note scale in Photoshop and tried using a bunch of different fonts to see which one would be best. Since my organ is destined to have a "vintage" look, I found fonts that would be both easy to read and yet somewhat old-fashioned.

After getting everything the right size during several test printings on plain paper, I got a sheet of 3M Multipurpose Transparency Film (CG6000) and ran the lettering through my inkjet printer, making sure to specify settings in the print dialog window such as "Flip Horizontal" (to get a reversed image), 600dpi resolution, best print quality, grayscale, etc.

This particular kind of transparency film is made for both laser and inkjet printers. Even though I was using an inkjet, I printed on the smoother/glossier laser printer side, NOT the rougher inkjet side. My testing indicated that I needed to print on the smoothest surface.

Here's how it looked after it came out of the printer.

At this stage, I found it very important to let the film dry for a least an hour or two, even overnight was OK. Trying to apply the lettering too soon will result in smudging. (Fortunately any accidents can be quickly fixed with a bit of sandpaper.)

Being impatient to see the results, I cut out a strip of letters for the smallest 17 notes. The larger 3 notes had a bigger font size, and it would take longer for the ink to dry enough to keep from smudging.

I carefully aligned the strip with the trackerbar surface, and with the ink-side down against the wood, I wrapped it all up in Scotch tape, being careful not to let the strip slide and smudge.

I then used a spoon to rub/burnish the lettering, making sure to move it repeatedly around each area from all different directions. I used the side of a screwdriver shaft to rub the little "flat" symbols, just to be overly safe.

 Here's how it looked when I pulled the film off. You can see that there's very little ink left on the film.

The font I finally decided on was "Maximilian Antiqua", which has a slight hairline shadow on the leading edge of some of the letters. I was happy that this was still defined. Perhaps all that sanding with 1500 grit sandpaper helped.

These are the test pieces I used in my preliminary test. I also tried a couple of different stain colors to see which I liked best, and to see if the stain would cause the inkjet ink to smear - which it did not. I finally got some use out of my failed tracker bar attempts after all.

Here's the final, yet unstained result. I think I'll let the ink dry for a few days before I stain and finish it.

Tracker Bar Plan "B"

 As my first attempt at this ended in failure (I widened out the holes too much), I decided to have another go at it today.

I used the same setup as before, using my digital caliper jig to get the holes drilled accurately.

I originally planned to do the base of the tracker bar using John Smith's honeycomb method, but advice from Melvyn Wright ( http://www.melright.com/busker/ ) pointed me at using an all-wood method, described in a post by Walt Lysack on Melvyn's website. After studying Lysack's post I decided that I'd do a slight variation on his method.

 I cut a length of hardwood about 7/8" x 1" to use as the top and center, and a piece 1/4" x 1 3/4" to use for the tubing connection base.

I drilled the center holes about 1/2" deep into the block, then sliced off the tracker bar top.

I then took the middle section (with its half-deep centerline holes) and glued it to the undrilled base piece.

I then drilled all of the offset holes through the middle piece/base assembly, again using my caliper jig.

This is what I wound up with:

And here's how the pieces looked once I finished with the Dremel tool spiral blade and some sandpaper:

Trackerbar pt. 1

 I started work on the trackerbar today. I first used my steel rule and marked off the position of the holes on a piece of paper. Just to be safe, I double checked it against the Melvyn Wright test roll holes to make sure everything lined up. I then made a photocopy of the template (just in case) and I cut it out and glued it to the top of a 1" x 3/8" length of the mystery hardwood that I've got so much of.

I made a jig for drilling on the drill press and figured out a way to attach my digital calipers. I first drilled all of the holes with a 1/16" bit to serve as a pilot hole, and carefully double checked my measurements with the caliper jig as I moved the piece along. Rather than "zeroing out" the calipers after each hole, I went with cumulative multiples of 6mm as I went down the row (9mm for the last 3 holes).
(On a test piece of wood that I had tried earlier, I zeroed out the calipers after every hole, and the minor variances added up along the way and threw the measurements way off.)

After I had all of the holes drilled to their proper size (5/32"/3.96mm for the first 17 and a "C"/6.14mm drill bit for the big three), I stood the piece on its side and cut it in half.

This left me with two identical trackerbar pieces, which I figured I might need if I really screwed the first one up somehow. As it turns out, the half that was originally on the bottom during my drilling, was pretty fouled up on the bottom side due to splintering from the drill bit. I could still use it in an emergency, but it would require a lot of sanding to make it sound.

Next, I set to work with my needle files and squared of the leading edge of the holes (after double and triple checking John Smith's instructional video to make sure I had them pointing the right way.)

Then I double and triple checked the video again and marked off the back of the piece so I wouldn't lose track of which holes widened in which direction.

To gouge out the back of the holes, I used a rotary cutting bit on a dremel tool and worked my way along very slowly and carefully.

 Here's what it looked like after all of the machining and filing.

I found that a rolled up scrap of 220 sandpaper made a good file for getting the little fuzzy bits out of the holes.

Easy fake brass finish

Thinking that I should probably give my steel crankshaft parts a coat of shellac to keep them from rusting, I thought I'd try an experiment with a mixture of shellac and a few drops of yellow food coloring.

I brushed two coats on a scrap of aluminum rod, one coat on the top of a razor blade, and two coats on part of a chrome-plated shaft, a ball bearing housing (avoiding getting shellac inside), a piece of steel from a ratchet, and part of a slotted screwdriver shaft.  I'm not sure how well the colorant will resist fading - especially prolonged sunlight exposure, but for the interior organ parts I may give it a go.

Back to the crankshaft

Now that I've got the reservoir/bellows assembly fairly complete, I figured I'd better return to my work on the crankshaft, since my first attempt with it failed so miserably.

I originally attempted to make the crank arms out of aluminum, but those turned out to be too soft, and the holes for the shaft widened out enough to make the shaft wobbly and useless. So I repeated the machining process, this time from a solid piece of 1/2" x 1/4" steel bar. This time, to ensure a tight fit, I drilled the shaft holes with a "D" sized drill bit (0.246"), which is slightly undersized for the 1/4" shaft. The shaft wouldn't quite fit into the holes, so I drilled them a little on both sides with a 1/4" bit, leaving about 2 or 3 mm. in the center at the smaller "D" size bore. I then polished the steel up so it would look as shiny as the chrome plated shaft steel, using 220 then 1000 then 1500 grit wet sanding, and finishing them to a near mirror finish with some white rouge polishing compound and a felt wheel on the dremel tool.

I then hammered the shaft sections into the crank arms and got a super tight and square fit for all the pieces. Even though there was no lateral or axial movement, I decided to play it safe and drill completely through the arms and shafts and tap in some 3/32" spring pins. I put a drop of loctite in each of the spring pin holes before inserting the pins. I would have used loctite on the shafts ends, but they were in so tight that the loctite couldn't find a gap to penetrate.

Here are most of the part laids out. I haven't attached the single arm/short shaft/rod end to the end yet, as I'll likely wait until I have it installed through the wooden panels. I'll also wait until I've got the whole pressure box assembly built, so I can tell how much to trim off the end of the shaft where the crank would traditionally go, since I don't want that part of the shaft protruding from the case.

Centerboard Valves and Reservoir Test

Last night, I finished up the bellows/reservoir box (except for the spring fittings). The centerboard valves still needed to be done, so I laid the two pieces of 7" x 1 1/4" blackout cloth over the rows of holes, and glued down a 1/8" x 1/8" strip between the valves to prevent any possibility of a cross draft from one valve that might disturb the seal of the other. I secured each end of the valve cloths with a spot of glue then fixed a couple of hardwood strips at each end to keep the valves and the center strip in place.

Here's a wider view of the completed valves

Then I put the self-adhesive foam weatherstripping around the base of the reservoir box, staggering the seams at the corners. I doubled the width of the weatherstrip, which left me about 1/16" too wide, but I decided that it was OK if there was a little excess on the interior side of the box.

I wasn't enthused with the grey foam, so once it was all stuck on, I decided to color the edges black

I used a Sharpie for the edges, which had the added benefit of hiding the rough seams somewhat

I think the black foam wound up looking less conspicuous, and once it was compressed, it looked like black felt more than foam.

I realized that I needed screws longer than the 3/4" No. 6 brass wood screws I had on hand, if I were to penetrate the plywood of the centerboard and get down to the hardwood strips below (Except for the side nearest the pipes, which has no hardwood below it). At first I was determined to make a trip to the hardware store for some 1" screws, but I got impatient and decided to use my 3/4" screws and countersink the screw holes in the battens by 1/4" deep using a forstner bit. I got everything put together and then gave it a test using a 2.5lb (1.1kg) weight to see how fast the reservoir emptied.

Here's a short video clip of the test:

Relief Valve and Reservoir Outlet

Continuing work on the reservoir box relief valve. I trimmed the base of the hinge and worked out a way to have it serve as the base of the spring as well. I attached the relief valve trigger arm to its base with a countersunk brass screw and some wood glue.

Then I stained everything with a mahogany stain, and made the valve seal out of a piece of chamois with a blackout cloth backing. My first two attempts at this backing failed. I first attempted using hot glue to try and seal the rubbery side of the cloth with the chamois, but the glue was too lumpy, and seemed like it wouldn't make a nice smooth sealing surface once assembled. So I then glued the cloth side to the chamois with wood glue, and then glued the rubbery surface to the bottom of the wooden pieces with the hot glue, but the bond did not hold. Finally, I found an old tube of contact cement, and that did the trick. I clamped everything for a few hours just to be safe.

Once the chamois/cloth facing was nice and secure, I worked on adding a piece to hold the spring in place. I used a small drill bit to carve grooves in the wood for the spring wires so I'd have a nice surface fit.

 While I was letting the glue continue to dry, I worked on getting a small piece of copper pipe cut. The 3/4" pipe was a slightly loose fit in the reservoir air outlet block that I had drilled with a 3/4" forstner bit, but over the past week or so, I've been applying layer upon layer of shellac in the hole hoping for a snug fit. I cleaned the copper pipe, cut about a 3" length of it and did a diagonal cut across the bottom of the pipe so I'd have decent airflow through the elbow of the block.

The repeated coats of shellac turned out to be beneficial, because I needed to use a hammer to pound the pipe into the block. It seemed like a nice tight fit, but just to be safe, I covered the joint with another coat of shellac, and shellacked the bare copper as well.

Returning to the relief valve... I knew there would be upward force on the hinge base caused by the spring, so I figured that a nut and bolt through the reservoir box lid would be most secure. I figured screws wouldn't be secure enough, especially since they'd only be grabbing the thin 1/4" plywood of the lid. Unfortunately, I didn't have any brass nut and bolt pairs, so I used a steel bolt, and figured out a way to completely hide it from view.

Here's the relief valve all put together, crazy curly-Q springs and all.

Looks like it will soon be time for a pressure test!