Etherwave part-swapping and upgrade discussion

The purpose of this thread is to discuss where there might be room for slight improvements in the playability, octave range, voicing or oscillator stability in the Moog Etherwave by means of parts-swapping or other slight modifications. 

There’s a bevy of posts across this forum from people trying their hand at making a Etherwave from scratch. Contributors experienced in Etherwave maintenance and repair will usually step in to assist these people with parts substitution, tuning procedures or build troubleshooting. On some of these occasions, a part or component is suggested that is of a higher specification than would be found on an Etherwave as purchased from Moog or its dealers. Others have gone as far as to produce aftermarket modules which, after installation, improve the overall usability and sound quality of the Etherwave. The best known of these modules, the ESPE01, has been used to increase the octave range, linearity and timbre of the Etherwave. Below I will collect other parts substitutions where it has been noted as an improvement on the stock component.

Antenna Linearization Coils

L1, L2, L3 - 10mH

L7, L8 - 2.5mH

L9 - 5mH

Thierry has suggested that, from his experience, replacing the J.W. Miller/Bourns 3-pi coils with 3-pi coils from Hammond will result in a more linear pitch and volume field. This is despite the respective datasheets indicating the coils are identical.

Variable & Fixed Pitch Oscillators

C1, C5 - 3300pF

These capacitors are the principal frequency determining components in both pitch oscillator circuits. The stock Etherwave uses either polypropylene or polyester (Mylar) 50V capacitors in this position. Polystyrene capacitors might seem tempting in this position for a scratch build but Thierry advises against it and prefers Silver Mica instead. Not sure if Silver Mica is superior to polypropylene in this regard.

L12, L13 - 47uH

These fixed inductors are in series with variable inductors L5 & L6 (47uH), respectively. L12 & L13 are ferrite-cored inductors with a DCR of about 3 Ohms. Thierry has suggested EPCOS B82144A2473J as an substitute. Without knowing the part number of the stock component, it’s hard to determine if the Epcos part has a superior temperature coefficient. The lower DCR of the Epcos part is not a factor.

Note: “The small 47uH and 22uH fixed inductors are not critical, but for temperature stability. Some experimenting with different types will lead to big surprises.“

I wonder what Thierry was alluding to...

C4, C8 - 1uF

These are electrolytic bypass capacitors. Unless the ESR is beneficial to the oscillators, it may be worthwhole to replace these with readily-available radial polypropylene capacitors.

C2, C6 - 15pF

The stock parts are tantalum composition capacitors. Polystyrene capacitors here may prove to be a lower-distortion upgrade.

Volume Oscillator 

C14 - 1800pF

Thierry has suggested Silver Mica in this position as well. No idea if it is superior to polypropylene or Mylar in this position.

L14 - 22uH

This is in series with 47uH fixed inductor L11. EPCOS B82144A2223K  has been suggested as a substitute to the stock ferrite inductor.

Detector

D4 - 1N4148

This is the component that extracts the difference signal from the two pitch oscillator circuits. There’s a plethora of diode substitutions, results of which would vary.

I think that covers everything so far...

wow, finally someone made the effort and gathered all these widely spreaded informations and put in a proper list. thanks very much.

Thank you. It only took 40,000 posts but now it’s all in one place.

I’ll probably revise the post soon. I’m thinking of adding a link to the “Hotrodding” document along with parts revisions (omit L4 & L10, etc). Then again, as we discussed in the other thread, it may be irrelevant as scratch builds of the Etherwave are likely a thing of the past.

I imagine I’ll also revise it once Thierry chimes in as well. 

I have an idea for L12 & L13 that may be an improvement if I can get it to fit under the hood, so to speak.

40'000 posts under the sea sounds pretty deep.

yeah, the hotrodding manual. that could need a update too! 

what also should be considered is the pcb layout itself. thierry stated clearly the importance of how the traces are made. later i saw a nice documentary about bob moog on analog synths and there how he talks about "those traces on the board".

Thank you for sharing that. PCB spiral inductors come to mind: 

I don’t think the hotrodding manual is really of any use, besides the (outdated) schematic. It seems to me that if somebody were to simply revise and upload a current Etherwave schematic we’d be better served.

"The purpose of this thread is to discuss where there might be room for slight improvements in the playability, octave range, voicing or oscillator stability in the Moog Etherwave by means of parts-swapping or other slight modifications."  - BBrinkman

I don't want to discourage you, more power to this thread!  But I do think the mod you'll get the most bang out of is buffering to remove oscillator coupling (the overpriced ESPE01, or my DIY YAEWSBM).  The rest is mostly (IMO, forgive me) lipstick on a pig. Buffering is two stupid transistors (FETs, or garden variety small signal BJTs) in the simplest configuration possible (source or emitter follower, no voltage gain).  Not sure why it's not a part of the stock EW in the first place, as coupling can always be reintroduced if desired, and in a more controlled manner.  (FredM said BobM was a minimalist circuit designer almost to a fault.)

Thank you Dewster.

I’m not offended by the lipstick on a pig sentiment; my thoughts are to consider the Etherwave as a platform for experimentation and then trying to milk the best performance from it. So perhaps lipstick on a cow is more apt, idiomatically speaking.

here’s something I’m considering, I’m interested in your thoughts:

what if L12 and L13 were wound adjacent to each other on the same single-layer air core (bifilar) but wired with reverse phase to each other? Their temperature drift would track each other (sharing the same core) and their phase relationship would be constructive, as far as our purposes are concerned. 

"what if L12 and L13 were wound adjacent to each other on the same single-layer air core (bifilar) but wired with reverse phase to each other? Their temperature drift would track each other (sharing the same core) and their phase relationship would be constructive, as far as our purposes are concerned." - BBrinkman

I'd keep them physically apart to keep coupling to a minimum.  And if you go air core there won't be much in the way of temperature drift.

But wouldn’t the two coils reverse-orientation make the coupling constructive?

Hello,

i just finished a diy Etherwave with some components changed.

I started with the schematic shown in the hotrodding manual, but changed the following parts:

- L5 and L6 changed to Coilcraft 7M3-104, eliminated L12 and L13

- L11 changed to Coilcraft 7M3-683, eliminated L14

- R2, R6 and R16 changed from 2K2 to 1k5

- R15 and R19 changed from 470R to 330R

- C11 changed from 0.47u to 2,2u, and going to ground, not to B+

- U1 and U2 were changed to TO220/1A parts, because the little TO92 parts became too hot (my opinion)

- added 47u electrolytics behind the voltage regulators

- used 4 liearization coils for volume and pitch as shown

- added buffers using MMBF5484 Jfets

- used styroflex capacitors in the place of C1 and C5

- added a Output Hi/Lo switch, a 39K resistor in series with R33, wich can be shorted with a switch

- added a "continuous output" switch, a simple switch between VCA out and B+

After tuning the theremin, i was happy with the results.

I never touched a stock etherwave so i can compare the performance this build only to YouTube videos i saw.