replacement rotors

[peteedwards]

replacement valve inner rotors.

IMG_2815.JPG

IMG_2814.JPG

These are 3D printed in 17-4 stainless steel then machined the contact surfaces on a manual lathe, honed and minimally lapped to the existing outer casing.
I designed them using geometry similar to the concept of CAIDEX rotors, which I developed with the generous guidance of Dr. Brian "Benny" Leonard, known here on the Trombone chat as sesquitone. He co-developed the CAIDEX valve. If you are not familiar with CAIDEX valves (acronym Constant Area Internal Ducts of Elliptic Cross Section) they are similar size to traditional rotary valves, but with constant cross sectional area throughout the valve ports, at the same time eliminating sharp corners within the air passage. Because of the geometry, air flow is not disrupted or restricted in any way. The sound wave encounters no sharp edges walls or turbulence, ensuring a more accurate acoustic pathway. The width of the valve allows for the bore to expand a little in the lateral direction, compensating for the fact that the valve diameter would not otherwise allow for 2 full-round full bore passages through the valve.
My valve design is a variation on this concept, using more of a four sided squished circle (or rounded quadrangle) rather than the ellipse of the CAIDEX, but the effect is the same.
3D printing allows for thin walls & low mass, less than 1/2 of the "standard" brass valves they are replacing and have a rotational moment of inertia less than 1/3. This means they are faster & can use lighter springs on the levers. 3D Printing technology allows for geometries such as undercuts and fine details which would be prohibitively costly if not impossible using traditional machining methods. It saves the expense of setup & machining on a 5 axis mill, each one can be different with no added cost. The 3D printing service is less than 2 week turnaround.
The results are really phenomenal. I really didn't expect a dramatic difference because the valves had already been modified ("scooped" to an equivalent cross sectional area), but I guess the elimination of shrp edges & corners in the sound path makes all the difference.
The sound of the horn is more centered & focused particularly in the low register. Playability is vastly improved, with narrower slotting throughout. Response is quicker & it is easier to play with a wider dynamic range. Slurring between open & valve notes is much smoother, due to the port geometry allowing for natural venting in the valve as it travels.
Best of all this is a risk-free, reversible modification. No major surgery need be done to an instrument to make significant improvements to the playability. No soldering, re-lacquering, re-linking. Just a drop in replacement.
For a prized vintage horn which can sound glorious but may be a bit more difficult to play, this would be the perfect solution.

[Burgerbob]

Very cool!!

[JohnL]

I'm curious - why 17-4 PH? Are you heat treating them?

[peteedwards]

I'm curious - why 17-4 PH? Are you heat treating them?

Honestly, it is the least costly metal material to print.
Printed brass uses a lost wax process, costing ~3x as much.
17-4 machines quite nicely and won't corrode like brass, and I can print thinner walls. It is a bit nerve wracking to tap. I may try printed threads next time, a little undersize to just clean up with a tap.
I'm not hardening them.

[peteedwards]

I'm thinking I could offer this as a service- a low-cost, low risk alternative to a full valve replacement. Think of it as angioplasty vs heart transplant surgery.
I'm thinking cost would be around $200 per valve, a bit less for 2 or more.

[Burgerbob]

I'm thinking I could offer this as a service- a low-cost, low risk alternative to a full valve replacement. Think of it as angioplasty vs heart transplant surgery.
I'm thinking cost would be around $200 per valve, a bit less for 2 or more.

How would this work? We send you a valve core to copy dimensions from?

[brassmedic]

If you lapped stainless into brass casings, I would assume most of the metal that was removed was from the casing. How is that reversible?

[peteedwards]

If you lapped stainless into brass casings, I would assume most of the metal that was removed was from the casing. How is that reversible?

I machined them very close to the original dimensions. Honed & deburred, the lapping was very minimal, I didn't use any abrasive at all, just a few turns by hand with oil. I guess I shouldn't have called it lapping if no abrasive was used.

[peteedwards]

How would this work? We send you a valve core to copy dimensions from?

Yes I would need a valve core for a day to reverse engineer, then a couple weeks later I'd need the valve section for fitting.
Of course if I already had the dims in my "library" the first step can be skipped.

At present I have dims for Bach 36/42/50, King 7B/8B

[ithinknot]

Reversibility might be a risky claim, but if the likely use case is upgrading/reviving old/worn rotors, then lapping seems like fair game. You could also offer marginally oversized cores of conventional design as an alternative to traditional valve rebuilds via plating.

Very interesting and thanks for sharing. I have the same design suspicions about this as I do the CAIDEX proper; the internal path geometry is super cool, but reducing the external sealing areas down to bare-minimum lips - far exceeding any alteration necessary just to provide a vent path - strikes me as "optimizing" for its own sake. Yes, it's officially lighter and faster, but it's unclear whether that's significant in a way that justifies the trade-offs. (No one complains about the mass or speed of conventional rotors, and indeed they keep buying alternative designs literally all of which are worse on both fronts but which have other favored characteristics.) The reduction in compression and reduced service life is, however, certain.

[peteedwards]

Reversibility might be a risky claim, but if the likely use case is upgrading/reviving old/worn rotors, then lapping seems like fair game. (You could also make replacement marginally oversized cores of conventional design as an alternative to traditional valve rebuilds via plating.)

I agree that all bets are off for reversibility for an overly worn rotor. Because the machining is done on a manual lathe, I need the casing to do precise fitting on a case by case basis, even with new-ish rotors. The old rotor would not be much worse off than it was prior. The printing is of course oversize to machine to final size.

Very interesting and thanks for sharing. I have the same design suspicions about this as I do the CAIDEX proper; the internal path geometry is super cool, but reducing the external sealing areas down to bare-minimum lips - far exceeding any alteration necessary just to provide a vent path - strikes me as "optimizing" for its own sake. Yes, it's officially lighter and faster, but it's unclear whether that's significant in a way that justifies the trade-offs. (No one complains about the mass or speed of conventional rotors, and indeed they keep buying alternative designs literally all of which are worse on both fronts but which have other favored characteristics.) The reduction in compression and reduced service life is, however, certain.

There is no wear on the outer diameter ('lips") of the valve when it is fitted correctly. The spindle has tighter clearance than the outer diameter.
Your other points are well taken, but keep in mind this is intended as a low cost alternative to the usual valve replacement surgery, or as a means of improving or resuscitating a vintage instrument that would be desired to be preserved in its original form.

[peteedwards]

At present I have dims for Bach 36/42/50, King 7B/8B

and Meinlshcmidt LB (.562) tenor

[WGWTR180]

At present I have dims for Bach 36/42/50, King 7B/8B

and Meinlshcmidt LB (.562) tenor

I sent you an email a few days back. This process is very interesting.

[brassmedic]

If you lapped stainless into brass casings, I would assume most of the metal that was removed was from the casing. How is that reversible?

I machined them very close to the original dimensions. Honed & deburred, the lapping was very minimal, I didn't use any abrasive at all, just a few turns by hand with oil. I guess I shouldn't have called it lapping if no abrasive was used.

Normally, valves are manufactured by making the core slightly oversized and lapping it into the casing to achieve a precise fit. So each casing is matched to a specific rotor that was custom fitted to that casing - usually stamped with matching assembly numbers. That's why you can't normally purchase a valve core by itself - it wouldn't fit. So are you saying you have found a way to get this same exact fit by machining the rotor by itself, away from the casing?

The spindle has tighter clearance than the outer diameter.

I'm puzzled by this remark. You seem to be implying that it's more important for the spindle to be airtight than the circumference of the rotor. I would think the opposite. Look at Hagmann valves - they don't even have a spindle on one side. The spindle is not what makes it airtight.

[hornbuilder]

Brad, you may be surprised as to how rotors are "matched" to casings in some shops. What you describe is certainly the preference, but it is not the case much of the time.

[peteedwards]

A rotor turns on the spindle which is the only part which makes metal to metal contact with the outer casing, the outer surface should have no contact. We are only talking about a difference of ~.0005" diameter clearance on the spindle vs ~.001" diameter clearance on the air sealing outer diameter, which is plenty close for "air tight", with a layer of oil.
Also remember that even at fff, with a practice mute in, the lungs can only generate about 3psi, so "air tight" in a brass instrument valve is something entirely different than in say, a pneumatic pump. (30 years ago I used to make compressor cylinders for diver pumps that had to be safety factor tested to 4000psi. They had a clearance of less than .00005" on a 4" diameter as I recall)

[TomInME]

I'm thinking I could offer this as a service- a low-cost, low risk alternative to a full valve replacement. Think of it as angioplasty vs heart transplant surgery.
I'm thinking cost would be around $200 per valve, a bit less for 2 or more.

$200 each, installed? That's a very interesting price point for instruments that aren't necessarily worth a full valve-section replacement but could use some improvement (like mine...).

[brassmedic]

A rotor turns on the spindle which is the only part which makes metal to metal contact with the outer casing, the outer surface should have no contact. We are only talking about a difference of ~.0005" diameter clearance on the spindle vs ~.001" diameter clearance on the air sealing outer diameter, which is plenty close for "air tight", with a layer of oil.
Also remember that even at fff, with a practice mute in, the lungs can only generate about 3psi, so "air tight" in a brass instrument valve is something entirely different than in say, a pneumatic pump. (30 years ago I used to make compressor cylinders for diver pumps that had to be safety factor tested to 4000psi. They had a clearance of less than .00005" on a 4" diameter as I recall)

Whatever clearance the spindle has is for a different reason. That's to ensure correct mechanical operation. The air seal is formed at the circumference of the rotor. A loose fit there will result in poor tone quality and difficulty centering pitches. If it's too tight it won't turn, but it does need to be as tight as possible. Having tight spindles will not correct for a leaky rotor body. Also, I've never seen a trombone valve where there wasn't some incidental touching of the rotor surface to the casing. Not required, but does happen.

[brassmedic]

Brad, you may be surprised as to how rotors are "matched" to casings in some shops. What you describe is certainly the preference, but it is not the case much of the time.

I have no doubt of that. I've seen all kinds of bad construction of allegedly pro level instruments. On my Bach 42B that I bought in 1995, the valve needed a big gap between the bearing plate and the casing or else the rotor would not turn. Clearly was not fitted correctly.

[boneagain]

I wonder if the seal might actually improve over the "standard" design?
Seems to me the shape would encourage more of the condensation/oil mix to migrate into the area that needs sealing. Wonder how that could be tested?

The reduction in mass seems pretty significant to me. I've spent quite a bit of time balancing springs and such so that the RETURN of the valve doesn't bounce too much and destablize the air column. Less mass should mean less bounce for the same amount of spring.

This is some REALLY nice work!!!

[peteedwards]

Less mass should mean less bounce for the same amount of spring.

This is especially true for way more massive tuba valves.
Besides the advantages of a smoothed-out soundwave path, although I'm not sure that part of the reason rotary valve tubas are chosen over pistons isn't because they add a bit of un-centered "fuzz" to the sound. This design would help eliminate that.

[Blabberbucket]

I like the idea. I think there is great potential for something like this, though I think CNC machining a new rotor in the original material is probably a better idea.

Doubling down on Brad Closes' concerns...

I am curious how this works well without lapping or re-turning the of the casing. If the rotor is worn to the point that it warrants replacing, there is almost certainly significant wear to the casing. This is typically due to worn or improperly fit bearings with excessive "play" allowing the rotor to tilt or move laterally/vertically within the casing when actuated. Given that the rotor is typically indexed by the stoparm drive to be fit one way and rotate roughly 90 degrees in a single direction, the wear to the casing will be uneven due to loose bearings and the pressure from the lever causing the rotor to tilt or shift within the casing, again due to loose bearings. If the rotor bearings were well-maintained, there would likely be no need to replace the rotor itself.

Plating repairs or manufacturing of a new rotor both typically involve either re-turning of the casing or significant lapping to remove uneven wear of the casing and create a well-sealing surface between the rotor and casing. The replacement rotor is typically slightly oversized from the original due to this process. It is important for the bearings to be well fit for proper action, but ultimately it is the mating surfaces of the rotor and casing the provide a sealing surface for the rotor to act as a "changeable tube" in the trombone.

If the casing is worn unevenly, which it likely is if a refit/remanufacture is due, a replacement rotor may improve seal but will not fix that issue.

[peteedwards]

Thank you for bringing this to my attention.
I suppose I should state that the process is reversible in *most* cases.
I would have to asses each valve & advise on whether the process will be reversible or not.
With my own 2 instruments I was able to fit them without altering the casings, even my 40 year old poorly maintained loosey goosey King valves. I measured the casing bores & made the spindles to fit the casing, rather than just copying the original cores, to eliminate the play I had in the original valves. The new valves seal perfectly, with .001" diameter (.0005" per side) clearance on the "air tight" surfaces. I guess the wear in my particular case was pretty evenly concentric. I will have to beware of valves with more uneven wear as you described.
Anyway my primary goal was to try out the new port geometry without spending $$$ and chopping up a perfectly good instrument. That goal has been achieved in spades.
As a machinist of 32 years, I can definitively say there is NO advantage to CNC machining over 3D printing in this case. The cost difference is huge. The advantage of being able to design features with complex geometry, some of which is simply not machinable, while keeping the rotating mass to a minimum, is great.
Material is not a problem I can offer these in brass, it just costs a bit extra because it is a 2 step print process (lost wax casting).

[peteedwards]

How would this work? We send you a valve core to copy dimensions from?

Option 1 (minimal downtime)
1) You send me the core(s) for measurement, A day later I send them back. Skip this step if I already have your brand/model in my library.
2) After approx 2-3 weeks you send me the complete valve section, I fit the new valve(s) and 1-2 days later I send it back.

Option 2 (minimal shipping cost)
You send the complete valve section, approx 3-4 weeks later I send it back with the new valve(s) fitted.

[AtomicClock]

Are there drawbacks (or advantages) to stainless valve cores? Maintenance? Sound? Galvanic corrosion?

[peteedwards]

Are there drawbacks (or advantages) to stainless valve cores? Maintenance? Sound? Galvanic corrosion?

The main advantage is cost, but also it's harder so it won't wear as fast as brass. SS is less dense so slightly less mass (negligible on this scale).
Galvanic corrosion with SS to brass contact is a concern, but highly unlikely with proper oiling, unless you have very acidic breath condensate.

[peteedwards]

Sound?

I have not done A/B comparisons with printed brass vs printed ss on the same horn, but I find it hard to believe it would make any difference. We are talking about a ~7% difference in density of metal in ~1-2% of the length of the trombone.

[Sesquitone]

replacement valve inner rotors.
IMG_2815.JPG
IMG_2814.JPG
These are 3D printed in 17-4 stainless steel then machined the contact surfaces on a manual lathe, honed and minimally lapped to the existing outer casing.
I designed them using geometry similar to the concept of CAIDEX rotors, which I developed with the generous guidance of Dr. Brian "Benny" Leonard, known here on the Trombone chat as sesquitone. He co-developed the CAIDEX valve. If you are not familiar with CAIDEX valves (acronym Constant Area Internal Ducts of Elliptic Cross Section) they are similar size to traditional rotary valves, but with constant cross sectional area throughout the valve ports, at the same time eliminating sharp corners within the air passage. Because of the geometry, air flow is not disrupted or restricted in any way. The sound wave encounters no sharp edges walls or turbulence, ensuring a more accurate acoustic pathway. The width of the valve allows for the bore to expand a little in the lateral direction, compensating for the fact that the valve diameter would not otherwise allow for 2 full-round full bore passages through the valve.
My valve design is a variation on this concept, using more of a four sided squished circle (or rounded quadrangle) rather than the ellipse of the CAIDEX, but the effect is the same.

*******


Just to follow up, here are some numbers for the CAIDEX valve rotor--taken from a Brochure describing the CAIDEX design in detail in comparison with traditional rotary valves. The green areas in the picture show surfaces "in contact" with the casing/bearing surfaces (via a thin film of lubricant). [The grooves on the spindles are primarily for reducing frictional area a bit. They have a secondary function of acting as a "reservoir" for the lubricant, which can migrate axially--unlike the helical grooves on the Meinlschmidt rotor that distribute the lubricant more forcefully.]

Note particularly the extremely small moment of inertia. And the "wide" natural venting pathways. Both also features of Pete's innovative design, along with his "round-cornered quadrangle" constant cross-sectional area ducts.


.

[braymond21]

Slightly off topic but do you think rotors could be printed out of other materials? I'd love to be able to design replacement rotors or pistons, especially for vintage instruments, but getting a metal-capable 3d printer would be far too costly. I was thinking if there was a material that could be printed on hobbyist-grade machines that would hold up well enough to work.

[Sesquitone]

Slightly off topic but do you think rotors could be printed out of other materials? I'd love to be able to design replacement rotors or pistons, especially for vintage instruments, but getting a metal-capable 3d printer would be far too costly. I was thinking if there was a material that could be printed on hobbyist-grade machines that would hold up well enough to work.

There are some printable high-density carbon-fibre/polymer products that you might think would be good candidates. The problem is that with valves you need very fine tolerances and "perfect" centering to give a tight seal without binding. And non-metal printed materials don't have anywhere near the dimensional stability for that. At best, you would need to print oversized, and then clean-up with machining on a precision lathe. And that is a treacherous process!

[TomInME]

Just got my 90's Bach back today (thanks, Pete!), and these are a substantial improvement. I don't know that it would measure up to top-of-the-line valves, but it was definitely worth doing. I've only had the chance to play it with bone quartet but it's clearly more responsive, the open side of the horn feels really nice, and soft entrances were improved a bunch.

Tomorrow I'll have some alone time and be able to really kick the tires - I'll post again after.

[TomInME]

Had a practice session on it today. Definitely an improvement and worth the money. I don't think it's up to the level of modern top-quality valves (partly due to other differences beyond just the cores) but it's a big step in the right direction.

The open side of the horn had the biggest improvement: everything is more stable/centered and you hear it in the color. Pedals might be a bit more open and require more focus, but they also have a little more core. Upper register still takes work, but the stability does help.

I need more time to fully evaluate the trigger register but it did improve significantly and is also more stable/centered, but not quite as much of a change as the open side (or maybe I'm noticing it less because the open side is better). Notes speak a little better and are easier to drop into. Color seems a little darker but it lights up just fine if you want it to, and there's less dirt in the sound. Low B is more stable, which was a big part of what I was looking for.

For the price, I think this is a great solution for horns like mine, a much much better value than a full conversion even if it's not quite at that level.

[peteedwards]

Hi Tom,
Thanks for the quick review. You are the first to play on these valves besides myself, and you confirm most everything I experienced.

[Sesquitone]

Very interesting and thanks for sharing. I have the same design suspicions about this as I do the CAIDEX proper; the internal path geometry is super cool, but reducing the external sealing areas down to bare-minimum lips - far exceeding any alteration necessary just to provide a vent path - strikes me as "optimizing" for its own sake. Yes, it's officially lighter and faster, but it's unclear whether that's significant in a way that justifies the trade-offs.

I want to address your "suspicions" about "reducing external sealing areas down to bare-minimum lips". One of the "problems" with conventional (open-channel sculpted-butterfly) rotary valves and intact-duct rotors with drilled holes for venting is that there is a tendency for a slight "catch" to occur in between attachment and slide alone notes (or vice versa) when lip slurring while actuating the valve, especially if a wide position change is also involved (causing a rush of air to occur). This is due to a very slight blockage of the air-stream in the intermediate orientation causing a build-up in pressure, which is then suddenly relieved when the other active orientation is reached. [The original Rotax valves were of an intact-duct design but with no venting whatsoever. Lip slurring as described was accompanied by very loud "popping"—like someone snapping bubble-wrap! The design was quickly changed to a more conventional sculpted butterfly with partial wrap-around.]

With intact-duct designs like the CAIDEX (with its elliptic cross-sections) or Pete's replacement rotor (with its "rounded-quadrangle" cross-sections), it's important to have as "open" a venting path as is practically possible—in order to try and minimise the valve-actuated-lip-slurred-notes "catch" problem described above. As a matter of fact, the lipped-tube geometry actually eliminates the catch entirely. In case you are thinking that Pete and I might tend to be biassed about this, let me mention that when well-known professional trombonist Ed Zadrozny was initially evaluating one of my other lipped-tube designs (the ULTRA valve—Unsymmetric Lipped-Tube Rotary Actuator), the very first thing he mentioned was the total absence of any hint of the "catch" problem, making such valve-actuated lip slurs much "cleaner" and reliable.

So how wide should the lips be? I have experimented with prototypes with lip widths as narrow as 1.5 mm. Provided the stops are aligned precisely, there is absolutely no loss of seal. To add a little "insurance", I usually use a width of 3 mm. This still provides a very "open" venting path. However, precise stop alignment is still mandatory; that's why the CAIDEX valve comes with individually adjustable stops (and a see-through end-cap so that the player can check, at a glance, that everything is aligned properly).

With the lipped-tube design and elimination of non-functional mass there are no "trade-offs" to worry about.



.

[Finetales]

Hmm...I wonder how this would do on a double (French) horn.

[TomInME]

Follow-up: these were definitely worth the price for an older instrument with standard bored-out cores (pic of the original below). It's a night and day difference. The instrument plays more open and consistently now with more core to the sound and less dirt/garbage. For me, it has a noticeably darker sound, but because it is more responsive and efficient I was able to switch to a smaller backbore and still get enough volume on low C/B. It's a bit too dark for my tastes with the Shires slide, particularly in the deeper end of the trigger register, so I'm pairing it up with the original nickel LT slide and liking that combo a lot - after a year of screwing around with equipment, I think I've found a resting place.

The trigger register isn't magically "fixed" but notes respond a lot more easily, and the color difference between low B and pedal B-flat is less pronounced. I don't notice a difference in the activation speed (the originals moved very fast), but because the notes stabilize more quickly, I can do faster things.

For the curious: below is a picture of the old cores, with red lines showing where the walls of the casing would be. Also note that the entrances/exits are abrupt corners rather than rounded because of the straight bore. Pretty clear why they aren't great to play on.

IMG_20240929_163128b.jpg