Roland JUNO-60 Chorus clone?

[VintageCharlie]

Hi folks,

i would be interested in a clone of a JUNO-60 chorus, as close as possible to the original design. With the same features, the ability to combine preset I + II and with the same bass boost / thick phasey stuff going on. I really like the TAL chorus VST plugin, which is an emulation of a JUNO-60 and would love to recreate that sound in an analog standalone box.

I stumbled over this project: http://gliglisynth.blogspot.com/2014/05/new-diy-project-standalone-juno-60.html

Looks like it is exactly what i am after, but i am not sure if this design reaches the same sound. I would like to contact the author, but could not find any contact info. In case GliGli is a member here, please let me know. It would be interesting to compare his clone to a TAL chorus - if it nails that sound, i am sold and will build this ASAP.

Also, i read some info on the DC-2 and it's ability to produce TZF just by a few minor mods - is there any chance a similar addition could be made to the JUNO chorus, as, from what i gather, they are supposed to be pretty similar designs.

Any advice is greatly appreciated!

VintageCharlie

[Mark Hammer]

1) Good choice.  Difficult choice, but good choice.

2) The Juno uses a pair of MN3009 chips, which are 256-stage BBDs.  The only MN3009 chips I have ever set eyes on are the two in my non-working Juno 106 in my basement.  Apart from that, I have never seen them or the 3209, for sale anywhere.  So, good luck with that, as they say.

You could use a pair of 1024-stage chips and simply crank the clock frequency up to get the same delay time, but you would likely need to use some sort of buffering between the clock and BBD to accommodate that ratcheting up of clock frequency.  The good news would be that, with a much higher clock frequency, the lowpass filtering on the BBD would not need to be nearly as aggressive.  But it would be a more complicated build.

3) There does not appear to be any difference between the two distinct delay paths used in the Juno Chorus, however they are counterswept by a common LFO, as in the DC-2.   In the DC-2, when the sample rate of one BBD gets faster and makes the output sharp, relative to the input, the sample rate in the other BBD is reduced, making the output flat, relative to the input.  In some respects, if you can identify the relevant timing/clock caps in the Behringer DC-2 clone, AND you have soldering skills to be able to get in there and swap them for a smaller value to shorten the delay time a bit, you may have what you seek.

4) The Line 6 Tonecore Space Chorus has an emulation of the multi-BBD chorus, as does the M5/M9/M13, and I suspect the MM4 as well.

[VintageCharlie]

Mark, thank you very much for the extensive and insightful reply!

Not to awake any false impressions - i am a noob at electronics, so i could just about pull of a monkey see, monkey do kind of project, so i am very happy that someone has provided schematics, layouts, etc. for such a build.

It would be ideal if it could be possible to have the following control scheme for the standalone JUNO-60 chorus. The original I, II and I + II presets. Then, if it would be doable without too complex circuitry, have a switch that switches between the original presets and speed/depth controls that let the user adjust the effect to taste, hence one would have kind of another preset that can be customized.

One more thing, would it be difficult to make the bass boost (which is essentially the result of the LPF, if i understand it correctly) switchable? I like the bassiness and fullness of the original JUNO chorus sound very much, but i can imagine that there are situations, where the extra bass might be an issue and hence it would be great if it could be dialed out with a simple switch for a more linear frequency response.

[Mark Hammer]

I'm looking at the Juno 60 and Juno 106 circuits, and unfortunately, they are a little too cryptic to be able to figure out what modes I and II do in an electronic sense.  However, given that there is no provision for asjusting rate and sweep width, I suspect they simply involve the CMOS switch shown in the schematic in orderto provide two different rate/depth combinations.  This is pretty much what the Boss DC-2 and the Behringer clone of it do, albeit with 4 combinations, instead of merely 2.  If that IS the case, that simply strengthens my recommendation to score yourself a Behringer unit,  For what they cost, you will likely be unable to even source the parts for your own build, so skipping to the happy ending is recommended.

Although I can't see it on the schematic either, I suspect that the "bass boost" is really nothing of the sort, but is actually a bass-cut switch that one either engages, for a swirl limited to mids and highs, or disengages to apply the chorus to the entire spectrum.  Again, if one could identify the relevant caps on the Behringer unit, it would be a moderately simply job to implement that....barring unreasonably small/fragile objects/places to unsolder.

FWIW, I generally include such a bass-cut control on choruses I make for myself, and it helps a lot to "de-wobble" the chorus and make it more suitable for bass guitar.  You will note that the Boss HF-2 Hi-Band Flanger is aimed at bass players, and uses both a BBD with fewer stages, to reduce the delay time range, and a little bit of high-pass filtering to keep the flanging effect out of the deep bass range.  Again, the goal is to keep the swirl, but lose the wobble.

As a complete aside, I don't know that I've ever seen a phaser analogue of the DC-2.  That is a phaser with two counterswept saets of phase-shift stages to produce an effect that doesn't go up or down in an obvious way , but feels like it goes everywhere at once, as the DC-2 does.

[VintageCharlie]

Actually i do have the DC-2 (got one for a great price a year ago) and the Behringer CC300. The Behringer sounds really nice on it's own, but to my ear it has more of that typical chorus pitch warble that i don't hear with the JUNO or the DC-2. The DC-2 sounds similar to the JUNO, but, as you say, has even less of the "movement" and also no phasey stuff (which i like a lot in the JUNO). I imagine one could get closer to the JUNO with an EQ that emulates the slight bass boost that corresponds to the EQ curve of the JUNO, but i haven't tried that out yet. The DC-2 also does not allow a combination of the presets. Seems like Nr. 3 and Nr. 4 correspond to the JUNO Nr. 1 and Nr. 2 quite closely. While the cc300 allows combinations, it just gets more and more warble, which, again, sounds great, but is not the unique sound of the JUNO. I'm not a big chorus fan, but to my tastes/ears the JUNO is in a league of it's own - subtle, but very apparent and musical (the thickness, phasiness and broad sound scape) at the same time, not cheesy or cheap like some choruses tend to sound to my ears/brain.

The guy who built the JUNO clone said he used chinese MN3009's (no idea if they are accurate reproductions or merely an approximation), but the result seems to sound very close to what the JUNO sounds like. Hence the parts availability might not be so dramatic after all?

[Mark Hammer]

If he can source 3009s, more power to him.  Myself, I've never found a source for the 3009 or the 3214.  They have datasheets, so I assume they exist, but I've never seen a single product using the 3214 - a 1024-stage BBD with 5 taps at harmonically unrelated points (a sort of baby brother to the 3011).

Not sure what you are using them with, but the Juno Chorus will be heard with a synth input, that will have very wide bandwidth that naturally accentuates the richness of the effect; same way that flangers sound amazing applied to mixed-down multitrack recordings, but rather uninspiring on one's guitar.

In general, a big chunk of the "character" of different chorus pedals is the delay range they aim for.  Although it IS possible to get different-capacity chips to overlap in delay range, a 256-stager will have only one quarter the delay time of a 1024-stager, clocked at the same frequency.  Hence be in closer to that border between flanging and chorus, where the swirl is more pronounced.

[YouAre]

Well, looks like I'm doing this project too!

Mark, based on page 7 of this document, we've got up to 20VPP coming from the LFO? It seems like the 3101's are getting +/-15v between the 2 of them, so 20VPP would make sure. I'm not too sure of that though.
http://www.houseofsynth.com/hos-downloads/manuals/Roland/Roland-Service-Manuals/Roland-Juno-60-Service-Notes-Hi-Res-Version.pdf

Also, would you care to comment on what's going on at the 3101 input (page11)? I'm familiar with the CE-2 and understand that the Juno Chorus is soooomewhat similar to the DC-2, but this setup is totally new to me. 5 transistors between the LFO output and this? Madness.

Thank you for the help on my super vague questions...

Regards,
Murad

[Mark Hammer]

I think (but will not swear to it, since I am looking at bits and pieces at a time on my tablet) that the 20VPP is unlikely since many of the electrolytic caps in the chorus board are shown as having a 16V rating.  Of course, given how sprawling the service diagram is, not to mention the fact that it is not all on one page, it's all too easy to get lost or overlook a point where voltages are dropped down from X to Y.

[YouAre]

I think (but will not swear to it, since I am looking at bits and pieces at a time on my tablet) that the 20VPP is unlikely since many of the electrolytic caps in the chorus board are shown as having a 16V rating.  Of course, given how sprawling the service diagram is, not to mention the fact that it is not all on one page, it's all too easy to get lost or overlook a point where voltages are dropped down from X to Y.

The "10/16" designation next to the caps in the clock driver selection would definitely suggest that.

Any thoughts on what's going on between the LFO output and the MN3101? Still not sure what all that circuitry is doing to the clock driver.

[Mark Hammer]

Well, the diagram shows each clock section fed an antiphase version of the LFO output, so it is more or less a variant of the DC-2.  My guess is that the additional circuitry simply shapes the LFO waveform to produce a particular kind of sweep.

[anotherjim]

In The Juno 6 manual ( which might be easier to read), the Chorus LFO output is given as...

I = 20Vpp @ 0.4Hz Triangle
II = 20Vpp @ 0.6Hz Triangle
I+II = 2.6Vpp @ 8Hz Sine (low pass filtered triangle)

So I & II are Chorale triangle sweeps and I+II is a shallower Vibrato sweep.
These 3 LFO types are the only differences between the modes I can see.

[YouAre]

Well, the diagram shows each clock section fed an antiphase version of the LFO output, so it is more or less a variant of the DC-2.  My guess is that the additional circuitry simply shapes the LFO waveform to produce a particular kind of sweep.

.

Because of my inability to get an inverter in Spice (to get the MN3101 simulated), I can't sim what's going on with all that additional circuitry. Even if I were to build it and look at it on a scope, I honestly have no idea what I'd be looking at when I scope the output of the clock.

Can anyone else comment on what that circuitry is doing to the LFO before it hits the clock?

[anotherjim]

I'm still looking at the Juno 6 scheme (can't get a Juno 60 one to load!)
I can see a power on reset that mutes the BBD output and holds the clock rate (to stop garbage spewing out until the bias has got to all the BBD cells). The rest looks like a linear-exponential converter(?) (similar to synth type CV control) that in turn controls clock frequency.

[Mark Hammer]

In The Juno 6 manual ( which might be easier to read), the Chorus LFO output is given as...

I = 20Vpp @ 0.4Hz Triangle
II = 20Vpp @ 0.6Hz Triangle
I+II = 2.6Vpp @ 8Hz Sine (low pass filtered triangle)

So I & II are Chorale triangle sweeps and I+II is a shallower Vibrato sweep.
These 3 LFO types are the only differences between the modes I can see.

I can see in the Juno-6 service manual those voltages, and that pressing the I+II brings in additional resistors to divide the voltage down.  In the Juno 60, the arrangement is sort of the same, except that a 4066 is used for the attenuation, instead of physical switches.  As well, the service manual says the waveforms are 20, 20, and 1.6Vp-p rather than 20, 20, and 2.6.

[iiimonfire]

I'm wondering if anybody (@VintageCharlie, @YouAre) has gotten this project to work?

I've been able to source some MN3009s (reissues from XVive, and NOS from Banzai) and have sourced some MN3101's from AliExpress (I know, I know...). For the most part, my voltages seem correct, but I am not getting any sound output at pins 7 or 8 of the MN3009s. Am I to assume that I have bad clock drivers?

The only difference between my circuit and the GliGli Synth schematic is that I've added a power section comprised of an LT1054 to step up from +9V to +-18V, which feeds into the two necessary voltage regulators (7815 and 7915) for +-15V regulated into the rest of the circuit.

[anotherjim]

Since this, TC electronics have a clone in pedal format - June-60 or something. It appears though that they have used 1024 stage BBDs and altered the modulation to better suit guitar. Also no "stereo".
Reviews are not very positive. SMD build, of course, so a good hacking challenge for somebody I'm sure. It is pretty cheap (TC are Behringer).

Roland picked preset LFO speeds and depth for sweet spots on typical synth tones - and they don't necessarily suit guitar that well either. The extra "easter egg" setting when forcing both modes I and II radio buttons to latch down is in fact supported by the circuitry. Modulation is then too deep for a synth but might be perfect on guitar.

[VintageCharlie]

At some point i heard about. the upcoming TC pedal and waited for it to come out, purchased it, tried it and put back into a drawer. It sounds nothing like the original effect. And the original sounded great for guitar - no idea, what they wanted to "improve". It has neither the same almost through zero like modulation, it is much more standard chorusy sound. And it has none of that great fattening effect of the original. It would be an ok effect, if you would not measure it by the Juno chorus.

[anotherjim]

If all TC did was to change to a longer BBD, then it wouldn't have a cat in hells chance of sounding like the original. It has a fixed wet/dry mix so the minimum delay time will matter. They would need to increase the BBD clock speed to the maximum to get closer to the shorter delay of the original. The chips can be clocked faster than the spec says, but I bet they didn't try this.