Stringed Instrument Tuner, Ideas?

[dtief]

@ philba - It should be equal. When the string looks like it's not moving coresponds to when you hear the beat frequency drop to zero.

- to clarify my previous post:

Here is an example of the tachometer / strobe I was refering to:
http://www.hotektech.com/Itlab1542B.htm

Not to be confused with an automotive timing light, which is sychronized to the RPM of the crankshaft.


@ ecerfoglio - I agree, a shorter pulse will improve the "sharpness".

A Luxeon LED would work great for this. Brightness wouldn't be a problem. The K2 can easily be run at 1Amp continuously, pulse is even higher.

Dave

[philba]

Yes, it should be equal until you consider that you shouldn't have to actually look at the string. I tend to look at the key/peg while I turn it. Some stringed instruments are easier to look at than others. A guitar is reasonably oblique if you are holding it like you play it. My son's cello is moderately easy. Violins are easy that way.

Still, I think setting up a strobe seems more effort that it's worth. So much easier to play a note on my piano (or turn on the tuner), hit the string and zero the beat. Fast and easy. Plus, when you zero in the beat, you don't have to wait - it's really obvious. There is a resonance that I can actually feel - especially using the piano.

[MrAl]

Hi again,

Dave:
Yes a high power LED would be nice. I think i can get away with
a small 20ma Nichia however, driving it with a high current pulse
around 60ma or so.

philba:
This is just another way of doing it. The end result would be a tiny
box you place next to or under the string and look at the string
as you turn the knob and pluck the string, guitar in lap.

[Bob Scott]

I found it interesting to watch guitar output on an oscilloscope. Harmonics crawl along the waveform, meaning that they are not phase locked or phase coherent with the fundamental.... Interesting!

[MrAl]

Hi Bob,

Havent done that in years now, maybe i'll take another look soon.

[MrAl]

Hi again,

I almost forgot to mention...

Another benefit from building your own products rather than buying
one already made at a store is that you can build it to your own
specifications, of which you may not even be able to find on the market.

A good example of this was a NiCd charger used for a drill pack.
The one that came with the drill was really crap, really, and i didnt
have a schematic so the only way to make it work was to either
buy one or design a new one and build it. I dont like the chargers that
are out there to purchase, so i was really stuck this time with either
designing and building a new one or junking the drill altogether.
The drill wasnt that old yet, so i decided to make a new design.
The new design would charge the cells when they needed it (after
a typical use of the drill) and also be able to replenish the self discharge
on a daily basis. So far it's been working very well and it's been like
2 years now. I hardly use the drill, but when i need it it's charged now.
Yes it took time and effort to build it, but using a micro controller it
wasnt too complex even though it has several charge modes and
means to adjust the daily self discharge replenish amount.
LATER
Just took a look on the web and can only find battery pack chargers
around 50 bucks or so, and im not even sure if they would work or
not for what i need. Total parts cost for my own design was around
5 dollars including the wall wart because i was able to use the old
wall wart that came with the drill for the new charger too.
I see lots of drill packs but not that many chargers on the web.

[CeaSaR]

Ok, I know this is an old thread, but I cannot recall who asked for the info I found.

At AMZ's site (Muzique.com) there is a note to frequency converter. Put the note
and octave in and you'll get the frequency for that note.

http://www.muzique.com/schem/freq.htm.

Hope this is of help to the person who needed it.

CeaSaR

[Dean Huster]

That was odd seeing this thread pop up a year later. But I guess I hadn't seen the "later" posts regarding the use of a frequency counter to tune a stringed instrument.

I tried that once, using a contact microphone and counter to tune a piano. A piano strikes three strings simultaneously on the upper octaves, two strings on a lower section and only one on the lowest notes. Because of that, you end up with a really NASTY waveform to work with, so just like a "real" piano tuner, you have to kill off the two extra strings and play with only one at a time. Still, you're stuck with two problems: a complex waveform that gives the counter multiple trigger points and a decaying amplitude. Restriking the note only makes things worse for the counter. It's easier to work with a non-hammer-strike stringed instrument such as a guitar, mandolin, harpsichord. You still have a decaying, complex waveform, but at least you only have one.

That's why a good piano tuner armed with a tuning wrench, rubber stops, a single tuning fork and his ear can tune a piano in an hour or so. Even under near-perfect conditions, a counter would take a lot longer and the desire would be to tune each note to eight digits of precision, even though a four degree change of temperature would skew each note and blow eight digits to smithereens!

[jimmy101]

Viewing the vibrating string with a strobing LED is certainly good for teaching people how a string instrument works. And how to visualize something that is moving periodically but too fast for the unaided eye to see.

I wonder, what is the on / off time for a typical LED? Seems to me that that is going to have a pretty large affect on what you want the duty cycle to be. Xenon flash tube strobes have very fast rise and fall times and can easily get total flash widths of 0.1mSec or less.

One thing to consider about accuracy of tuning any musical instrument is that the ear, even someone with perfect tone, is only sensitive to about 5 cents or so. Expert piano tuners can only tune a piano to about that accuracy (or perhaps a bit better). Tuning to better than 5 cents accuracy might be a fun intelectual indeavor but it won't make any difference in how the instrument sounds.

A "cent" is a 1% difference in frequency between whole notes. On a piano "middle A" (A4) is 440 Hz. A-sharp (A#) is 466.164 Hz. 5 cents is (0.05)(466.164 - 400) = 1.3 Hz. Since cents are percentages the frequence change depends on the note. So, in the middle of the piano scale an accuracy of +/- 1 Hz or so is all you need. Tuning the instrument any more preciely than that is a waste of time since even a trained ear can't hear the difference.

[CeaSaR]

I know what you mean Dean, but I remember a thread where someone asked about the frequencies of
guitar strings as they were trying to develop a tuner. I only found 3 threads: this one you are reading now,
(>FLUKE 8060A Guitar Tuning Interface<) and (>guitar tuner VU Meter failed<).

I think the thread I was looking for disappeared during that spate of spam that innundated the forum a
while back. Hopefully the person who was looking for the info will see this and be able to use it.

CeaSaR

[MrAl]

Hi there,


Wow this was a thread from quite a while back
Thanks for the links still.

[Dean Huster]

The top octave of a piano. Divide each frequency by two to get the note an octave lower.

A 1790.000 Hertz
A# 1864.654
B 1975.532
C 2093.004
C# 2217.460
D 2349.318
D# 2489.014
E 2637.020
F 2793.824
F# 2959.952
G 3135.964
G# 3322.436

There is an EXACT mathematical equation for moving from one half-step to the next, and based upon the standard A440, you can get the entire scale down to the lowest 32' pipe organ C all to way into the ultrasonics flawlessly. However, I've misplaced it. Lots of good that does!

[Bob Scott]

There is an EXACT mathematical equation for moving from one half-step to the next, and based upon the standard A440, you can get the entire scale down to the lowest 32' pipe organ C all to way into the ultrasonics flawlessly. However, I've misplaced it. Lots of good that does!

Well, there are 12 notes in an octave. If you have a piano, that includes 7 white keys and 5 black ones. They are all equally geometrically spaced. The black keys are not odd in any way. I can't understand why they are considered sharp or flat.

Every note in an octave is double the frequency of the previous octave's note. The "C" above middle C is double the frequency of middle C. So, the formula for any next note is:

Previous note * 2^(1/12).

The 12th root of 2, where 12 is the number of notes necessary to reach double (2X) the frequency. It is a geometric progression.
I can't think of an easy way to make a top octave generator out of this equation. AFAIK, all top octave generator ICs use methods that are only very close approximations.

You can probably use the same TYPE of formula to calculate all the sharps and flats by changing the 12 to a higher number like 24 to get notes AND half steps. I assume that they are half steps if A sharp is the same frequency as B flat. Anybody know?

Musicians can't do math. The word "Octave" connotates 8 notes in an octave. Do, Ray, me, Fa, So, La, Te, Do, right? 8 notes? No, we counted "Do" twice. The last Do is actually the start of the next octave, so there are only 7 notes in an octave if you are a musician.

We amateur mathematicians know that there are really 12.
I expect some feedback because the formula does not agree with some of the note frequencies posted by Dean..

Bob

[MrAl]

Deans top octave scale:
A 1790.000 Hertz
A# 1864.654
B 1975.532
C 2093.004
C# 2217.460
D 2349.318
D# 2489.014
E 2637.020
F 2793.824
F# 2959.952
G 3135.964
G# 3322.436



Hello again,


Well we seem to have migrated to the subject of notes and frequencies, which
is also interesting.

It looks rather obvious that Dean meant to type
"1760.000 Hz" instead of "1790.000 Hz"
that's all.

It's true that the next note in the chromatic scale is the last times the
twelfth root of two (2^(1/12) as already noted) but that is only for the
equal tempered chromatic scale. There are other scales that would come
about if we started with some normal 7 note western scales and tuned our
chromatic scale (inadvertently) with those scales instead such as what
Bach and others would have done. In this way, the notes dont come out to
have that same relationship of the twelfth root of 2 but are slightly different.
Supposedly (i never looked into exactly why this happens) when the 7 note
scales do ray me fa so la te do are tuned so that they sound just right and
exactly in harmony with at least one other 7 note scale the relationship
between adjacent notes is not the twelfth root of two anymore.
Computers however, often use the twelfth root of two because it's easy to
calculate all the frequencies of all the notes on the traditional chromatic
scale using that mathematical formula:
NextNoteFrequency=PreviousNoteFrequency*2^(1/12)
and yeah, A=110 or whatever will get you started and you can also divide by
the twelfth root of 2 to get notes lower.

The reason some of the keys are black on the piano that represent flats and
sharps is because of the diatonic scale. The piano was built with the
diatonic scale of C in mind and with this scale notes that are flat or sharp
do not belong to the actual scale itself and so are made black and smaller.
This was probably because of the common interest in the diatonic scale at the
time because there are numerous other ways to organize the entire chromatic
scale. I think the diatonic scales came from the early Greek era, and was
diatonic, and early church music was diatonic. It's interesting though that
once we switch to some another musical keys the black keys dont make as much
sense anymore. Maybe the first musicians mostly played in the key of C and maybe
D minor? Probably the first instruments had a limited number of notes that could
be played on them. When other instruments were designed they would have
had added notes which would have sounded very strange to many people except
the musicians themselves. This may have shaped the way the piano was made.

As music attempts to mimic life, there will no doubt be other scales out there
too that do not adhere to the rule of the twelfth root of 2. Even in contemporary
music there are slides that encompass an infinite number of frequencies that
dont belong to the main scale involved in the main music theme.

[philba]

Jimmy, that's why using the beat frequency is good. The number of beats per second is the frequency difference in CPS between the string and the reference. You can get it within fractional CPS.