[pianotech] brass rail duplication

[Encore Pianos]

What I will probably do is just drill new holes with counter sinks away from
the old locations, then drill the holes in the wood rail.  Run a shim of
.035 under the brass plate ( I've thought of using short snips of piano wire
of .035), then screw the plates in place.

Will  

-----Original Message-----
From: pianotech-bounces at ptg.org [mailto:pianotech-bounces at ptg.org] On Behalf
Of Bruce Browning
Sent: Wednesday, September 12, 2012 12:03 AM
To: pianotech at ptg.org
Subject: Re: [pianotech] brass rail duplication

Hi,
One solution, to save time now, is to drill out the holes which attach the
brass rails to the wooden action rail, using drill approx. 0.030"
larger thus giving some degree of adjustment.
Annoying but doable.


On 12/09/2012, Encore Pianos <encorepianos at metrocast.net> wrote:
>
>
> To the list:
>
>
>
> A few weeks ago I sent two brass rails from a big old Henry F. Miller 
> upright to a supplier (who shall remain unnamed) for duplication.  
> After several weeks I got them and put them back on the wooden action 
> rails, mounted the original parts onto them with the brass rail butt 
> plates, put everything together and went to my customer’s house to put 
> it back in and go from there.
>
>
>
> The first thing that jumped out at me was the fact that the hammer 
> shanks were sitting a good 1/4” off the hammer rest rail.  Moreover, a 
> number of notes were now blocking against the string because the let 
> off had changed also.  It was interesting to note that a number of 
> hammers were wedging lightly against the wooden top of the damper head 
> because there was no longer sufficient clearance.
>
>
>
> Allow me to establish a fixed point before you put your thinking caps on:
>
>
>
> The bottom of the original brass rail was resting on the shelf of the 
> action rail when screwed into place.
>
>
>
> The replacement rail, when screwed into place, was also resting on the 
> shelf of the action rail in the same place as the original.
>
>
>
> So screwing the rails into place did not change that particular 
> relationship, and allows me to use that as a point of departure 
> because it is the surest point of reference since it has not changed 
> from one rail to the other.
>
>
>
> I took several measurements with my digital .001 caliper before 
> getting back to the supplier with my findings.  They were:
>
>
>
> 		
> New
>
> 		New
>
> 		
>
> Hasenjaegar Miller upright
>
> Old
>
> treble
>
> Difference
>
> 	tenor
>
> Difference
>
> 	
> 	
> Rail
>
> Rail
>
>
>
> 	Rail
>
>
>
> 	
>
> Measure bottom of bracket to center pin v
>
> 0.886
>
> 0.851
>
> 0.035
>
> smaller
>
> 0.854
>
> 0.032
>
> smaller
>
>
> Center of screw hole to center pin V
>
> 0.671
>
> 0.650
>
> 0.021
>
> smaller
>
> 0.640
>
> 0.031
>
> smaller
>
>
> bottom of bracket to top of individual mount
>
> 1.032
>
> 0.999
>
> 0.033
>
> smaller
>
> 1.000
>
> 0.032
>
> smaller
>
>
> bottom of bracket to top of brass plate pin
>
> 0.547
>
> 0.528
>
> 0.019
>
> smaller
>
> 0.535
>
> 0.012
>
> smaller
>
>
>
>
>
> I took measurements at 3 points on each rail, they were pretty 
> consistent, variations no more than .001 to .002, the above figures are
the average.
> My
> relatively new caliper is certified to within .0005.  When I compare 
> my readings against a micrometer, variations of .001 or less.  So my 
> caliper is sufficiently accurate and consistent.
>
>
>
> The center pin V is the slot that holds the butt center pin to mount 
> it to the rail with the butt plate. You can see that that distance is 
> .035 smaller on the treble rail, and .032 smaller on the tenor rail
>
>
>
> The center of screw hole to center pin v is .021 on treble rail, .031 
> on the tenor rail.  This measure is compromised by the fact that I was 
> eyeballing the center of the hole, but both are smaller than the 
> originals
>
>
>
> The following measurements are not critical, but instructive nonetheless.
>
>
>
> The bottom of the bracket to the top of the individual mounting arm is 
> about
> .033 shorter on the new rail. As long as this arm is not too long or 
> too short, this measure is not critical
>
>
>
> The bottom of the bracket to the top of the brass plate guide pin is 
> .019 smaller on one, .012 smaller on the other.  This is not critical 
> either, so long as the butt plate is able to full cover the center pin.
>
>
>
>
>
> As you can see, everything is smaller.  As best I can judge, the most 
> accurate and meaningful measure is the first, the bottom of the 
> bracket to the center pin v.  Since the bottom of the bracket is 
> sitting on the shelf with both rails, this can only mean that, in 
> reference to everything else in the action, the center pin v is .035 
> lower than on the original rail.  This means that any part of the butt 
> and the hammer is now sitting .035 inch lower.
>
>
>
> The butt leather is now .035 closer to the jack top, because that 
> position has not changed.  Presumably that is why the hammer shanks are
sitting ¼”
> off the rail now.  The reason why the hammer bottoms are now touching 
> the top of the damper heads.  And, of course, the geometry regarding 
> the intersection of arcs between the butt leather profile and the arc 
> of the jack has changed, and not for the better.
>
>
>
> I spoke with the fellow who made it at the supply house.  I asked him 
> if he had a caliper or other means of measuring needed values to assure
accuracy.
> He told me that he did not, said that he eyeballed it when it was 
> done, and he thought it was good enough.  Apparently they have an 
> indexing system on their milling machine that allows them to duplicate 
> rails without using measurements.  The essence of it is that he has no 
> reliably accurate means of checking his work or whether or not the 
> machine has drifted or worn its way out of tolerance.
>
>
>
> I sent the rails back to be done again, asking that they buy a caliper 
> and check their work. Their technician, not the same fellow, called me 
> to tell me that the he had measured the distances I had given on the 
> new parts with his caliper accurate to .001.  He said both the first 
> and second sets of rails came out essentially the same, and varied 
> from the original by 3 or 4 thousandths.  So he thought my caliper was
suspect.
>
>
>
> We argued about who was right for a while.  He thought it was 
> sufficiently accurate, they had been doing this for 50 years, and they 
> had never had any problems before. He told me I should be able to make 
> it work and besides, what other choice do I have, since nobody else makes
them?
>
>
>
>
>
> Finally, the 64,000 dollar question:  Would any of my dear readers 
> find a tolerance of .035 acceptable, or is that value totally out to 
> lunch?  If you were in my shoes, where would you be finding yourself 
> sitting?
>
>
>
> Thanks for your help.
>
>
>
> Will Truitt
>
>
>
>
>
>
>
>


--
Bruce Browning
The Piano Tuner