EWIUSB.COM
Electronic Wind Instrument Information and More

Technical Information

The technical documentation that comes with the EWI USB is a bit on the sparse side, and naturally there is little information about what is going on under the hood. To the left you will find a list of links containing pictures of the inside of the device, reverse-engineered NRPN and System Exclusive data, and other information that is either not provided by Akai, or corrections to their documentation. Further on this page you will find a review of the device itself and a copy of the specifications as stated by Akai.

A Technical Review of the EWI USB
By Jason M. Knight

The device itself is remarkably simple. The use of electroconductive keys is radically different from most midi controllers. What does electroconductive mean? There is a grounding plate that sits under your right thumb, and a second one (that is semi-optional to touch) next to the "octave rollers". When you touch the actual keys or rollers, you complete an electrical circuit through your body tripping a "keypress". This has the advantage of no moving parts increasing the life of the device, though excessively dry skin or air, as well as oil buildup on the keys can cause issues. Most players agree that skin lotion alleviates most of the "dryness" issues, though I've found the chemicals in them can cause issues with the keys themselves... Personally I've found that cleaning the keys with a 50% watered down solution of pine-sol clears up the patina without damaging the metal, though it can cause some discoloration.

As I determined during dissassembly the internals of the EWI USB are remarkably simple. Three single layer circuit boards, only one of those boards having any "meaningful" componants on it the heart of which being to very small surface mount IC's. This was not a suprise to me as from an electronics standpoint there is little need for a USB version of the EWI to be much more complex than your average USB game controller. Three analog inputs and 17 switches are not exactly rocket science these days, nor is MIDI output. I've still not identified the primary chip in the EWI, though I SUSPECT it is something like a PIC18F4550 or other PIC controller. If I was going to build one from scratch (which I've been considering doing for years) that's my current weapon of choice, since it retails for under $5, can easily handle Midi over USB, has more than enough built in Flash and RAM to handle turning switch inputs into a midi stream, and includes at least a dozen more inputs than you would need. It's actually why I'm suprised the USB model costs as much as it does, as they can't cost more than $100 to actually make - I could probably bring one in under $50 if so motivated and had access to mass fabrication... Makes me wonder what a toy like the Casio DH-100 would cost if you decided to make one today as a USB controller.

The bite sensor is incredibly simple - the two spring-steel sheets are separated by a sheet of plastic, effectively making a one layer capacitor. As you bite down the metal sheets are bent together reducing the distance between them and the plastic sheet, changing the capacitance. It is SO incredibly simple I'm shocked to see it on a commercial product. The old saying "The simple solution is often the best" was definately in mind when they made this.

The pressure sensor is what is called a single flow backpressure system. at the mouthpiece airflow is split into two tubes. One tube is a "closed" system - air does not flow through the tube, it only detects increases in air pressure. This allows a Motorola MPX02010GS 1.45PSI Max transducer to be used as the pressure detector without major concerns about moisture. The second tube just runs straight to the bottom of the instrument and out. This is where all air breathed out goes. This dual-tube single-flow system is also a very elegant and simple solution to the huge problem of measuring breath pressure without shorting out your sensor or other componants, and it also reduces the amount of pressure that can be applied to the sensor preventing overload.

There are two simple strain guage "bend" sensors located around the right thumb. This is one of the few spots I disagree with their design as the lack of a thumb rest makes the instrument very difficult to hold on to, resulting in my constantly hitting the top one. While I'm slowly breaking myself of that habit and adjusting to resting all of the weight on the neckstrap (Hell, I can play Bari Sax without a neckstrap and without resting it on the floor!) the placement of these sensors is semi-awkward - though I've found that mapping them to "CC21 Growl" and "CC23 Flat Roll" means accidental contact is not the end of the world, and in fact can increase the expressiveness of playback with software like Mr. Sax T which recognizes them.

Like most cheap woodwinds, the neckstrap it comes with is pretty much rubbish... Not exactly a shocker since if you went out and bought a $300 el-cheapo made in China Cecilio Tenor Sax you'd get almost the same quality... Actually, that's not fair, the cheap chinese neckstrap would at least have a "recurve" on the hook to reduce the odds of it flopping off accidentally and be more substantial than nylon spagetti straps. The large open hook falls off the unit too easily, annoying since the lack of a right hand thumbrest turns C# into a case of the dropsies unless you constantly apply strain on the neckstrap. First good purchase for one of these is a soprano sax neckstrap that can be tightened down as short or shorter than your normal bolo tie. It's one of those little details that if they had put more than 50 cents into, it would make the unit feel more refined. Thankfully the hoop to which you connect the neckstrap is a nice big robust chunk of nickle plated steel meaning most any lanyard hook will work with it.

Internally there are three small steel weights likely there to make the EWI feel less like a toy. From a balance standpoint I found them to actually be more hindrance than help increasing the difficulty in not constantly dropping the unit - so I removed them from mine.

All the concerns of holding onto it can probably be dismissed by most people as much of it stems from my early saxophone instruction in the late 70's and early 80's, where I was taught the thumbrest is for the weight, the neckstrap is there for if you **** up. (like most New Englanders, my instructor's use of language was colorful at best, obscene at worst) The EWI seems to have been designed from the start with the opposite philosophy. The key to holding it seems to be to create a pressure triangle between your right thumb and the neckstrap connection - resulting in more pressure on the neckstrap than I've dealt with playing Baritone Sax. Despite it being a total featherweight, playing it without a neckstrap simply isn't viable - since resting it you would also stress the cord and plug the air hole. I consider that a design flaw, YMMV.

The parts used in it's manufacture are extremely solid and well made, and I can honestly say that in terms of design and materiels used no corners were cut. One of the many jobs I've held over the past few decades was in plastics manufacturing and design, and the extruded polystryrene used in the outer case is a rather robust thickness, somewhere over an eigth of an inch where most companies these days would use half that. The black thermoset resin casings for the pitch benders are the only part that are a bit thin, but they are sufficiently re-inforced from underneath it is not a concern. Internally they also used thermoset resin for the junction between the silicon mouthpiece and the air tubes. This choice was likely because of thermoset's resiliance to liquids and stability over time. Where most plastics in contact with moisture give up their oils becoming brittle, thermoset remains fairly stable. The only drawbacks of thermoset are the lower melting point and that it is much more brittle than PET or extruded polystyrene. Both are non-issues in this design as you would never reach temperatures where you would worry about the former, and the mouthpiece junction block is protected by both the black poly shell around the silicone mouthpiece, and the case itself... Though I would take extra care not to damage that piece when/if you take the mouthpiece off for cleaning - though at least they made the "shell" of the mouthpiece out of PET (Polyethylene terephthalate) and the part you actually put your lips on out of silicone - both materiels being washer safe.

The use of silicone for the mouthpiece is an interesting choice. I would have dismissed it as a option because silicone is gas permeable, but at the pressures we are talking that is not really an issue and it is watertight. (Waterproof but gas permeable is just one of the oddities of silicone) Within the past decade there was an innovation in silicone molding that moved it out of the plumbers box and into the manufacture of goods like kitchen untensils and baby pacificirs. Everything that makes it ideal for home goods makes it an ideal flexible mouthpiece materiel. Tear resistant, wear resistant - if it's durable enough to replace latex in baby pacifiers being effectively bite-proof, it's more than durable enough for use in a mouthpiece. The EWI 3000 introduced the silicone mouthpiece, but it wasn't "perfected" IMHO until the EWI 4000s. The texture difference between the black mouthpiece of the EWI 3000 and the white one on the 4000s & EWI USB is noticeable to some, the newer version being softer and more resistant to wear. For all intents and purposes this part of the EWI USB is identical to that of the 4000s - likely on the philosophy of "if it ain't broke". I do wish the end was a little narrower, but given the choice of materiels used in it's construction going thinner would likely make it prone to mechanical failure akin to the piss-poor design of SATA connectors. Given the choice between having it tear to shreds in a year or having it last several years, I'll take the extra width between the lips... uhm...

The silver finish paint on the lower half is remeniscent of early 80's computers for me, but the finish seems durable enough. Stranger is that the black top piece and the mouthpiece shell appear to both have been painted with a metallic sparkle black - I've already worn off that sparkle paint in spots on the PET shell in just a month of ownership - though you really would have to look close to see it.

The only structural weak spot I can find in the design is one common to most electrical appliances, the plug. Thankfully they went with a type B, which is the strongest USB Plug in terms of transfer of force from cable to unit which is why it's common on devices prone to vibration - like printers. Regrettably internally the only thing the USB plug itself is connected to is the circuit board via solder. Much like power plugs on laptops this is going to be the most common point of failure for older units as the stress of moving the device while it is plugged in will be translated to that board weakening the solder until it lets go. The little rubber clip to hold the cable should alleviate a lot of those stresses, and that they recessed it into the chassis instead of having it just sticking out loose also increases durability - Yet it is still a likely failure point. On mine I put a coat of two part liquid epoxy around the connector to the board to spread out those stresses, then also put some between the board and connector and the bottom shell to transfer energy/stress that direction - AVOIDING putting any on the solder side of the board should I have to go in there and re-solder in a few years. It's an easy bit of extra protection that again, I'm suprised they didn't bother with at the factory - unless of course they are looking for it to fail so they can charge for repairs in a few years... or lowering the price point to where it is cheaper to just buy a new one.

The main circuit board is mounted in an odd fashion - it is basically held in ENTIRELY by the boards around it's plugs. The plug from the top half of the case, the plug to the USB connector board, the plug to the grounding plate and pitch benders, and the plug to the board that has the rollers all intersect in a manner that holds the main logic board in place. It's kind of an odd design to not see any form of screws or reinforcement apart from a few plastic standoffs to hold it... But it probably simplified assembly and it does make replacing that one board if need be a lot simpler. Given that there are sixteen screws holding it together on the outside, I wouldn't worry about it, though my experience with those types of connectors are that they can tarnish after a few years - see the video connector on certain older laptops. Sometime in the future we'll likely be seeing people advocating the use of watered down tarn-x on them to get them working again... When a better answer would be some WD-40 on the end of a Q-Tip.

Still for the handful of faults, it has all the strengths of the classic EWI at a price finally within reach of the massses. I'm a New Englander, we are not easily impressed, in the habit of finding fault with everything and ripping even the best products to shreds with complaints. Seriously if the handful of minor issues mentioned above is the best I can come up with for complaints, they did a damned good job. Now if they just had more to offer for documentation than a handful of pamphlets...

Hardware Specifications

  • Dimensions:23.125" W x 2.75" D x 2.75" H
  • Weight:1.3 lbs
  • Power:~100mA, 5V via USB
  • MIDI output channels over USB:1
  • Note keys:13
  • Octave rollers:2 mobile, 2 fixed; 5-octave range
  • Plates:2 ground, 2 bend
  • Terminals:Slave connector x 1 (MIDI over USB)
  • Accessories:USB cable (3m)
    Neckstrap
    Cleaning cloth
    Software DVD
    Quickstart Manual