I apparently never posted this to the list, intending to write it up and submit it to the Journal. Since I still haven't yet done that, and someone asked, here's the gist of it. After the demise of Terry's differential expansion gage, I got to thinking - and you know what kind of trouble that can lead to. As usual, what I started out doing was different from what I ended up doing, which was different from what I should have done from the beginning. After flailing around doing it inside out and backward for about a week, here's what I decided I should have done in the first place, and how I would do it if I were building another one... I already had a $6 linear dial indicator from Harbor Freight that I had purchased last year for the inevitable dedicated tool I figured I'd be building before long, or just for general purpose confusement. It was just too cheap to pass up, and this looked like a fine use for it. I made a maple frame of "U" channel for a slip fit of the panel scrap I intended to use. So far so good. Choose a well quartered 8mm plank, somewhat dense, with relatively tight grain that's uniform all the way across. If there's only a 50mm width that's uniform, rip it to 50mm. It should react slower than a light loose grain piece, so it won't get too far ahead of the wood in the soundboard panel when it's actually used as an MC gage. It shouldn't move as far as the loose grain stuff for a given MC shift either, and your gage can be longer for a given expansion scale to, hopefully, increase measurement accuracy. Now, what's the actual expansion/contraction rate and how to scale it to the dial indicator? Cross cut the chosen plank into strips of the right width (mine's 50mm) to fit in the channels in the frame pieces, leaving a couple millimeters free top and bottom so the strip won't bind. It's just supposed to slide easily, but not sloppy, in the tracks. Cut enough strips to fill the frame length. Stabilize them at some measured MC in the hot box, house, or shop; wherever the temperature and RH is stable enough for a couple of days until they quit moving. Somewhere around 10%MC would be good. Record the cross grain width of one of the pieces as an indicator, and mark it as the test piece. Or do them all for comparison averaging. Record the MC from temperature and RH%. Dry the pieces in the hot box for a couple of days until the test piece quits changing dimension (and/or weight) at around 6%MC or so. Record temp and RH% with the same instruments you used for the high MC measurement, and figure the new MC. You now have a dimension @ a specific MC for two MC values that aren't near the less trustworthy high and low limits of your RH% measuring device. I assume the error in an electronic hygrometer will likely be in the same direction and close to the same rate for both measurements, so the proportion of dimensional change to computed MC% change should be valid whatever the numbers actually are. This may not be the case, but it seems likely. If you have the facilities to do this with a sling psychrometer, that would probably be better. Now the scaling. From the expansion figures I got, I figured I had room in my frame for a strip of the proper length to change 0.010" per 1%MC. That seemed sensitive enough to me to be useful, and simple enough to figure out on the gage without a conversion table. I can read it directly off the dial. From measurements taken on the test piece, find the expansion rate per inch per 1%MC increase, and glue enough pieces in line to nearly fill the frame, and mark the proper length to have a piece that will expand 0.010" per 1%MC. Put a thin maple cap on one end of the strip for the dial plunger to ride on without denting, and mount it in the frame with the gage so the gage reads what you've just measured the MC at. I mounted mine by driving a small wedge with a bit of glue between the frame edge and the panel at the mark I had made indicating the appropriate length. That way, the waste end can expand and contract without getting in the way of anything, and My frame can be a bit longer than actually needed in case I want to replace the panel piece some day. Now, start testing it against as accurate a temperature and RH% means as you have available at different humidity levels to certify it's accuracy. The sucker seems to be pretty close, and since it's unconstrained, it won't suffer compression set either over time, or from extremes of humidity like most of the designs I've seen will. As long as I don't overheat it or contaminate it with something that affects it's moisture capacity, it should last a very long time. So far, I like it. Here's the photo. Photo: https://www.moypiano.com/ptg/files/attachments/28/96/b7/66/Dsc00001.jpg Alternate URL: http://tinyurl.com/u4ah
This PTG archive page provided courtesy of Moy Piano Service, LLC