|
What sort of finish should I put on my marimba bars? Should I use estapol or polyurethane, or should I oil them or simply leave them be?
What sort of finish should I put on my marimba bars? Should I use estapol or polyurethane, or should I oil them or simply leave them be?
I definitely would steer clear of polyurethane or estapol for the bars. I have a marimba with bars finished with polyurethane - I wish they were not. I would never do it again. You cannot change it without having to completely retune the entire set of bars, and in the most difficult possible way. A solid protective coat can look great to start with, but over time it scratches and fades, then you are left with a substandard looking bar - possibly with chips on the finish - and no easy way to fix it. A natural oiled finish is much easier to maintain, and can look just as sensational if you put the effort into sanding back the bar faces with fine enough grits. Wax should be ok too as long as it's not the thick build up a coat on the surface of the bar type. You can just leave the bar with nothing at all, but it is probably a good idea to use a little oil to protect the timber. My preference is a natural oiled finish using an oil based furniture polish like marveer or Scandanavian teak oil.
What sort of timbers can I use to build marimba frames and bars?
What sort of timbers can I use to build marimba frames and bars?
Well for the frame, it doesn't really matter, but you are looking for stability, durability, ease of working, and possible a good price as well. Ash is a popular one but for me as long as its not pine I tend to go for the durable but cheaper options. Good seasoned oak makes great main planks for the top parts of the end pieces which get a few knocks - if you can get/afford it.
As far as the bars are concerned - To be clear - you can get a sound from just about any timber at all - it's just the quality and potential range of pitch that varies. Harder timbers tend to work better and withstand being hit more. Not ALL hardwoods sound any good though. For the "make a marimba" prototypes I have used "Durian" which is a medium hardwood available in most hardware stores around Australia - cheap. For a commercial quality instrument Honduras rosewood and more recently African padoak are the traditionally accepted timbers. Others can work ok - I have heard Brazilian rosewood bars, and they seem pretty good but I'm not sure what genus that is - there are often a few variations between popular names and their official genus. Timbers are often misrepresented. For a full range marimba I would always go for Honduras rosewood (dalbergia stevensoni) given the chance. It sounds good and works well across the whole range. African Padoak is ok. It sounds as good overall. In my opinion it is slightly better in the lower ranges but perhaps not quite as good up high. It is especially good for bass instruments - they are low so it sounds good and also because it is cheaper and easier to get hold of big planks. In the top octave + half of a concert instrument though the timber can occasionally sound say 5% worse - not a lot. The main issue is that the padoak is significantly lighter than the rosewood which means that the smaller bars can jump a bit when you play them hard and even displace the string.
What size holes should I drill in the bars, and can I use wire instead of string to suspend them?
What size holes should I drill in the bars, and can I use wire instead of string to suspend them?
The size of the nodal holes will make very little difference to the bar's sound. The only possible problems are these. ONE - If the hole is too small for the string going through it, it will make a tight fit, which can dampen the sound a little and make threading the bars difficult. TWO - If the hole is way too big for the string. This causes a problem with "bouncy bars" with too much free play. It also makes the bars sit lower to the frame, and it may be low enough to make them actually clunk against the timber struts when they are struck. In general because the eyelets used in most of the "makeamarimba" designs are not quite as tall as ideal, I like to do everything I can to prevent this, including drilling the holes only a tiny bit bigger than the string, and also drilling them closer towards the bottom face of the bars. By the way - the eyelet thing is not a design fault, they just dont seem to make them with a longer screw shaft without also making the eye diameter too big. With the holes drilled well and the tube used as described in the building guides though, you should not have any problems.
I most certainly would steer clear of using wire of any sort. ONE - it will actually fatigue faster than you might imagine. TWO it has no flex or give at all, so the slight variance of the hole to the exact nodes which is caused by material non uniformity and angled holes etc becomes more of a problem. With string the small vibration which occurs 1-4mm away from the nodes is just taken up by the flex and movement. THREE - the hard surface of wire - even insulated wire - will cause an undesirable rattle. The inside of the hole can vibrate and clunk against the hard wire whereas a softer and more flexible surface will not create a sound. The best string to use is a 3-4mm braided nylon cord or thin cotton "sash cord".
I have heard that the nodes on the bars should be 2/9ths from the ends, so I made mine that way - does this mean your bar supports are in the wrong spot?
I have heard that the nodes on the bars should be 2/9ths from the ends, so I made mine that way - does this mean your bar supports are in the wrong spot?
This issue comes up on occasion and I find it concerning. I have had a few people talk about the nodal points being 2/9ths and so forth and it worries me as I'm not really sure how that conclusion is arrived at. I'm aware that the Madin style marimbas suggest using this method, but it really is not a very good one, and if you check the Madin books you will find that his own measurement tables for bar nodal points don't actually always use the 2/9ths ratio at all!! Whether you use 1/4 or 2/9ths is immaterial as it really is a very rough guide only - just to be used as a quick guess when actually tuning the bars. The idea of having a uniform physical distance for the nodal points at 2:9ths of the bar's length is not strictly correct. It is a good starting point, and in the average instrument's middle to high range it is probably pretty accurate. This is because bars in this range have just a little shaping of the arch. Most professional instruments however have the true nodal points closer to the end of the bars as the bars get lower in pitch. This is to do with the shape of the bars vs their length. Because an unshaped bar would need to be an unpractical length to achieve low notes, we simply cut them shorter, and remove more timber from the arch. The more timber removed from the arch, the lower the pitch of course, and as an extra effect, the node tends to move slightly towards the bar end. Mr. Madin's designs are simpler and are for DIATONIC instruments, so using longer bars with less cutting of the arch makes sense. It is not practical for a chromatic instrument which has two rows of bars though, because longer bars would make reaching the second row difficult.
The following is to help you understand how it works, if you do not already: A bar of uniform cross-section - ie. No arch cut at all - will have its nodes at 1/4 the bar length. The bar ends are antinodes and the bar center is an antinode. The wave travels at uniform velocity along the length as the thickness is uniform. As the 1st mode of the bar is one complete cycle along the length, the nodes are exactly halfway between the antinodes - ie the center and the end - ie 1/4 the length. As we start lowering the pitch of the bar by making it thinner in the middle we are effectively slowing the waveform down in that part of the bar. This has the same effect as making the middle part of the bar longer so what WAS a bit like this BEFORE shaping the arch: Bar middle = M________NODE_______A = antinode.Now is acoustically like this: Bar middle = M____________________NODE_______A = antinode.Now you will see that in terms of a ratio the middle part of the bar is much longer than the part outside the node. Of course the actual physical bar length has not changed by shaping the arch - just the acoustic length - so because the ratio needs to be the same, the node physically shifts towards the outside edge of the bar so a bar with a significant arch will physically be something more like this: Bar middle = M____________NODE___A = antinode.Same total length as version 1 above but with the node further out. NOW… That means if you have cut your bars to the correct length and thickness and tuned them to the correct pitch, the nodes WILL NOT be at 1/4 OR 2/9 (except maybe the top few which will be close) There can be some slight variations in the exact physical ratio in using different timbers for the bars, but the differences are surprisingly small. The low C of the P3 should have its node very close to 80mm from the far end (player's perspective), and 80mm from the end closer to the player to the closest hole - the further hole will be more because of the angle. The P3 was built from scratch by me and the bar measurements were taken exactly from an actual commercial instrument which has African Paduk bars. After tuning the P3 bars the nodes were found using the "salt" test - where a little salt or fine sawdust is sprinkled on the nodal area. The bar is then struck whilst not at all suspended at one end - you have to grip the fingers at one end and leave the other free. The particles will accumulate in a line over the natural node. Not surprisingly the nodes found using this method were all very very close to those of the commercial instrument the bar measurements were taken from. There is always a little variation (1-4mm say) due to the fact that timber is inconsistent by nature. When designing an instrument from scratch with non standard bar sizes, you would tune the bars first, then find the frame measurements by lining up the bars with their correct spacing and positioning all the bars so their nodal lines follow the best average you can make. This is pretty much what I did with the P3 although the measurements turned out almost identical to the commercial instrument. That is to be expected as the nodal points are (as explained) pretty much primarily a function of both length and pitch. You should find that the theoretical 2:9 point is much closer to the actual measured position on the top notes where there is much less arch shaping. As far as I'm aware there is no practical way you can shape a bar of specific pitch to place the nodes where you want them without changing the bar's length. If you are at all unsure about where the actual nodes exist within your bar, you should do the salt test to find out. You could also try just gripping the lowest bar (where the effect is the most noticeable) firmly between thumb and finger at the MEASURED 2/9 spot and striking it. If the node is not actually at that spot (it shouldn't be) you should feel the bar vibrate between your fingers, and the vibration of the bar will be dampened. Now for comparison try gripping the bar in the same way, but a little further out. (say 80mm from the bar end for the low C on the P3) The sound should be less dampened and you should feel less vibration in your fingertips. If you discover your nodes are more than say 5mm from the points my frame dimensions allow for I will be very surprised indeed. If that IS the case then I really cannot think WHAT could possibly cause THAT situation. As it is, there is a considerable amount of adjustment room built into my frame designs anyway. ONE - the A frames are a little longer than actually required, so the rows of bars can be slid a little towards the smaller or bigger end. TWO the timber struts are 19mm thick which gives a fair amount of room to simply screw the eyelets in a little more towards one side or the other.
|
