OT: thoughts and other brain-droppings on low volume mfg

[Synthoid]

Oh I don't know... it's always good to have extra shelf space.

 

 

[mate stubb]

More droppings, things you deal with BESIDES just the manufacturing process -

 

For awhile I made custom modular synthesizer panels in the MOTM format. Nothing more than a robust piece of aluminum with paint, graphics, and holes in the right places.

 

After sourcing metal, the correct baked on paint, and a screen printer who could handle fine printing graphics, there is:

 

- taking orders, caching them for the next panel run. Tons of emails to track

- doing custom designs on the PC for new panels

- offering the custom designs to the community to get a minimum viable quantity to manufacture

- calculating prices based on how many were ordered (if a lot of people ordered the same design, there was a price break)

- collecting the money

- ordering painted blank panels, paying for them, receiving them, ordering screens from printer, shipping panels to screen printer, receiving them, sorting them, drilling thousands of holes

- packing and shipping them all over the world. Oy! Customs forms! Some want insurance, some don't. Multiple trips to the post office.

 

It's a lot of work. I did it for love of building synths, and so I could afford to pay for my own custom panels!

[OB Dave]

Not my field but wasn't 3D printing designed to facilitate affordable low volume manufacturing?

 

3D printing is way too slow, way too expensive, and the material too brittle for most manufacturing scenarios. It's really great for the design process though. When I had the knob made, the first step in the process was to hire a Solidworks guy to make a 3D model of the knob and I had prototypes made using a 3D printing process called PolyJet. It was the most amazing thing. Contractor emails me the file containing the drawing. Used a viewer app to examine the drawing, then emailed up to a company in Anaheim, three days later parts arrived in the mail. Having those samples and the drawing made it much, much easier to then go around to plastics houses to get bids on injection molding. 3D printing greatly sped up the design and bidding process, but it would have cost a small fortune to actually produce the knobs on a 3D printer.

 

Moore Law appears to be dead for microprocessors, but it still applies for emerging technologies like 3D printing. I've been wanting to buy a small 3D printer to tinker with but every year they get much cheaper and much better.

 

I look forward to seeing where this is all headed. I do think great things will happen for on-demand 3D printed manufacturing processes, but we have a bit of a wait ahead of us.

 

There are also on-demand CNC processes that are pretty remarkable. My favorite anecdote is some guy who's into obscure 1950s English automobiles. I forget the brand, they're long out of production. There's some suspension piece that's notorious for cracking, and there's no replacement parts to be had. Unobtanium, but he managed to get a 3D CAD drawing of the piece made, and has made the drawing available for other Hill freaks, so when their suspension piece breaks, they download the file, send it over to one of the many lights-out-manufacturing CNC mills and gets a brand new part fabricated.

 

[OB Dave]

By the way, for those of you in the Southern California area, there's an interesting trade show held at the Anaheim Convention center about three weeks after NAMM. It's quite a shock to go back there so soon after NAMM and attend a trade show that's not crowded or loud. Anyway, it's actually 6 or 7 trade shows rolled into one and it takes up the whole convention center: rapid prototyping, plastics, electronic components, manufacturing, packaging, factory automation, and occasionally really creepy medical devices and surgical machines. It's free to attend, you just need to preregister on the web.

 

There's always a bunch of the latest-generation 3D printing machines there. One of the most impressive things I saw there was a 3D printer that printed out a little car wheel, perhaps for a toy. But it was able to print multiple materials in one go, so the rim was made of some sort of chromed plastic, the tire was rubber, and it had some detail work done in some other color. Apparently it takes most of a day to print, but still pretty damned impressive.

 

Perhaps we should organize a KC hang and RedKey lounge around this trade show as well...

[Josh Paxton]

I can relate to this thread more than ever since I've started pursuing a little side-hustle involving unique art objects that I've learned to make. They're made of paper, so the material cost per unit is nearly zero. But even though I've automated part of the process with a die cutting machine, it still takes me anywhere from 15 to 45 minutes to make one. That time will improve as I get better at making them, but still I'm estimating that the minimum I'll be able to charge for them to make it worth my while is $20 or $25. So the trick will be finding people who are willing to pay that for what basically amounts to a small cut-up piece of paper. Wish me luck!

[MathOfInsects]

I can relate to this thread more than ever since I've started pursuing a little side-hustle involving unique art objects that I've learned to make. They're made of paper, so the material cost per unit is nearly zero. But even though I've automated part of the process with a die cutting machine, it still takes me anywhere from 15 to 45 minutes to make one. That time will improve as I get better at making them, but still I'm estimating that the minimum I'll be able to charge for them to make it worth my while is $20 or $25. So the trick will be finding people who are willing to pay that for what basically amounts to a small cut-up piece of paper. Wish me luck!

 

Much luck, and may you be better at it than I am.

 

Signed,

Math "Buy High, Sell Low" of Insects

[RichieP_MechE]

Good thread, Dave, on the considerations of lower-volume manufacturing. I agree with others that these days, crowdfunding is the way to go for these sorts of projects.

 

Not my field but wasn't 3D printing designed to facilitate affordable low volume manufacturing?

3D printing is way too slow, way too expensive, and the material too brittle for most manufacturing scenarios.

This was true maybe 8-10 years ago but is not really the case anymore. You can get very durable prints from processes like SLS printing, and the prices are more reasonable as well. Of course, for the average joe, the prices I pay for 3D printed parts at my day job (designing robots) may seem high, but compared to machining a complicated geometry out of plastic, it is definitely more cost-effective and just as functional with the material properties I need. Even FDM printers like the Makerbot can produce parts with adequate strength for certain everyday uses. Speeds have improved as well; depending on part size and quantity, some 3D printing services offer 24-48 hour turnaround.

 

3D printing in metal (DLMS) has also really advanced to the point that GE is using 3D printed parts in aircraft engines and Boeing is 3D printing strutural components for the 787. Of course the cost for a DLMS machine runs into the millions of dollars and parts can still be expensive. However, some of the rapid turnaround manufacturing places I use (such as Fast Radius) are actually recommending 3D metal printing over machining in certain cases if the volume of parts is right.

 

The next nut to crack with 3D printing I think is quality of finish right out of the printer. Many if not all require secondary operations to get smooth surfaces and to remove support structures. Any part that comes out of a DLMS machine needs secondary machining. The multi-material printers are getting better, but something that could print a "finished" product all in one go like a populated circuit board is a lifetime away at least.

[cphollis]

A interested bystander in all of this, but a fascinating discussion. At least for me.

 

Here is this boutique market of MI consumers. Here is this other reality of what it takes to put a product in the market. A fascinating dynamic, at least to me.

 

The vendors are sometimes driven by extreme passion. Hey, I've figured out this great thing. Pay me.

 

The consumers are generally not wealthy. They work for a living. So maybe they splurge on the new thing, maybe not. Mostly not.

 

Personally, I'm fascinated by the MI products that do well. Ventilators, for example. And Aspen's gig with his SSv3.

 

All good. I'm a free-market capitalist by nature, so it's all grist for the mill.

 

 

[WesG]

These threads are dangerous. I have a bad habit of scratching my own itches. That can get expensive, quickly. Luckily I've become hooked on vacuum tube organs recently......nice change from the day job. Although I accidentally got pulled into R&D on some organ-related stuff........argh. LOL.

 

I'm curious what percentage of musicians are "makers". There's a bunch in this thread.

 

[GRollins]

I'm curious what percentage of musicians are "makers". There's a bunch in this thread.

 

Full woodworking shop, here. Also do some mechanical stuff. Which leads me to...

 

I'm a luthier. That term may not mean anything on a keyboard forum, but a luthier is someone who builds or services stringed instruments. I build my own basses and have a guitar in progress that I'd like to finish if I can get the time. I don't do kits. I start from raw sawmill planks and carry all the way through to a finished instrument.

 

I also design and build analog electronic circuits. That comes in handy if you're fond--as I am--of electric guitars and amps.

 

As far as keyboards go, I've got a dead Kurzweil K2500 that I'd like to bring back to life and, as of yesterday, I've found myself confronted with the opportunity to snag a Hammond E-143 for free. See separate emergency "oh, hell, what do I do?" thread currently in progress.

 

Not to mention the usual parade of family members bringing me broken whatsits and asking, "Poppa, can you fix my [fill in the blank]?" Plus plumbing, carpentry, electrical, automotive...you guys know the drill (ahem). As I've mentioned elsewhere, I'm currently up to my behumpus in a master bathroom rebuild, which is on hold so I can finish an African ribbed mahogany desk top while the weather is nice and I can do some of the sanding outside. Also working on damaged finishes on two guitars with the same reasoning. Etc. etc. etc.

 

I've had 48 hour days on order for what seems like forever, but the factory is backordered. No telling if or when I'll get 'em.

 

Grey

[jverghese]

Hi,

 

Interesting thread.

 

I used to make a rack version of the ARP ProSoloist, mainly used by Genesis tribute bands and other hardcore fans. To me the most cumbersome thing about these projects is the logistics, i.e. ordering the materials, arranging and labeling endless component reels, trays, etc.

 

If you have to use subcontractors (which you often do), initial costs can be prohibitive. Circuit board films, pick-n-place programming, custom mechanical parts, etc. Thankfully, production files can be reused for future production batches.

 

I've spent the last two years building a new house, so very little activity on the synth building front. However, I have a number of dsp algorithms lying around which I'd like to make something out of (has to be a hardware product, I'm not into plug-ins). Got to find a new electronics subcontractor, though, as the one I used previously is no longer around.

 

I've had 48 hour days on order for what seems like forever, but the factory is backordered. No telling if or when I'll get 'em.

This. And watch out for signs of burning out.

 

-joachim

 

[NTL2009]

Very interesting thread, I know it's a little old, but I'm new here and came across it in a search. My career was in electronics manufacturing (from one up protos, to 10's-to-100's for production support equipment, to millions of consumer product volumes), so I can really relate to what Dave has brought up.

 

The Kickstarter approach may have merit, as mentioned this allows you to gauge demand with actual commitments before making any purchase commitments. But this is still a niche within a niche within a niche, so it may be a long shot to hit enough volume to make it worthwhile.

 

But maybe another approach could work - what about we put our collective brains/experience together to design/document a DIY 'kit'. Even in single-digit quantities, these little micro-controllers are cheap (~$5 for an ESP8266, which also has wifi, making it useful for all sorts of other things - or maybe to configure via browser?)) and easy to program (and even a non-programmer can be walked through how to load code that we could make available), there are cheap 16:1 analog mux boards available ( < $1 each in 10's). Maybe a PC board to mount some recommended sliders, and some drawings for an enclosure, and I think just about anyone who can handle a soldering iron and isn't too afraid of computers could get one going, or knows or could hire a tech to build it. Offered as "instructional" and there wouldn't be any certification requirements (use at your own risk, etc).

 

example sources for parts:

 

http://got.by/2mj5i <<< "6x1 Analog MPXR board WITH headers (un-soldered) generic - ~ $0.90

 

https://www.banggood.com/NodeMcu-Lua-WIFI-Internet-Things-Development-Board-Based-ESP8266-CP2102-Wireless-Module-p-1097112.html?rmmds=search&cur_warehouse=CN

 

https://www.amazon.com/dp/B010O1G1ES/_encoding=UTF8?coliid=III7RMVIX1RKA&colid=2A7YD8SVABYEF&psc=1

 

 

If you take a look at this following post, you'll see it doesn't seem all that hard:

 

https://forums.musicplayer.com/ubbthreads.php/topics/2911258#Post2911258

 

A DIY approach avoids all the hassles and real-world issue of bringing something to market, up front investment of $$$$$$ and making a profit. Collaboration and part selection and a PC layout(if needed - it's mostly wiring switches/pots to the mux/controller - very little circuitry as such) would make it so much easier for a novice to pick up on it.

 

My personal interest is I'm looking into buying a single manual clonewheel, but I want a lower manual & pedals set up for home, and I want the 'full Hammond experience' with 4 sets of drawbars plus pedal drawbars, and I can't get that in a (compact - I don't want the extra octave presets) single manual.

 

Thoughts? If there is interest, what would be a good place to collaborate, and then maybe provide summary updates here on this forum? Githib, a google group or something?

 

 

[Dave Bryce]

So that's what I'm doing wrong...not multiplying by four to begin with...

4 x landed BOM is probably not enough to bring the average product to market, unless you wanna do it on a shoe string.

 

 

Imagine your unit cost if you had to factor in the overhead expenses associated with having employees.

Yeah, there's that.

 

 

More droppings, things you deal with BESIDES just the manufacturing process -

 

For awhile I made custom modular synthesizer panels in the MOTM format. Nothing more than a robust piece of aluminum with paint, graphics, and holes in the right places.

 

After sourcing metal, the correct baked on paint, and a screen printer who could handle fine printing graphics, there is:

 

- taking orders, caching them for the next panel run. Tons of emails to track

- doing custom designs on the PC for new panels

- offering the custom designs to the community to get a minimum viable quantity to manufacture

- calculating prices based on how many were ordered (if a lot of people ordered the same design, there was a price break)

- collecting the money

- ordering painted blank panels, paying for them, receiving them, ordering screens from printer, shipping panels to screen printer, receiving them, sorting them, drilling thousands of holes

- packing and shipping them all over the world. Oy! Customs forms! Some want insurance, some don't. Multiple trips to the post office.

 

It's a lot of work. I did it for love of building synths, and so I could afford to pay for my own custom panels!

 

...and then there's that pesky marketing stuff, because if no one knows about your product, etc...

 

dB

[MurMan]

I've made it to the brink of launching a small company to sell a small electronic device four separate times. The devices did very different things, but were all based on simple to medium complexity embedded microcomputer designs. Last one was cloud-connected smoker monitor/controller with a smart phone app. Remote control, cooking profiles, graphics, etc.

No - not that kind of smoker!!!

 

The product requirements, design, development, and proto testing parts were great. But, when the time came for the big commitment to a production run came, I backed out. My work experience of production nightmares, ECO's, and field returns was sobering. But in the end, it was all the other functions, like sales, marketing, and running the business that kept me from pulling the trigger. If I had teamed with the right partner, things might have been different. Hardware is tough and covering everything by yourself is just too much.

[Threadslayer]

If I had teamed with the right partner, things might have been different.

Don't bet on that. You can end up with an underperforming/dishonest partner and then you're stuck doing his work as well as your own - but for half the profits. (Been there. Done that. Still pissed off about it.)

[MurMan]

Don't bet on that. You can end up with an underperforming/dishonest partner and then you're stuck doing his work as well as your own - but for half the profits. (Been there. Done that. Still pissed off about it.)

Agreed. That's why I wrote "right partner". Sorry about your bad experience.

 

It's often said that a business partnership is like a marriage. You've got to trust your partner, but you might know if they are trustworthy until years into the relationship. I had a few potential partners, but I didn't trust that they could hold up their end of the bargain. They seemed to be "interested", but I was looking for "committed".

[mfnickster]

@jverghese Joachim, I'm guessing I'm too late to buy a ProSoloist Rack+ from you (or maybe you don't even hang out here anymore) but is it possible to get your DSP algorithms to implement in, say, Reaktor or Pure Data?

 

Many thanks!