Had some lessons learned of course, but next time with a few changes, could hit the start button on the CNC after setup at 4 PM, and come in the next day at 8 AM, and be 100 % complete, unattended. This is just the prototype. Imagine going to production with a full design, unattended machines (other industries call these robots). As I have said many times before, these machines are great at making critical parts, but we need a Luthier to make this work. Imagine, as John Lennon liked to say.
Steve Wishnevsky can make odd guitar things - but typically does a solid body guitar in a few days.
Not having any skill I would take a Kremona $69 piezo pickup, and have a little battery operated Karlson coujpler from 12mm Baltic birch, neodymium coax speaker, a TPA3116 amp, 20v tool battery and $10 tool battery dock. I think the sound of the Karlson front chamber would be harmonious with a nylon string guitar. My hands are numb in old age so all a dream - but was never a good player anyhow so little loss
Not having any skill I would take a Kremona $69 piezo pickup, and have a little battery operated Karlson coujpler from 12mm Baltic birch, neodymium coax speaker, a TPA3116 amp, 20v tool battery and $10 tool battery dock. I think the sound of the Karlson front chamber would be harmonious with a nylon string guitar. My hands are numb in old age so all a dream - but was never a good player anyhow so little loss
I briefly experimented with DIY optical pickups about 15 years ago before I discovered the Graphtech saddle piezos. I tried the usual optical interrupters, the plastic things with a slot in them. A few could be found with the base of the pickup transistor pinned out so that some bias could be applied increasing the sensitivity in interrupter use. Bias is needed to make the transistor linear enough to recover audio from a vibrating string. I also tried my own DIY pickups, both infrared and visible light.
The biggest obstacle here, and one that I could never overcome is optical 60 Hz pickup. Incandescent lights were bad, but a florescent bulb could be detected 20 feet or more away if any of the light got into the pickup. I can't imagine what todays LED bulbs would do with their tens of KHz modulated emissions.
There are some maker boards on the market that do "sonar rangefinding" Some of these will output a linear signal which will contain audio if pointed at a guitar string. No further experiments were tried.
The simplest single string pickup I have found is one of the tiny magnetic speakers used in PC's for the "beep." Their frequency response is usually only a KHz or two wide, but if you rip the thin steel diaphragm off you are left with a coil of wire wrapped around a magnet. It makes a fair pickup for a steel stringed instrument.
I am trying to stay away from pickups, to get the natural sound of the guitar, but nothing will be ruled out just now. I am curious to see how much natural sound will be provided from the horns, since they are solid maple and part on the guitar body itself. If they provide good tone, then maybe a pickup and transducers could be used, that would be much easier than using a mic , and also greatly reduce feedback issues.
Joe,
I think you are missing the harmonics, which go past 16kHz.
Here are a few screen captures of all six strings of an Ovation Elite, with roughly 20 year old strings strummed open, just captured using the internal mic on a Mac computer, using both RTA and Spectrograph (SMAART software).
The dB scale is not calibrated (it is reading higher than the actual SPL) but the relative response is informative.
As you can see in the spectrograph, the low range takes time to build up, and decays more slowly than the upper harmonics, and the second harmonics of the lower notes are louder than the fundamental frequencies.
The upper harmonic response does not drop steeply until past 12kHz.
With new strings, I'd expect as much as +6dB in the upper harmonics.
Anyway, without comparing your horn to the speakers in a box, you will have no idea how much gain the horn is adding or subtracting.
Art
Attachments
Ok got it Art, the harmonics, thanks much. You are turning into my Audio mentor. As I get to these high frequencies it is not as peaky, but the dB goes down quickly, to the point where there is not any volume gain above 10kHz. I did a crude experiment with some wood that was about 2 feet long and approx 3" x 3". I also included a crude compression chamber. The small speaker by itself of course had very weak bass, but when I mounted it to the wood, it sound like the bass player had just joined the band, very obvious bass improvement. My horn is five feet long and captures full 1/4 wavelength of low frequencies, so it should be even better, I hope. Is this logical?
Check out Eva Cassidy doing "Over The Rainbow". Sadly she passed away years ago. Her voice is pure beauty. I thought nobody could do this better than Judy Garland, but maybe not true. She sings along with only an acoustic guitar, stunning. Then the strings join in, does not help. I love violins, but only when necessary. My point is a great acoustic guitar sound is hard to beat, with vocals. The only other instrument that can match or exceed this is of course a piano. How many other instruments can do this by themselves? Sometimes music is just over processed.
Joe,
1)Without an SPL and drive voltage reference, we don't know don't know the "gain" is over the driver in a box.
2) Small speakers are near omni-directional, the front and back waves are opposite polarity, cancelling them out. At high frequencies, you would hear a narrow "beam" front and back, a "figure of 8" pattern.
It would have been easy to build a horn from foam core (or cardboard or thin plywood) and duct tape (Gorilla Tape is my favorite..) to test your design, though it looks from the pretty routing you finished that you are close to test mode.
What are you using for measuring response of your speaker/horn tests?
3)Hope is "logical" if one has already simulated, or better yet, built an operating device proving an advantage.
With a weak motor (low Bl) driver, no phase plug, relatively low throat to mouth ratio with a very small mouth relative to Fc, my expectations would not be high for a horn/pipe as you have designed to achieve a favorable outcome.
By favorable outcome, I mean +3 to +6dB voltage sensitivity gain over the range of 80Hz to 16kHz compared to a speaker mounted in a bass reflex enclosure of similar volume as the horn occupies, with reasonably smooth (+/- 3dB) frequency response.
Bjorn Kolbrek wrote a good introduction to horn theory:
https://www.grc.com/acoustics/an-introduction-to-horn-theory.pdf
Hope your horn does better than I think it will!
Art
Last edited:
Hello Art - I like that article, I have seen parts of it quoted before, but not the entire article, need to study further. Here is more ammo for you, let me have it! A good way to learn. I did build a proto from PVC and cardboard, five feet long, and had a 90 degree turn, it was much louder than the speaker. It was mostly pipe, with a horn on the end, very crude, I do not trust it. Not being wood, the tone will be much different. Plus not exponential at all. I tried to build a proto of the actual shape using cardboard and duct tape, too complex, kept collapsing.
I decided to cut the real wood horn, with a 1:8 mouth to bell ratio, got this from Klipsch. The mouth is small, but this is coming off a larger compression chamber tuned at 89 Hz. It is truly exponential, given the precision CNC equipment used, as shown in the video.
It is similar in size to a saxophone, but of course does not have a reed. I will test it when complete. All I have to test is a dB meter and my ears. I will know very quickly if it works, or is a failure. I may need to get bigger, better speakers. I have some room. My mic and amp may also need to be upgraded. The patent legal claims refer to a general electronics package, so I can change whatever I want there, I just need to get the horns to work! I am out of town, so can't test right away. Thanks!
I decided to cut the real wood horn, with a 1:8 mouth to bell ratio, got this from Klipsch. The mouth is small, but this is coming off a larger compression chamber tuned at 89 Hz. It is truly exponential, given the precision CNC equipment used, as shown in the video.
It is similar in size to a saxophone, but of course does not have a reed. I will test it when complete. All I have to test is a dB meter and my ears. I will know very quickly if it works, or is a failure. I may need to get bigger, better speakers. I have some room. My mic and amp may also need to be upgraded. The patent legal claims refer to a general electronics package, so I can change whatever I want there, I just need to get the horns to work! I am out of town, so can't test right away. Thanks!
Joe,
1) If you mean the speaker with no box, completely understood.
If you mean the speaker in a box with the enclosed volume of the horn, the horn may be louder at some frequencies, less at others, you have not determined which.
2) You are making a hybrid, like an electric guitar/amp, so "tone" is important, but loudspeakers designed for reinforcing microphone signals accurately should not have "wood tone", they should be as flat response possible, "what goes in, is what comes out, only louder".
3)The actual shape is not of particular importance for initial prototyping, but the expansion rate is. It is fairly easy to tape together four pieces of cardboard that have the same cross-sectional area per given distance from the throat as your design calls for, the shape is of less importance than the area.
4) The "bell" of a horn is the mouth, the other end is the "throat". If you meant throat to mouth ratio, what Klipsch horn uses that ratio?
5)Not sure how you determined "compression chamber tuning", you might be thinking of reactance annulling. Leach's paper linked in #2 of this post is useful:
tune the rear chamber for reactance annulling
The throat to mouth ratios of musical horns is many times higher than 1/8, but the mouth is too small to support the low frequency note(s). What you hear (or measure) are primarily upper harmonics. The spacing of the harmonics are multiples of the fundamental, giving the impression of the note, even though it may be -20 (or less) dB SPL than the fundamental.
6) You can do testing with a dB meter, tone generator, and a volt/ohm/meter, but accurate results take a lot of time, which is a limited resource. The dB meter has little utility for measuring the output of a musical instrument or reinforcement from a loudspeaker.
The screen shots a few posts back showing a guitar's response took literally a few seconds, and gave a permanent record of information, while listening is subjective, and auditory memory often fleeting.
You could get similar information about your speakers (and guitar) using REW (Room Equalizer Wizard) free software, which includes tools for generating audio test signals, measuring SPL, impedance, and phase response plus many other features.
REW - Room EQ Wizard Room Acoustics Software
Even if it doesn't turn out as planned, wire some sort of equalizer into your signal path to see if it can be made to work.
The EQ does not need to be mounted inside the guitar for a simple test. I used an outboard rack mounted graphic EQ because I had one.
One of those fancy DSP boxes would probably offer more control. You will eventually run into feedback issues or driver Xmax issues in your quest for SPL, but EQ can help a lot.
The EQ does not need to be mounted inside the guitar for a simple test. I used an outboard rack mounted graphic EQ because I had one.
One of those fancy DSP boxes would probably offer more control. You will eventually run into feedback issues or driver Xmax issues in your quest for SPL, but EQ can help a lot.
Oh hell, yes of course I meant throat to bell, sorry Art! I do not have my specs with me, but I have a 1:8 ratio from Klipsch, but one of their small speakers. The large speakers are not far off from that, of course the throat and bell are much bigger. Look at a tenor sax, you can see the ratio is similar. OK maybe 1:10 or 1:12 which I could do, but not 1:40 like a damn tuba! Also look at an Alphorns. They are VERY long, but the throat to bell does not look like a great ratio, and they send sounds to great distances. Ok I know, they have mountains!
Yes I do mean a horn compared to an open speaker, so much better. Remember, I only have the guitar body to work with. If I can't make this space work, I am toast.
I hear what you are saying about mics, but I have heard it myself. Plastic sounds like **** and wood sounds so much better with the same source, why?
I agree shape is meaningless, thankful for that, because I need a crazy shape to fit the guitar.
Compression chamber tuning is based on sizes of a resonant chamber...like a reed, of course you know all this...again, the sound from the guitar soundboard remains, this is in addition to that...Thanks! Joe
Yes I do mean a horn compared to an open speaker, so much better. Remember, I only have the guitar body to work with. If I can't make this space work, I am toast.
I hear what you are saying about mics, but I have heard it myself. Plastic sounds like **** and wood sounds so much better with the same source, why?
I agree shape is meaningless, thankful for that, because I need a crazy shape to fit the guitar.
Compression chamber tuning is based on sizes of a resonant chamber...like a reed, of course you know all this...again, the sound from the guitar soundboard remains, this is in addition to that...Thanks! Joe
Hi Joe K
here's a couple of interesting articles
Acoustic input impedance | Inside the Saxophone
http://www.xlrtechs.com/dbkeele.com/PDF/Keele (1973-09 AES Preprint) - Optimum Horn Mouth Size.pdf
here's a couple of interesting articles
Acoustic input impedance | Inside the Saxophone
http://www.xlrtechs.com/dbkeele.com/PDF/Keele (1973-09 AES Preprint) - Optimum Horn Mouth Size.pdf
The DB Keele paper copies Olson who is God. (Keele extended to more cases than Olson thought essential; Keele had a computer.)
The length of the horn is not a primary parameter for a wide-range horn! The length comes out of the required/optimum mouth area, the desired throat area, and the flare rate. I have made VERY effective 80Hz horns in 13 inches of length; using a huge cone and usually arrayed mouths. Yes, hand-size mouths on car door speakers may work out near 45 inches length, but that is incidental; and the hand-size mouth is gonna reflect most frequencies but aid a few. See the 30:1 dips in the sax paper.
The length of the horn is not a primary parameter for a wide-range horn! The length comes out of the required/optimum mouth area, the desired throat area, and the flare rate. I have made VERY effective 80Hz horns in 13 inches of length; using a huge cone and usually arrayed mouths. Yes, hand-size mouths on car door speakers may work out near 45 inches length, but that is incidental; and the hand-size mouth is gonna reflect most frequencies but aid a few. See the 30:1 dips in the sax paper.