Page 1 of 1

Notes from the Workshop

Posted: Wed Oct 23, 2019 7:31 am
by Brendan
Hi Friends,

I know new product releases have been quieter of late, but that doesn't mean nothing's happening. Quite the contrary! I'm currently in a purple patch of workshop development where it feels as if many strands are coming together: a virtuous circle of improving skills, discovering new materials, gaining fresh insights, mastering new technology etc. As a result, various projects that got stuck because of some seemingly intractable problem have been resurrected in a new improved form that is more user friendly and potentially commercially viable.

I don't want to say too much before anything is released, because unanticipated problems can always crop up and detail what seems a dead cert. But the improved harps I've been creating include the TwinHarmonica series, both chromatic and diatonic based.

The ones that use chrom reedplates include the DoubleChrom, Tremolo Chrom, OctaChrom, Quartertone Chromatic, X-Reed Chromatic. Past versions were bulky with long pathways to the reeds, resulting in the dreaded Helmholtz Resonance Coupling issue. They worked in lower range keys, but were difficult and expensive to make. I was never fully satisfied with them, and most have been discontinued awaiting a new approach.

About a month ago I had an insight that unlocked the door to making all these instruments much more compact, about the same size as a normal chromatic - despite having twice the number of reeds. Now I'm deep in the time-intensive task of making new versions of all of them. There's still quite a bit of work to do before releasing anything, but I'm confident in the basic design concept, and the prototypes are living up to their promise.

The smaller TwinDiatonic harps are getting a makeover too, with a completely fresh approach. As with the chromatic versions, my current TwinDiatonics (Switch-Harp, OctaHarp, UniBender etc) have the mouthpiece some distance from the reeds - resulting in the same Helmholtz issue of flattened pitch in the higher notes, plus more difficult bending.

Based on a new concept altogether, my latest versions get the reeds much closer to the mouth, helping alleviate both those problems. In one way the design is simpler, in another more complex, and it's still a work in progress - but the overall concept is certainly theoretically superior to what I've made in the past. That's very important in itself, but getting all the tiny details right to turn good theory into reliable working harps always takes a lot of time and hard work. Once I'm satisfied they are rock solid I'll make plans for releasing the new versions.

One recent invention of mine, the Gasket-valve, is proving very helpful in enabling others to improve. I'm finding it can apply to every harp which needs valves to do its thing. The Gasket-valve not only makes fitting valves far faster and easier, it also improves the accuracy of how they fit. This is crucial on some of my new models, which have very tight spaces inside for valves to swing.

On a completely different tack, quite out if the blue I recently got sent this extraordinary instrument by its owner to examine, figure out, and hopefully reverse engineer:

The Pipe Harmonica
https://youtu.be/jEXi8pqlZKY

It's quite a challenge, as the mysteries are locked within and the instrument can't be dismantled. Quite a few people have wanted to recreate the Pipe Harmonica in the 60 years or so since this prototype was made, but no one has succeeded. However, working together with another person who's been trying to crack its secrets for several years, our collaboration has yielded quick success in 3d printing the first pipe pairs that sound on both blow and draw breaths. There's still lots of work involved in this project to make a similar 3 octave Pipe Harmonica to Theodore Folsom's original, but the fundamental principle of how it works is now understood. I'm finding it very interesting learning about the parallel universe of the Flue Pipe.

Ok, back to tweaking, fettling, micro-adjusting until the sweet spots for each design are reached. If only I could get Alpha Zero on board to help, things would go a lot faster!

Re: Notes from the Workshop

Posted: Thu Oct 24, 2019 11:41 pm
by Alan
I always enjoy hearing about your current projects! Thanks for the update.

Re: Notes from the Workshop

Posted: Fri Oct 25, 2019 1:01 pm
by EdvinW
Nice to hear about the progress!

This place HAS been a little quiet lately, and looking back i find I haven't posted in almost two months. Mostly this is due to my work, where it's becoming increasingly clear a PhD thesis doesn't write itself. (Who would have thought, huh?) I hope to get some more time for the forum again shortly.

I have some harmonica related stuff going on though, actually including some ideas for tunings for your double chroms. It will be exciting to see what the new designs look like!

Your final comment about Alpha Zero is interesting. I've been thinking a little about how machine learning (and other maths-heavy techniques) could be used to aid harmonica design. There's no real progress yet, but maybe some day ;)

Re: Notes from the Workshop

Posted: Fri Oct 25, 2019 10:53 pm
by Lizzy
Good news Brendan! Always fun to read about your progress :D . One question: i was curious when you in April mention something about an alternative method for getting easy overbends on stock harps, That you finally arrived at a good solution on this. You invented some sort of assembly that you could retrofit onto existing harps and i think it was something different than the overblow booster. Do i remember it correctly or is it an abandoned solution? :roll:

Re: Notes from the Workshop

Posted: Mon Oct 28, 2019 12:47 am
by CrawfordEs
Great news.
Thanks for the update.
I’m waiting for the day I can try a unibender and maybe a new switch harp.

Re: Notes from the Workshop

Posted: Fri Nov 22, 2019 1:24 pm
by EdvinW
Brendan wrote:
Wed Oct 23, 2019 7:31 am
...
I don't want to say too much before anything is released, because unanticipated problems can always crop up and detail what seems a dead cert. But the improved harps I've been creating include the TwinHarmonica series, both chromatic and diatonic based.

...

Based on a new concept altogether, my latest versions get the reeds much closer to the mouth, helping alleviate both those problems. In one way the design is simpler, in another more complex, and it's still a work in progress - but the overall concept is certainly theoretically superior to what I've made in the past. That's very important in itself, but getting all the tiny details right to turn good theory into reliable working harps always takes a lot of time and hard work. Once I'm satisfied they are rock solid I'll make plans for releasing the new versions.
Will the new design be superior enough to allow a half valved twin chrom?

The answer, no matter if it's negative or affirmative, might qualify as "too much". Just asking in case you'd want to share! Otherwise ignore this post ;)

Re: Notes from the Workshop

Posted: Sun Nov 24, 2019 7:42 am
by Brendan
Edvin wrote: "Will the new design be superior enough to allow a half valved twin chrom?". Definitely.

As I go through the stages of developing an idea, prototype parts accumulate. When, as normally happens, I come up against a tough problem that inhibits further progress at present, or even appears to kill the idea altogether, I put them in a small box or tub and store it away. Various kinds of plastic food packaging tubs get repurposed in this way: I must have close to 100 storing bits from various projects in my workshop!

Over time, new insights, skills or materials often revitalise these past projects, and it's always a nice surprise to retrieve one of the tubs and rediscover what's inside.

Yesterday I thought of a new way to make a 30-reed diatonic with 3 reedplates. This is an 'old' idea that I first dabbled with over 10 years ago, sharing making methods with my Hungarian friend and collaborator Zombor Kovács. Together we produced the excellent but complex limited-edition X-Reed MB30 in 2015:

https://youtu.be/B5PoDkHdLx8

The MB30 was so hard to make we gave up on it after the initial batch of 50 harps, but the idea of adding an extra reedplate to a 10-hole diatonic in some simpler way has persisted. I have several 'idea tubs' of different approaches relating to it. It's surprising how many options there are once you delve into it, but none of mine has yet crossed over to being worthy of release.

The basic idea is to add a third reedplate with suitable x-reeds to allow the normally 'straight' reeds to pitch bend with interactive-reed bending. In Richter Tuning that's the lower 6 blows and upper 4 draws. If you adopt a regular-breath tuning like PowerBender it's relatively simple, but the breath crossover at hole 7 in Richter Tuning makes it much trickier. Since Richter is what 99% of harp players use, it's always been my aim with any 30-reed harp idea to come up with a good Richter version.

You can easily experiment with this concept yourself. Just get a C harp and the blow reedplate from an A harp. Make a triple plate sandwich, with the A reedplate in the middle of the two combs and the two C plates on top and bottom as normal. It creates a fat diatonic which can be played for testing purposes.

With your mouth over a vertical double hole on the whole harp, the standard C blow and draw notes sound. Using hole 4 as an example, you can play blow C, draw D and bend the D down to C#. You can't bend the blow C down with your lips over the whole harp, but if you just play the top chamber, suddenly you'll get a satisfying blow bend down to Bb.

Nice! But that's when it gets difficult. How do you develop this simple idea into a pleasing, workable harp? First up, Richter Tuning creates a problem. You can't bend the upper 4 draw reeds without swapping reeds around. Then there is the ergonomic aspect: the fat, awkward shape in your mouth, the inconvenience of tilting the harp to get extra bends...

I made a version of this simple two-comb sandwich construction for Paul Bowering, one of our forum members. To make the harp playable without tilting I added a mouthpiece at the front plus two sliders, which selected the x-reed bends when desired. It worked, but the added complexity and bulk of the harp meant it ended up as just another interesting curiosity as far as I'm concerned.

I've tried so many other approaches, but all involve a considerable amount of modification work, special comb parts, reed swapping and, generally, valves. I've got lots of working prototypes, but nothing yet has ticked enough of the boxes to convince me that it would be attractive for players to buy.

Yesterday's 30 reed idea has been fun to explore, and I'll continue with it today. Construction is pretty easy and the sound is good, but the harp remains bulky, still requires reed swapping for Richter Tuning, and needs a mouthpiece: all negatives that will likely preclude it
from seeing the light of day as a commercial model.

But hope springs eternal that I'll one day have a 30-reed Eureka moment that doesn't end up dwindling into Ho Hum So-Soness. Maybe today? That's what gets me up in the morning :-)