Asher Guitars WD Music Products Amplified Parts Mod Kits DIY Nordstarnd Pickups Warmoth.com
Asher Guitars WD Music Products Amplified Parts Mod Kits DIY Nordstarnd Pickups Warmoth.com
Asher Guitars WD Music Products Amplified Parts Mod Kits DIY Nordstarnd Pickups Warmoth.com

Small Acoustic Archtop

Discussion in 'Tele Home Depot' started by richa, Jun 19, 2017.

  1. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    Not much carving left. Close photographic inspection reveals some areas that need tweaked. I find I use my finger tips more when I'm working on it - amazing how good they are at picking up very small irregularities. But for some reason I see things in pictures of it that I don't see when I'm working on it.

    20170702_174206.jpg
     

  2. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    Now I've got some planning to do. Might locate the f-holes. Might set the plates aside for a while and come back to them with a fresh critical eye before doing the bracing. That would give me a chance to root around some more for an approach to tuning the plates that might be slightly better than random. It seems to me that I could wait until I'm ready to glue the plates to the sides and add bracing then. This would allow me to glue the top first and record audio samples of the impulse response (i.e. - a tap) and run them through the frequency analysis tool in Audacity to help guide the process of thinning the bracing. It would be better than nothing. The back would still be a shot in the dark. Though I might be able to clamp things up well enough for a not useless sample of the body as a whole before committing to gluing both sides.

    Also need to think about the f-holes. When I did this one I had a few things going for me.
    20161222_212958.jpg

    The sides were solid built-up from cheap plywood 1/2" thick. So effectively they were not acoustically active. The top plate was not thinned in the area where the f-hole was - so I had about a 1/4" thickness to work with. All of that meant that the calculations for the Helmholtz resonance were actually pretty indicative of reality (which is not normally the case for acoustic guitars). So setting the low frequency response was somewhat better than a shot in the dark. Of course I still got the plates/bracing too stiff. The end result was that the low frequency response is quite good, but a bit disconnected from the rest of response. Sort of like what people complain about with badly designed ported speakers. :)

    Anyhow I'm thinking the only way to set the low frequency response on this build will be to make the f-holes under size and enlarge them experimentally (maybe using Audacity again). Might even be able to string it up before applying finish and just do it that way.
     
    Zepfan likes this.

  3. Barncaster

    Barncaster Poster Extraordinaire

    LOL! I'm a dash A man myself.
    IMG_3636.JPG
     
    richa likes this.

  4. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    Hah! Under normal circumstances I would agree. But I just call 'em like I see 'em. :)
     
    Barncaster likes this.

  5. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    Circumstances have not been conducive to making progress on this thing. But that worked out because it gave me a chance to come back to the plates with fresh eyes and a bit more will power to tweak.

    I jokingly referred to "photographic inspection" earlier but it was only half joking. Typically when I'm trying to take a picture of something I play around with the orientation relative to the lighting to get a better picture. I noticed in the process that things show up in the pictures that I wasn't really noticing before. So today I went back and started closing examining the front plate while holding it such that the bench light raked across the surface casting shadows. I concluded from this that while the arch is smooth long-wise, it has a somewhat abrupt "break" side-to-side. Fortunately this is excess material that can still be removed. This is probably a good thing...

    So is it possible to tell anything useful about a plate using a spectrum analyzer phone app (the app is cleverly called Spectrum Analyzer)? Got me, but I gave it a try and the results were at least suggestive. I tried two basic tests (an a number of variations that I won't go into much detail about because they didn't really add much).

    Trigger Alert: Tap Tuning discussion to follow. If you find this topic upsetting please do not proceed. :)

    First I tried a holding the front plate by the edge of the lower bout and tapping it at the bridge location with my knuckle to get an impulse response for the FFT program to chew on. This program allows you to collect an average of samples over time which is kind of handy since I could tap it a dozen or so times and average out the variations. This gives the free plate impulse response. I'm not interested in quantitative results - just qualitative. I'll say more about that shortly:

    Screenshot_20170709-152539.jpg

    Then I tried to simulate the effect of the plate being glued to the sides by clamping it to my form. This would not be perfectly representative, but it would at least be indicative. What I'm primarily curious about is what changes compared to the previous test.

    20170709_155841.jpg

    Screenshot_20170709-160143.jpg

    So what do I know now that I didn't know before?

    Starting with the limitations of the test:
    • The results cannot be viewed as quantitative in terms of absolute db values. This would required a mic calibrated to a reference source with carefully controlled distances etc. etc.
    • The relative height of the peaks probably must also be taken with a grain of salt. Again the response of the phone mic is an unknown. On top of that the program allows you to apply a correction to the display that is similar to the way the human ear responds to sound.
    • We are talking about individual plates - not the guitar as a system.
    • The bound plate test probably introduces some artifacts that are not representative of being glued to the sides.
    • I don't have data from a "known good build" to compare to. So it's a bit of blind man's bluff.
    Channeling bits and pieces of things than can be found from Alan Carruth, other sources, and speaker design:
    • Low end response is shaped by Helmholtz resonance and this only comes into play on an assembled body. So that will have to wait. But basically it's controlled by the volume of the container, the thickness of the material at the air port (f holes), and the size of the air port. Because guitar bodies flex (compared to say a ported speaker cabinet) The Helmholtz equation breaks down. But it still happens and there are some fudge factors that can get you to a ballpark calculation. Apparently you want this to be a bit above the low open E string (maybe 90-100Hz). When I did my tenor guitar this played out quite well. But the rest of the frequency response didn't quite come together with it.
    • Above that you have the main monopole resonance mode of the plate (e.g. - the whole plate moves in and out). This mostly affects the 100-200 Hz range. It's what you have the most direct control over with the thickness of the plate and the bracing.
    • Above that you have a bunch of stuff that I have no idea how to influence. Unevenness in the carve can probably negatively affect this because it makes the higher order resonance modes asymmetrical (or non-existent) and generally keeps the plate from sounding "musical". So you can try getting samples opposite each other on the axis of symmetry. They should not be grossly different. Beyond trying to get a smooth uniform carve I'm not really going to mess with this. I'll "get what I get".
    So then:
    • The blue curve is effectively background noise. The yellow curve is the average response of all the times I whacked on the plate.
    • It's fairly easy to see that the free-plate and bound-plate responses are not completely unrelated (though there are significant differences). This is good because with an archtop plate they should be sorta kinda related. With a flattop they are different animals. So I won't really say any more about the free plate resonance because it doesn't much interest me. It's easier to test as you are working though so I'll probably keep using it as a reference (especially that peak in 150 area) but I will use the bound plate resonance to get the final determination of whether to stop tweaking.
    • The bound plate resonance has a lot of energy starting around 150 Hz to 400Hz and isn't limited to a single peak. This seems good to me except...
    • If I was shaving the bracing I would be very tempted to just call it good - but I haven't even put the bracing on. Adding bracing will shift the response up. Humpf.
    • This is a small guitar so I should expect it to be a bit mid-rangy.
    • The results of tapping in locations opposite each other on the axis of symmetry seemed to indicate that the carve does not have serious issues in that regard. e.g. - it seems uniform.
    • Also - as I noted earlier I still have some excess material that needs to be removed from the plate before I start bracing. So maybe this isn't too far out of whack. I guess we shall see what we shall see.
     
    wayloncash, Zepfan, Daniel94 and 2 others like this.

  6. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    Fussed around with the top plate some more with thumb planes and scraper. Evened out the side-to-side curve and really don't want to go any thinner. Pasted one of the old FFT's over the new one. The frequency response shifted down by about three semi-tones from this exercise. Cutting the f-holes should lower it a lot. Then adding the braces will raise it back up a lot. Wouldn't it be sweet if those two things more or less balanced out because I'd like to to be too much higher than where it is now. :lol:


    Free-upper-bridge-comparison.jpg

    Seem's to me there's a chicken and egg problem here. I'd like to wait until the guitar is assembled to take the f-holes to their final size. But the bracing needs to be trimmed before that. Might be able to do final trim of the f-holes and braces based on a clamp-up. Dunno yet. Might just make the f-holes the same size as the "prototype" and hope for the best.

    But the very next thing will be to revisit the back plate and give it the same treatment the front-plate just got.
     

  7. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    F holes.

    20170714_220815.jpg
    20170714_221254.jpg
    20170714_221822.jpg

    Not optimistic this coping saw is gonna cover itself in glory. Would be better if I had a finer blade. Hmmm...even better would be a cable saw.
     
    Zepfan likes this.

  8. printer2

    printer2 Poster Extraordinaire

    May 24, 2010
    Canada
    Leave enough room for error and cut out the middle then fine tune with needle files. I would show my resonator which I put f holes in.
     
    richa likes this.

  9. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    Good call. Being very conservative. Even at that with this wood the saw is leaving some "opportunities" on the back side of the plate. But it's cleaning up nicely with files from the front. I'll be blacking the inside edges too so it should look fine. Just ordered some different blades too. Won't do me any good this go around but they'll be on hand the next time.
     

  10. printer2

    printer2 Poster Extraordinaire

    May 24, 2010
    Canada
    I stopped half sentence. I had my pictures up on photoshop but they turned nasty, I did a resonator guitar and I photo documented the build. Hopefully I can get the pictures up back soon. Yeah, the first one that I did took a while, the second one less time as I knew what the tools can do and took a bigger bit out of the hole. I so want to start a archtop, but besides the other projects I got going I have the desire for a uke. In the process of cleaning my shop so I don't trip over stuff too much. Maybe a archtop uke? I am sure I have wood for one.
     

  11. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    The words having been uttered have a certain narrative necessity to them. You *have* to do this. :)
     

  12. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    I think I could have done about as well with my teeth than the blade on that saw...

    20170715_082601.jpg

    But all's well that end's well...ish

    20170715_144942.jpg

    Need to do a bit more fussing then on to the bracing. Yea!

    Just realized something else I jugged up. I had intended to leave the area that would sit over the neck block at full thickness...or at least a lot thicker because it will have a cutout for the neck. That area is pretty weak and won't have the benefit of being glued to the lining. Might beef it up a bit just around the edge of the cutout.
     
    Crafty Fox, wayloncash and Zepfan like this.

  13. Zepfan

    Zepfan Friend of Leo's

    Nov 30, 2013
    Horn Lake, MS
    Those turned out great.
     

  14. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    Thanks Zep! I'm pretty happy with it. I like way this particular design looks even better than I thought I would. It suggests alot more than is actually there.
     

  15. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    Backing up for the big picture

    20170715_215739.jpg

    And for my daily geek out: Bottom is the first measurement, middle is after some thinning, top is after the f-holes.

    Huh. For those keeping score adding the f-holes shifted it about another two semi-tones down. Changed the distribution of energy quite a bit too. Don't know what to make of it but curious to see what adding the braces does.
    Free_downside-bridge-fholes.jpg
     
    Zepfan, Robert.II and Mat UK like this.

  16. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    Making braces today. Well it's been a giggle. The cool cats like to split the braces from the billet so the brace ends up following the grain as much as possible. Bonus points if you split it two directions (perpendicular to each other). If you do this pure angelic song will emanate from the finished guitar and rainbows will shoot out the f-holes. Ok, I can see how this would create the strongest most stable brace possible from a given piece of wood. And it sounds kinda fun to try. But a few things don't quite add up. For starters the brace is curved so it's not like you won't be exposing end grain regardless of what you do. Then there's the question about whether run-out along the side of the brace would even matter since it's not load bearing in that dimension. I can sort of see that having the grain cut through the brace at an angle would still weaken it though. Anyway I'm game to give this a go just for grins - and within the limits of the materials I have.

    BTW - remember that old cartoon (Loony Toons or something) where trees went into a factory and each tree was made into a pencil? Ahem. :eek:

    I don't have a wedge shaped billet sectioned radially from a spruce tree. I have a piece of spruce about 1" x 1 7/8" x 6' I pulled out of an off-cut bin for $1.00. 18" worth is shown here. Notice the word "AIR" written on it - hinting to the hopeful dumpster diver that this was graded as aircraft spruce. Perhaps. It's very tight grained and well quartered. But the problem with a piece that has already been dimensioned is that the grain almost certainly runs at some angle to the sides.

    20170716_141816.jpg

    So here's the split. I've planed the face on the lower piece in the photo flat.
    20170716_143123.jpg

    Ending up with this. The curved vertical lines are saw marks. The growth rings run parallel to the planed surface at the top. If I split it again perpendicular to this one it won't be tall enough. So I will be living with whatever grain run-out I get along what will be the narrow edges of the brace. [Note from future - I did take a piece of scrap left from squaring this up and split it that direction just to see how bad it would be. It was actually not bad at all so the grain runs pretty straight in that direction. By the time the edges are squared there is not enough here to get two braces so I'll use the other half for the second brace.

    20170716_143157.jpg

    So after cutting both braces and trimming with the bandsaw/plane...
    20170716_145931.jpg

    We have two braces (bottom) and a lot of scrap (top).

    Thinking "this can't be right" I googled around a bit and found this. I do this after I have made a dog's dinner out of my piece of spruce - because that's just how I roll. o_O

    Well in that link he didn't do much better so I feel somewhat satisfied that this may be par for the course (not that I'm quite happy about it :rolleyes:).
     
    Last edited: Jul 17, 2017
    Mat UK likes this.

  17. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    By the way, I am using the Robert Benedetto book "Making An Archtop Guitar" as a general guide. This isn't the guitar he lays out in that book but it still provides a starting point.

    Worked out where the braces will go. They need to intersect the the bridge "feet" so that sets one point (where the dots are). And I picked a value for the neck end of the braces of 13/16" center to center which is the same value Benedetto uses on the larger guitar in the book. It seemed more important to hit those two points and let the "wide" end fall where it falls in the absence of a better approach. I guess I could have maintained the same angle relative to the center line, but I have no reason to believe all of dynamics will scale linearly with the size of the guitar. So I just picked.
    20170717_183813.jpg

    Then transferred the curve to the brace with a drafting compass.
    20170717_184340.jpg
     
    wayloncash and Mat UK like this.

  18. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    Notice that I didn't get a perfect split on the braces. There's a growth ring that sort of pops in and out along the length. But its pretty close. Also the band of heartwood down the center of the plate curves. Not ideal and more pronounced than I thought it would be. That wood is a bit stiffer than the rest so it would have been better to have it centered. Que sera la vie. :oops:
     

  19. NotAnotherHobby

    NotAnotherHobby Tele-Afflicted

    Oct 27, 2015
    Da' Magic Mittin'
    Interesting tap analysis.

    I've done some viewing of violin making (man, I'd live to do that as a long-term hobby). After they make the top and bottom plates, they put them on something like a vibrating plate and sweep the frequency slowly. They add some small bearings (or something) in the hollow of the arch, and when the plate hits some sort of resonance, the bearings will form a pattern.

    Oh, and I want your planes.
     

  20. richa

    richa Tele-Afflicted Ad Free Member

    Apr 23, 2016
    Washington
    I like my planes :).

    Is this something like what you saw with the violins?


    If it is they're Chaldni patterns. Apparently they provide some good objective reference points in that the pattern is associated with a particular resonance mode of the plate. So it will tell you that the plate is capable of producing that mode and I suppose with experience will tell you something about what to change if it's not developing right or at something other than the target frequency for that mode. It would be fun to play with sometime. Looks like they will sometimes just use a speaker and hold the plate over the speaker. But it's primarily a low/mid frequency thing. I think I'd have to be planning to make a lot of archtops to be worth going down that road. Hmmm.... :D
     

IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.