Help with true bypass in boss flanger

[Jrockpsycho]

Trying to true bypass my boss rbf-10. Would also like to remove the noise gate if possible. Any ideas are much appreciated. Thank you!

[ElectricDruid]

From that schematic, doing the true bypass doesn't look too bad - unless it's all tiny SMD parts.

Remove the FETs that connect to C and D, so Q2, Q3, Q7. Then add a jumper from D to S across Q3 and Q7. That should wire the pedal permanently On, irrespective of what the footswitch is doing. You can then wire a 3PDT true bypass switch around the whole board.

Where's this noise gate you mention? I'm not seeing anything on the schematic that looks like one, and none of the controls seem relevant either.

[garcho]

Is the noise gate envelope follower the "remote" section, and Q3/Q7? Maybe tap the output at R39 to bypass the gate? Not sure, for sure.

[ElectricDruid]

No, that "Remote" is for an external footswitch to flip the flip-flop and switch the pedal effect on/bypassed. Q18/Q19 are the flip-flop.

[Jrockpsycho]

Hey druid thanks! Ok so I understand which fets to remove, then you say to install a jumper from D to S? Where is S? You mentioned doing it "across" Q3 and Q7, what exactly does this mean? Thank you so much.

If anyone else has ideas about the noise gate let me know. 👍

[ElectricDruid]

Hey druid thanks! Ok so I understand which fets to remove, then you say to install a jumper from D to S? Where is S? You mentioned doing it "across" Q3 and Q7, what exactly does this mean? Thank you so much.

A FET has three pins; Drain, Source, and Gate - eg D, S, and G. "Gate" is the one like "Base" on an ordinary transistor, the one that switches it on or off. You need to connect the other two so it acts like it's always on. On the schematic, that's connecting them from left to right, horizontally. Of course, in reality I don't know how they're arranged, and you'd need to check the pinout for the device. The pins may not be in DGS order, could be DSG or GDS or whatever.

[antonis]

you'd need to check the pinout for the device.

Drain & Source are interchangeable for the vast majority of (symmetrical) JFEts..
(at least, for audio frequencies..)

So, OP just has to locate Gate pin..

https://instrumentationtools.com/topic/how-to-do-testing-of-junction-field-effect-transistors/

[garcho]

The block diagram makes it look like there's a gate after the delay's low pass. Just trying to help at least a little after making a complete fool of myself

So, more foolishness... what are IC11 and Q13 doing? And Q11. Roughly speaking: with audio signal IC11's output turns Q13 on, which turns Q11 off? No audio signal, Q13 is off, so Q11 is on, and shunts the signal to Vcc/2?

Also, what are Q5 and Q6 doing? Not trying to hijack the OP, just curious.

[Mark Hammer]

Yep, no noise gate.

[garcho]

Hahaha, I really won't ever learn.

[antonis]

shunts the signal to Vcc/2

Which Vcc/2, IMHO, is formed by a very bad design..

[garcho]

Can someone explain what's going on with IC11, Q13, and Q11? And Q5 and 6?

Which Vcc/2, IMHO, is formed by a very bad design..

Why do you think it's a bad design? Isn't it just two 100K resistors and an op amp buffer? It seems very standard.

[ElectricDruid]

The block diagram makes it look like there's a gate after the delay's low pass. Just trying to help at least a little after making a complete fool of myself

Not at all. You're the only one who's really digging into this.

So, more foolishness... what are IC11 and Q13 doing? And Q11. Roughly speaking: with audio signal IC11's output turns Q13 on, which turns Q11 off? No audio signal, Q13 is off, so Q11 is on, and shunts the signal to Vcc/2?

Yes, I think you might have found the missing noise gate.

As far as I can see, Q11 doesn't serve any purpose except to short the signal to the midpoint voltage, like you said, which would presumably mute it. There's a kill switch thread elsewhere on this site at the moment which does something pretty similar. That would mean that the mess around IC11 and Q13 must somehow create some sort of envelope follower signal (again, like you suggested), although it doesn't look like any version of that I've seen thus far. That's a good thing! Always up for learning a new way to do old stuff! I might have to sim it to see what it's up to.

I didn't *see* it because I didn't *recognise* it. But you came at it with fresher eyes and spotted what I'd missed, so fair play for that. Respect.

Also, what are Q5 and Q6 doing? Not trying to hijack the OP, just curious.

Q5 and Q6 are much simpler. They just short out resistors in the output op-amps, altering the feedback resistance and changing the gain. That's why they're controlled by the "A" signal, which comes from the Level switch. Basically, without them switched on, there's a gain of three (R35 or R42/47K vs 100K+47K) and with them switched on there's unity gain (47K vs 47K).

[antonis]

Why do you think it's a bad design? Isn't it just two 100K resistors and an op amp buffer? It seems very standard.

An opamp buffer directly driving a 47μF capacitive load, not only isn't standard but it's highly indictable..

A series resistor (between op-amp output and C48) should be placed..
https://www.analog.com/en/analog-dialogue/articles/techniques-to-avoid-instability-capacitive-loading.html

or better, C48 should be removed..
(c52 is quite adequate for Vcc/2 stability..)

Dunno if RFB-10 was designed to be powered of battery so I don't make a comment about R66/R65 values..

[garcho]

Not at all. You're the only one who's really digging into this.

I'm not trying to be a jag, I thought the OP was interested, it was mentioned twice.

EDIT: maybe you were being nice and talking about me not being a fool, ha! Sadly, not true. 

Would also like to remove the noise gate if possible

If anyone else has ideas about the noise gate let me know. 👍

It's a bit like a comparator as a one-shot envelope trigger/closer, yeah? As opposed to an actual envelope follower. Since it's just supposed to cut off a little whine/hiss/hum I suppose that makes sense enough. Maybe just remove Q11?
Why do you want to remove the gate "feature"?

[Mark Hammer]

The IC11/Q11 combo makes no sense to me.  First off, this is Boss.  They have all the FETs one could want, if only left over from the true noise gate found in the CE-1, and yet Q11 is a bipolar transistor.  They certainly use FETs here for mode switching and bypass,  Second, IC1B and IC2A use the exact same pre-emphasis/de-emphasis network that Boss used for noise control in many of their BBD-based pedals.  So why supplement with a gate?  Alternatively, why NOT supplement in their other BBD pedals if it's that easy and useful?  Third, there is nothing in that particular subcircuit that would suggest a usable time constant.  Gates can certainly respond instantly to a supra-threshold signal, but they need to fade out somewhat gradually, or else the audible result is choppy and sputtery.  Does R83/C22 do that?  I don't know, but I doubt it.

Now, it would certainly not be the first factory schematic ever produced that had an error or two in it, although Boss circuit diagrams are generally far more accurate than, say, Maestro's.

But I support/second garcho's suggestion.  Lift one end of Q11, and listen closely to whether it makes a difference in noise level.  I'm not hopeful, but I've been plenty wrong before.

'll just close by saying that I'm envious.  I've got flangers up the wazoo here, so I shouldn't be coveting anything, but that half rack series was very nicely done.  I have the phaser and the 5-different-overdrives units.  The M-Bus provides some interesting opportunities, including adding actual envelope-following.  I added that to my phaser.  The same bus would also permit use of other modulation sources, like S&H, random, stepped, or simply a second LFO mixed in.

ADDENDUM:  Proof positive that not only have I been wrong before, but I'm wrong again.  This is from the RBF-10 service manual.

[Jrockpsycho]

"Dunno if RFB-10 was designed to be powered of battery so I don't make a comment about R66/R65 values.. "

No battery, only 9v wall power

"Why do you want to remove the gate "feature"?"

I will be using two of these in stereo on the entire mix of a track, and don't want the track gating during the fadeout

[Jrockpsycho]

Which makes me think... How difficult would adding stereo input be? It already has stereo output...

[garcho]

don't want the track gating during the fadeout

I know that feeling.

How difficult would adding stereo input be? It already has stereo output...

Stereo is just two parallel mono signals. Will the flanger be doing anything different to either channel? If not, then there's no point. If yes, then you build two flangers in the same enclosure. If you want stereo inputs, say, because you don't want to break a stereo signal chain, then I would just sum left and right into mono and be happy with the two outputs already there.

[Nitefly182]

"Dunno if RFB-10 was designed to be powered of battery so I don't make a comment about R66/R65 values.. "

No battery, only 9v wall power

"Why do you want to remove the gate "feature"?"

I will be using two of these in stereo on the entire mix of a track, and don't want the track gating during the fadeout

Flange before adding the volume automation?