rubner tuners

[Marty Jacobson]

By way of clarification and apology, I have been working from bad information, and based on realizing that, I will change my tune a bit, but I do stand by my previous statements.

I just took several sets of the normal Rubner tuners I have on hand, 23mm spacing, and dropped them into pegheads drilled for both 29/32" and 23mm spacings with my CNC machine. Either configuration works fine with Golden Age, Schaller, or Rubner bushings.

Where I went astray was that the reference set of Rubner tuners I used to check drop-in fit on vintage Gibson A models is a reproduction set made by Rubner and somehow that set is a different spacing. Evidently there was a miscommunication somewhere along the line. I was not involved in developing them, just asked to provide feedback on them. Given my experience with their work being very reliable and precise with all the other sets I have used, I wrongly assumed they would be standard spacing (especially since they were made expressly for the purpose of being drop-in replacements on vintage instruments). They are ever so slightly more than 23mm in spacing, but I'll have to ask what the actual spacing is. Sorry for making that assumption, but I'm glad to have realized my mistake, since it makes life easier.

[tom.gibson]

For what it's worth, I've also had a drop-in fit with Rubners on holes drilled with the 29/32 Stewmac jig.

[Dave Cohen]

That small difference between 23mm and 29/32" is multiplied by 3 from the first post to the last one. The cumulative difference is .05625", which is enough to be troublesome, IMO.

I stand by my numbers. I worked them through and checked them before posting. The difference between exactly 23.00 mm and exactly 29/32" is still 0.0008". Multiplying by three for the total center-to-center distance between the outermost posts still gives a difference or error of 0.0023", or approximately 0.002", to the nearest thousandth of an inch. If anyone thinks they can lay out and drill their tuner post holes to better than within +/- 0.002" with calipers and a drill press, I'd like to see that. Needless to say, I am skeptical. Further, even that kind of precision with a CNC is difficult. Unless you spring for a very spendy CNC, the manufacturers usually don't claim repeatabilities better than +/- 0.001". That, and the repeatabilities are also limited by the rigidity of the platform.

Regarding the Jacobson retort about measuring to 0.91". I most certainly did not say that I do that. The 0.91" figure was for the sake of argument. The 29 and the 32 factors both have literally 2 significant digits, i.e., digits expressing the results of measurement. The rule for expressing quotients of measurements is that the quotient can have no more significant digits that the factor having the least number of significant digits. Those rules typically appear in freshman chemistry, or at latest, sophomore physics, maybe even in high school science courses. Look 'em up. The whole point of the argument was that neither of the numbers as stated,, i.e., 29/32" and 23 mm, have enough reported significant digits to specify the needed precision of the tuner post spacing. Presumably the manufacturers make the tuning machines using greater precision, probably to +/- 0.1 mm (=+/- 0.004") at worst, or better, +/- 0.001". Further, as I also stated in my earlier post, it is not clear to me that the manufacturers weren't using the 23.0 mm measurement from the beginning, implying that from the standpoint of commercially available tuning machines, there is no difference.

[peter.coombe]

Good grief. No wonder I always work in metric.

[Kieran]

Saw a good range of beautiful machines heads on display at their stand in Frankfurt last week,also for mandola with longer spacings between posts which could be suitable for CBOM's.

Also look out for a mandolin version of Schaller's high end 'Grand Tune' machines sometime in the near future.The guitar grand tunes have been out a while.


Kieran
http://www.moloneymusic.com

[Marty Jacobson]

Regarding the Jacobson retort about measuring to 0.91". I most certainly did not say that I do that. The 0.91" figure was for the sake of argument.

I never said you did. I said, if your argument were valid, then, for argument's sake, you could.

The 29 and the 32 factors both have literally 2 significant digits, i.e., digits expressing the results of measurement. The rule for expressing quotients of measurements is that the quotient can have no more significant digits that the factor having the least number of significant digits. Those rules typically appear in freshman chemistry, or at latest, sophomore physics, maybe even in high school science courses. Look 'em up.

This isn't freshman chemistry. 29/32" is a nominal value (as is that of any other fractional value called out in a specification), which means it is a name for a specification, not a measured value. Just as drill bits are labeled "A" through "Z", screws 0000 through 10, and drill bits from 1-100. This is established industry practice dating back over 150 years. Machinery's Handbook is more applicable, in which everything is detailed to four significant figures, if not five. Detailed tables are provided for every instance. They are usually posted on the walls of any decently equipped shop which does precision work.

My confusion, which I have already apologized for, has absolutely nothing to do with a preferred measurement standard. It was a miscommunication which lead me to an incorrect conclusion. I actually work in metric a lot more than English units these days, because microns are more convenient than ten-thousandths of an inch at the scale I am often working.

[Jim Adwell]

I stand by my numbers. I worked them through and checked them before posting. The difference between exactly 23.00 mm and exactly 29/32" is still 0.0008". Multiplying by three for the total center-to-center distance between the outermost posts still gives a difference or error of 0.0023", or approximately 0.002", to the nearest thousandth of an inch. If anyone thinks they can lay out and drill their tuner post holes to better than within +/- 0.002" with calipers and a drill press, I'd like to see that. Needless to say, I am skeptical. Further, even that kind of precision with a CNC is difficult. Unless you spring for a very spendy CNC, the manufacturers usually don't claim repeatabilities better than +/- 0.001". That, and the repeatabilities are also limited by the rigidity of the platform.

Regarding the Jacobson retort about measuring to 0.91". I most certainly did not say that I do that. The 0.91" figure was for the sake of argument. The 29 and the 32 factors both have literally 2 significant digits, i.e., digits expressing the results of measurement. The rule for expressing quotients of measurements is that the quotient can have no more significant digits that the factor having the least number of significant digits. Those rules typically appear in freshman chemistry, or at latest, sophomore physics, maybe even in high school science courses. Look 'em up. The whole point of the argument was that neither of the numbers as stated,, i.e., 29/32" and 23 mm, have enough reported significant digits to specify the needed precision of the tuner post spacing. Presumably the manufacturers make the tuning machines using greater precision, probably to +/- 0.1 mm (=+/- 0.004") at worst, or better, +/- 0.001". Further, as I also stated in my earlier post, it is not clear to me that the manufacturers weren't using the 23.0 mm measurement from the beginning, implying that from the standpoint of commercially available tuning machines, there is no difference.

You are correct. I was doing the calculation wrong.

[Dave Cohen]

I never said you did. I said, if your argument were valid, then, for argument's sake, you could.



This isn't freshman chemistry. 29/32" is a nominal value (as is that of any other fractional value called out in a specification), which means it is a name for a specification, not a measured value. Just as drill bits are labeled "A" through "Z", screws 0000 through 10, and drill bits from 1-100. This is established industry practice dating back over 150 years. Machinery's Handbook is more applicable, in which everything is detailed to four significant figures, if not five. Detailed tables are provided for every instance. They are usually posted on the walls of any decently equipped shop which does precision work.

The point of my argument was not that 0.91" could be used; it was exactly the opposite, i.e., that 0.91",or two significant digits, is not sufficient precision to drill a set of holes that will properly accommodate a set of manufactured tuning machines. Nor did I even imply that installing tuning machines is freshman chemistry or sophomore physics or high school science. Those courses/texts are where you can first find the rules for significant digits, but if you have an aversion to such courses or texts, you can find the rules in other places, e.g., Wikipedia, as well.

A 29/32" drill bit may well be a nominal size bit, although at some point, a measurement had to be made, or at least referred to, or duplicated. It is a bit of a stretch to call tuner post spacing a nominal size. Maybe if one uses a jig or some software to accomplish the spacing, the spacing becomes a de facto nominal size. I don't use a StewMac jig or a cnc, so I need three significant digits to drill for and install tuning machines.

[mariongee]

To Marty and David,
HUNH? I just thought they were pretty!

[Dave Cohen]

If you replace a set of tuners in a mandolin, the new tuners have to fit the tuner post holes. If you are building a mandolin, the tuner post holes have to fit the tuning machine posts. That is what the argument is about.

For a long time, Schaller mandolin tuners occupied the best quality/price ratio slot. In the last several years, opinions have changed on that. Regardless, Schaller tuners illustrate the issue of tuner post fit which was raised in this thread. Schallers were/are advertised by American distributors as having a 29/32" spacing between adjacent posts. The Schaller firm originated near Nuremberg, Germany, but is now located in Bavaria. If you go to the Schaller company website and look in the online catalog, you will find a drawing of the Schaller mandolin tuners with dimensions included, and yes, the dimensions are in mm, decimal points and fractional millimeters included. The spacing between adjacent posts is given as 23 mm. That means that for all practical purposes, the 29/32" spacing is the 23 mm spacing. Calling that spacing a "nominal" designation and assuming that 29/32" and 23 mm are different is what has generated the confusion about fit.

It is a truism that in crafts, one generally tries to avoid measurements, instead using reference points/lines and various "tricks" to obviate the time and uncertainties involved in making measurements. In that case, measurement, along with actually trying the tuner post fit, is what resolves the dispute.

[Hendrik Ahrend]

If you replace a set of tuners in a mandolin, the new tuners have to fit the tuner post holes. If you are building a mandolin, the tuner post holes have to fit the tuning machine posts. That is what the argument is about.

For a long time, Schaller mandolin tuners occupied the best quality/price ratio slot. In the last several years, opinions have changed on that. Regardless, Schaller tuners illustrate the issue of tuner post fit which was raised in this thread. Schallers were/are advertised by American distributors as having a 29/32" spacing between adjacent posts. The Schaller firm originated near Nuremberg, Germany, but is now located in Bavaria. If you go to the Schaller company website and look in the online catalog, you will find a drawing of the Schaller mandolin tuners with dimensions included, and yes, the dimensions are in mm, decimal points and fractional millimeters included. The spacing between adjacent posts is given as 23 mm. That means that for all practical purposes, the 29/32" spacing is the 23 mm spacing. Calling that spacing a "nominal" designation and assuming that 29/32" and 23 mm are different is what has generated the confusion about fit.

It is a truism that in crafts, one generally tries to avoid measurements, instead using reference points/lines and various "tricks" to obviate the time and uncertainties involved in making measurements. In that case, measurement, along with actually trying the tuner post fit, is what resolves the dispute.

Schaller is still located near Nuremberg, in the Region of Franconia (part of Bavaria). Schaller's mandolin string post spacing is - indeed - 23 mms, just like the vintage F5 (Waverly) spacing from the '20s, also 23 mms. (Except for early Loar tuners and some individual examples with wider spacing.) Now, did they really shoot for 29/32'' instead of 23 mms - for all practical purposes?

[Dave Cohen]

Schaller's website says that they weren't founded until 1945. The de facto 23 mm spacing was established before that, so I have no way of knowing how that spacing was established without doing some digging that frankly doesn't interest me. 29/32" is an odd little fraction for a spacing designation, especially when you consider that the originators of those mandolin tuners would have had to abandon rational fractions and use decimal measurements to more decimal places in order to carry out the machining. That, and the fact that 23.00 mm is within 0.0008" (or ~0.02 mm) of 29/32", give me the opinion that the tuners were originally made to 23.00 mm specs.

[Marty Jacobson]

It is a truism that in crafts, one generally tries to avoid measurements, instead using reference points/lines and various "tricks" to obviate the time and uncertainties involved in making measurements. In that case, measurement, along with actually trying the tuner post fit, is what resolves the dispute.

Indeed.

[mariongee]

I understood what the argument was about, and I got the math. I was just trying to lighten things, by my mistake, with humour..Making fun of my redneck lack of technical knowledge. I am not a machinist by trade, and the only time I have to deal in thousandths of an inch is when I rebuild a centrifugal chiller, which is very rare these days.
I make fun of myself quite a bit. It reminds me I'm not missing the biggest joke: me!

[John Flynn]

+1 for Rubners. I installed a set on an OM I used to have. Very nice. I would put them a bit above the top end Grovers. When I ordered them, they made more different models for mandolin than they show on their website, such as F-style, slotted headstock tuners and plainer, less expensive models. The website seems to just show the really fancy ones. You might contact them to see if they still make the others.

[Kieran]

Here's the mandolin pages from their current printed catalogue, and a gold set with rosewood buttons.



Kieran
moloneymusic.com

[John Morton]

Schaller's website says that they weren't founded until 1945. The de facto 23 mm spacing was established before that, so I have no way of knowing how that spacing was established without doing some digging that frankly doesn't interest me. 29/32" is an odd little fraction for a spacing designation, especially when you consider that the originators of those mandolin tuners would have had to abandon rational fractions and use decimal measurements to more decimal places in order to carry out the machining. That, and the fact that 23.00 mm is within 0.0008" (or ~0.02 mm) of 29/32", give me the opinion that the tuners were originally made to 23.00 mm specs.

Here's my machinist take on this: it is true that 29/32" is odd, but in the world of English units there is a custom of designing with fractional dimensions, even 32nds if there is a critical spacing issue. But for a German product, 29/32" is improbable. The difference from 23 mm across 4 pegs amounts to .0022" or so. It follows that if you split the difference in spacing (.906", pretty much) and drill the holes .005" oversize, everything will clear.

Measuring across the shafts I have found spacing variations showing that production/assembly methods often exceed these small tolerances. IMHO a design is faulty that requires such an exact hole spacing. Eyeballs on a ruler are unreliable inside .005". If you really have control to .001", and the fit really matters, you should measure everything and map out your pattern.

[Ron McMillan]

Here's the mandolin pages from their current printed catalogue, and a gold set with rosewood buttons.



Kieran
moloneymusic.com

Kieran do you sell them? I'm going to be in Britain for a short visit in June and a friend has asked me about picking up a set of A mandolin Rubners for him.

ron

[Kieran]

Possibly very soon Ron - will PM you.

These Rubners going on this 20 year old KM-160S.The Korean set didn't do too bad up to now.

[Mike Floorstand]

I just replaced the original 12:1 tuners on a 1921 Gibson A3 with Rubner 15:1, now much smoother. It was not a drop in fit as the Rubner bushing is wider - fortunately I have a reamer. 5 screw holes instead of 3 but since it's now worm over the screw holes are nowhere near the old ones. Kent at Rubner US was super helpful helping me track the package when Royal Mail made a mess of the delivery here in UK.

[hank]

Mike why did you choose the worm over configuration instead of worm under like the original waverlys. The top tuner shafts are so long on the F model tuners it looks like it could be more easily bent. I bought a pair of F model worm under for an F5 kit build before the worm over were available and am not sure if I'm missing something about why use worm over or under other than trying to match the set your replacing.

[Mike Floorstand]

Hank, to be honest when I ordered them I didn't even think about whether they were worm over or under. But these are A model tuners so all the shafts are the same length, I found I was able to install them as they are intended - worm over - without any problems. I could have installed them worm under to match the original set but then they would turn the wrong way.