Lucky find that was passed on to me from a retiring brass orchestra musician!
Right proper acid synth, eh?
Dual VCO-based synth with 4 pole low pass and analogue VCA
As far as I know, the EWI3000 was released before the rack mount EWI3020M, which offers fewer synthesiser elements (one instead of two fully voices) but more modulation options, including some rudimentary modulation “matrix” (Well, you can use the LFO not only for vibrato like on the EWI3000). Unlike the 3020M (and the EWV2000, for that matter), the 3000 also misses the switchable bandpass filter network at the end of the signal chain that can make your sound go “reedy” or “flutey.” But it sports midi, takes far less space than the rack mount version, and has two full synth voices that can be layered or used for chords.
Manual: https://www.patchmanmusic.com/manuals/EWI3000mOperatorsManual.pdf
What’s in a voice?
The sweet thing about the 3000 is that you get a lot of Curtis Analogue Synth goodness with it. Each of the two voices sports a CEM3374 dual VCO, which you can also find on Oberheim’s Matrix12/Xpander, Rhodes Chroma, and the Akai VX600, amongst others. For the MIX/VCF/VCA section you get an CEM3378, which is equally iconic. The VCOS offer Trio, Saw and Pulse Wave and you can hard sync VCO1 to VCO2. There is a modulation VCA for the VCOs that allows frequency modulation of VCO1 by the TRI wave of VCO2 *or* the filter envelope (which is cool for hard sync sounds). Technically, the setup is pretty much like on the EWI3020M or the VX600 but these can have VCO2 frequency modulate the filter cutoff for FFM, while ADSR-to-VCO-pitch is done with a mod matrix setting. What the EWI3000 is also missing is pulse width modulation, unless you’re using the breath controller (which can be set to a midi CC such as mod wheel or aftertouch as well!). The waveforms per VCO can only be selected (not mixed), and an external input signal can be used in their place. Both VCOs can be balanced (you can blend between them but not have both at, say 80%, which makes overloading the filter impossible), and the 4 pole low pass retains passband level at higher resonance.
Afaik the 3378, like the 3379, uses a variant of the CEM3328 (which, for instance, is used run Doppler’s take on the Xpander filter). While on the EWI3020M, the VCA of the the 3378 is used for volume control while the final VCA is built around a Mitsubishi M5206P, the EWI300 has the same VCA order but I still need to measure if they are doing the same thing. I hope the CEM does volume control because I think the Mitsubishi sounds better with ADSR control.
The modulation elements are all generated digitally. There’s an LFO, solely for vibrato, and two well scaled ADSR envelopes. These don’t have the fastest impact but have a good feel IMO. Apart from that, you get to route breath (either from the wind controller or per midi) to various synth parameters, including cutoff, resonance, pulse width, pitch etc. While I really would like to have one or two dedicated LFOs for these purposes, you can make do with midi automation or using aftertouch, for instance. Finally, you can set the volume of your “Source 1” and “Source 2”, which is what each synth voice is called here, for mixing and layering. While this seems luxurious (which vintage CEM synth isn’t anyways?), you can use one synth “source” for sculpting a transient and other halfway flexible things.
Modifications
Apparently, there were some modified EWI300Ms, mostly sporting CV outputs, so you could use the breath controller with a modular synthesizer. Nyle Steiner offered some mods, and AFAIK Matt from Patchman Music did so as well. My interest would be the other way round, though, and have the synth module controlled per CV and gate inputs. Making the VCOs and the filter cutoff respond to 1v/oct signals is quite easy (I did this on my EWI3020M, which I still need to document, ahem…), and tinkering CV inputs for PWM and VCA control would be easy also.
At the moment I’m still reluctant to drill new holes into the plastic case, mostly because I already have a EWI3020M, which I partly butchered (it was not in the best state when I got it anyways), so any more extreme incursions are on hold for now. Things I’d do, however, up the signal level, so the VCA can be driven a bit, PWM CV inputs, Filter FM options, extended signal mixing options.
Some sources
https://www.patchmanmusic.com/NyleEVIConversion.html
https://www.matrixsynth.com/2009/03/akai-ewi3000-with-modded-3000m-and.html
https://www.ibiblio.org/pub/multimedia/emusic/info-docs-FAQs/wind-controllers-FAQ.html#3.4
https://modwiggler.com/forum/viewtopic.php?t=150391
IC INFO
The only source for a service manual seems to be this here: https://www.vstservice.com/p-663-akai-ewi3000-service-manual-ewi-3000-ewi-3000.aspx Though, if you search for the AKAI VX600 service manual (synthxl website), you get a clear idea of some of the internals.
http://www.synfo.nl/datasheets/CEM3374.pdf
https://github.com/alanbog/3374-VCO
http://www.synfo.nl/datasheets/CEM3328.pdf
http://www.buchty.net/ensoniq/files/cem3379.pdf
http://datasheet.buhieen.net/M5206P.pdf
