Carrier bleed-through / Oscillator "mute"

[John Lyons]

I'm getting some bleed through from the oscillator.
Is there a way I can mute the oscillator?
I have it hooked up with a carrier null but I can't
quite kill it all.

[swamphorn]

The FSK input? It switches the role of timing resistor between R₁ and R₂, so if you only have one resistor the input serves as an active low or high oscillator brake.

[ElectricDruid]

Good idea.

[John Lyons]

Here is the actual schematic I'm using.

[Eb7+9]

I'm getting some bleed through from the oscillator.
Is there a way I can mute the oscillator?
I have it hooked up with a carrier null but I can't
quite kill it all.

best-case carrier suppression is 55db with this ic ...

[John Lyons]

Is there a way I break a connection to stop the oscillator?

[Mark Hammer]

When I made an EPFM ring modulator a couple years ago, I couldn't null the modulating oscillator enough to arrive at silence when I wasn't playing notes.  Because the circuit feeds the RM signal to an op-amp mixing stage, I simply inserted a back-to-back diode pair in series with the signal to act as a quick-and-dirty gate.  I used Schottky diodes, such that the whine had to exceed about +/-200mv to become audible.  Of course that will introduce some crossover distortion, but it's not as if a ring modulator is ever going to sound especially pristine anyway.  The diodes will subtract from the amplitude of the final result, but then you can always compensate by increasing the value of R7 to nudge the gain a little higher.

[nocentelli]

the parasit studio sonic reducer (v1 and 2) has a neat envelope detector gate thing on the oscillator - it's a single opamp job rather than the function generator thing, not sure if the principle could be applied

[Eb7+9]

Is there a way I break a connection to stop the oscillator?

try using the frequency-shift keying function to select between R1 (your choice) and R2=open
R2=open will obviously kill the CCO

[ElectricDruid]

try using the frequency-shift keying function to select between R1 (your choice) and R2=open
R2=open will obviously kill the CCO

That's exactly what Swamphorn said. It's *still* a good idea!

What we need is a signal detector circuit that can switch that FSK input.

TBH, Mark's "two diodes" solution is much simpler and probably just as effective.
 

[Rob Strand]

If you look at the Maestro RM-1A and RM-1B you can see it's an old problem.  They use a JFET based noise-gate.

I guess the question is: Is the excess carrier simply the -55dB feedthrough?, or, it is caused by layout or other unforeseen circuit issue?

[John Lyons]

Rob
Yes, it's bleed through via crosstalk as it comes
though faintly while in bypass.
I was hoping to ground out the oscillator
while in bypass which I can jump off the
footswitch but so far I have mor found the way
to do that. Can you see a way to do it?
If I break the connection to the 68K as
the speed control I can kill the oscillator
but that would mean using a 4PDT switch.
(Bypass, LED and Make/break 68K)

Love the Maestro/Oberheim Ring mod.
If that dang Reticon SAD IC was "Available" I'd make more
of them.

[Rob Strand]

Rob
Yes, it's bleed through via crosstalk as it comes
though faintly while in bypass.
I was hoping to ground out the oscillator
while in bypass which I can jump off the
footswitch but so far I have mor found the way
to do that. Can you see a way to do it?
If I break the connection to the 68K as
the speed control I can kill the oscillator
but that would mean using a 4PDT switch.
(Bypass, LED and Make/break 68K)

Love the Maestro/Oberheim Ring mod.
If that dang Reticon SAD IC was "Available" I'd make more
of them.

I'm not sure I get exactly when it occurs.

The -55dB figure applies to the oscillator breaking through in effects mode.

If it's true bypass the effect should be completely removed from the audio signal path.  So muting the oscillator shouldn't be necessary.  If you get the oscillator breaking through in bypass there has to be a problem with the grounds.   You might be able to put a large cap, say 220uF across the XR2206 power pins but that might just be hiding the true problem in the layout.

Another way the oscillator can get in is via the wiring to the frequency pot and the wiring to the cap select switch.   Those need to be kept well away from the audio wiring.

If you are getting oscillator breakthrough with the blend pot set to clean only I would be a little suspicious of connecting R8, the 56k from pin 3 of the XR2206 to VR.   You might need to connect that back to it's own VRef.  Again increasing C10, the 47uF VR cap, might help.

[John Lyons]

Cool, thanks for that checklist.
I'll run down those.

[duck_arse]

If I break the connection to the 68K as
the speed control I can kill the oscillator
but that would mean using a 4PDT switch.
(Bypass, LED and Make/break 68K)
 

no sir. dpdt, with millenium indicator. use the mill as a control to switch whatever you like. see R.G.'s mill and mute mod.