Re: [Harp-L] Tolerances on Harmonicas and the Manufacturing process.

["Vern Smith" <jevern@xxxxxxx>]

I agree to all that Harvey has said about tolerances and 
manufacturing processes.  I also agree that small reed 
tolerances are better than large ones.

However, I raise the following questions about the 
importance of variations of the clearance between the reed 
and slot:

 If the nominal clearance is .001". Reeds vary in length 
from .75" to .30" for an average of about .50".  The width 
of a reed is about .08" Thus the clearance area around the 
average reed is (.5 + .5 + .08)" * .001"  = .00108 sq in.

However, consider the gap of the opening reed in the same 
chamber.  A typical gap for a .5" long reed is .008"  The 
area through the gap (two triangles of .5 * .008" plus .08" 
* .008" at the tip) is .00464 sq in.  ...almost five times 
the clearance area.  If by means of tolerances or embossing 
you close or open the clearance by .0005" or 50%, you have 
changed the clearance area by  .00054 in sq.  However, you 
have changed the total leakage area by only  .00054 / 
(.00108 + .00464)  =  9.4%

Not only that but the opening reed (especially the long 
ones) open even farther when you blow or draw.

Thus the gap of the opening reed contributes the vast 
majority of the leakage area in an unvalved diatonic.  I 
think that this remains true even if you quibble with my 
exact numbers.

Q. What then is the big deal with clearance tolerances and 
embossing?  If they are significant, it must not be as 
simple as total leakage.  It must be how completely the flow 
is shut off when the reed is passing through the slot.

Q. Could it be that we subjectively perceive differences in 
performance that are not there?

Q. Has anyone measured differences in clearance and then 
compared performance in a test using "blind" players?

Q. What have I overlooked?

Vern

----- Original Message ----- 
From: "Harvey Berman" <cscharp@xxxxxxxxxxxxx>
To: <harp-l@xxxxxxxxxx>
Sent: Tuesday, May 26, 2009 8:26 AM
Subject: [Harp-L] Tolerances on Harmonicas and the 
Manufacturing process.


It seems that all the conversations about Harmonica out of 
the box quality and consistency have to do with tolerances.

Tolerances mean that when you specify a product, or you 
manufacture a product you specify the allowable deviation 
from the standard you wish to achieve. This is a plus or 
minus situation, and the tighter the tolerances, the more 
expensive the product. When two or more components interact 
with each other, as in the Harmonica reed plate and 
harmonica reed tolerances are very important. For example, 
If the reed tolerance is on the plus side, and the slot 
tolerance is on the minus side, then you have a real tight 
fit, and a great harmonica. If it is the other way around, 
and the tolerances stack up wrong, then you have an airy, 
less than perfect product.

I figure that this is why harmonicas are not all the same 
out of the box. The average tolerances, I am guessing 80 
percent, are the ones that you get out of the box which work 
as designed. The tight tolerance matchup, let say 10 percent 
are the ones that play great out of the box, and which we 
would probably pay more for if we could consistantly get 
them. The loose matchup, the other 10% are the ones that 
everyone is bitching about.

What is the answer? Maybe there could be 3 prices of lets 
say a Marine Band. Lets say $30 for the Normal Marine Band, 
and $40 for the tight one, and only $20 for the loose one 
(which is still with the original specs) Of couse, to 
separate these, probably adds another manufacturing step, 
and in turn raises the costs.

Before anyone slams me for now knowing what I am talking 
about, I will admit that I know nothing about the Harmonica 
Manufacturing Process, but I spent many years in other 
manufacturing environments. All the numbers I quoted are 
made up, and all of this is IMHO. Tolerance theory is real.

Harvey Berman
_______________________________________________
Harp-L is sponsored by SPAH, http://www.spah.org
Harp-L@xxxxxxxxxx
http://harp-l.org/mailman/listinfo/harp-l