i know there's been a million threads about the Ibanez/Boss flip flop circuits, but i can't find any addressing this particular op-amp circuit, and i've tried some of the main fixes for these issues to no avail, so thought i'd reach out before try replacing all the associated chips and transistors.

the schematic is in post 5 of this thread: https://www.freestompboxes.org/viewtopic.php?t=4179

there is some excellent reading on this circuit at http://www.geofex.com/Article_Folders/bosstech.pdf that really helped me understand how it worked and has been my main reference for troubleshooting.

issue is simple - the effect switching on and off only works intermittently, and favors the off position. all effects and functions work perfectly normally when the pedal is actually turned on. this is with a known good 9V power adapter, thought the issue is the same when using a battery.

the switch i assume is quite well used, and has accumulated the 'bounce' that everyone talks about, but i believe the issue lies elsewhere, as the switch makes perfect contact across itself every time it's actuated, and shorting directly across the switch with an alligator clip produced no better results than by using the switch. holding down the switch, or switching faster or slower, also produces no change in on/off success rate.

i tried replacing C47, the cap across the switch, with a 10nF up from a 1nF, as this seemed to work out for every single person i found who tried it... aaaaaaaand it somehow made it WORSE! now it is impossible or damn near impossible to 'blip' the flip flop into on mode, it stays off perpetually. it seems it wants to turn on because the LED gives the tiniest little flash, but it never latches. i have confirmed that the switch is indeed bringing the junction of C45 and C46 to ground, but it's still not flipping the flop!

measured every resistor in and around the circuit, all measure within a 1% tolerance. replaced C43-46 with new 1000p caps, no change. tried shorting the switch to ground through a 100R resistor like i see in some circuits, no change.

now here's the kicker - when i short pin 11 of IC6E directly to ground, it turns on EVERY time! same with pin 9, makes it turn off every time. and when it's on, the switch has no problem turning it off, every time. it just won't turn on with the switch. so at this point, shunting the junction of C45 and C46 to ground will never turn it on, but shunting pin 11 of IC6E to ground turns it on every time. and i've confirmed that the on and off states produce symmetrical voltages on the IC6 pins: the ON state gives me 0V on pin 8, 3.3V on pin 9, 5V on pin 10, and 0V on pin 11, and the OFF state is a mirror image of that.

i should mention that when i push the switch, there is a little blip on each of those IC pins of perhaps 0.1V, just not enough to actually flip the flop.



so....... could this possibly be an issue with anything but IC6? any other ideas?

Q4 seems to only exist for the LED circuit, but i was thinking it could perhaps be unbalancing the circuit if it's not working properly... i'm not really sure what proper readings would be, but when the circuit is ON i have 0.74V on the base, 0V on the emitter and 0.012V on the collector... and with it OFF there's 0V on the emitter and base, and 5.58V on the collector.

what would be some typical cold resistance readings for Q4 and IC6?

Quote from: mxrshiver on August 25, 2024, 05:22:03 PMnow here's the kicker - when i short pin 11 of IC6E directly to ground, it turns on EVERY time! same with pin 9, makes it turn off every time. and when it's on, the switch has no problem turning it off, every time. it just won't turn on with the switch. so at this point, shunting the junction of C45 and C46 to ground will never turn it on, but shunting pin 11 of IC6E to ground turns it on every time. and i've confirmed that the on and off states produce symmetrical voltages on the IC6 pins: the ON state gives me 0V on pin 8, 3.3V on pin 9, 5V on pin 10, and 0V on pin 11, and the OFF state is a mirror image of that.

To me that means the gates are working but the circuit is on the threshold of working.

As a sanity check, try soldering another switch across the existing switch pins.

At first try the original C47=1nF but I would try to get it going with C47=10nF, perhaps even try 47nF.

The operation of these circuits also depends on the C43/C44 and C45/C46 values.   I did some analysis a few years back.   Perhaps try *reducing* C45 and C46 to 470pF.

The presence of C49 and C53 raises my eyebrows a bit that they had problems with the reliability circuit.

You might want to checkout this thread out because it deals with a similar circuit:
https://www.diystompboxes.com/smfforum/index.php?topic=131880.0

Things get complicated.    You can see I used a slightly different circuit for the MOSFET case, the one with the diodes, because the Boss/Ibanez circuit doesn't work inherently well with MOSFETs.  Some of the issues probably extend to gates *however* at the end of the thread I mentioned MOSFET capacitances and the gates won't have that issue - so perhaps there's room to tweak that circuit.

QuoteOnce the feedback cap was chosen the trigger caps needed to be small to stop glitches on the falling edge, say less than 220pF.   To my surprise I could drop the trigger caps down as low as 22pF.  So the circuit worked with fairly large and small feedback to trigger cap ratios.  In the end I decided the feedback caps 470pF to 1nF and trigger caps 100pF.

The existing 1nF feedback caps (C43, C44) are probably a good start.  From this quote perhaps my earlier idea of using smaller trigger caps (C45 ,C46) is definitely something to try.

If that doesn't work  I'd be tempted to also raise or remove the 200k resistors (R49, R50).
[see below reducing the resistor helps.]

Given things can get tricky with the details the circuit might need a closer analysis with the gates.  The above are only casual remarks based on previous analysis of the transistor and MOSFET versions.

Other links:
https://www.diystompboxes.com/smfforum/index.php?topic=127813.msg1228447#msg1228447
https://www.diystompboxes.com/smfforum/index.php?topic=129239.0

I had a quick look at it.

The gate version of the flip-flop doesn't seem to work.  I'll need to play around with it.  It feels very much like the problems faced with the MOSFET based flip-flop.

I tried the "forced steering variant" I used on the MOSFETs and it worked first go.

No guarantees but here's a mod which might push things in the right direction:

Use the recommended cap values.
If 180k doesn't work try 150k.   If you like try 150k first then 180k if that doesn't work.
120k isn't recommended since it's too close to a non-working configuration.

FIXED.

C47 to 18nF, R49 and R50 to 180K, and, it seems most critically, C45 and C46 to 220pF.

progression of changes and results:
- C47 to 18uF, resulting in ON/OFF switching success rate changed from about 20-30% before, to 0% for ON, but 100% for OFF after
- R49 and R50 to 180K, C45 and C46 to 370pF, resulting in switching success rate of about 30-40% evenly balanced between ON and OFF
- R49 and R50 to 150K, resulting in switching success rate of about 20-30%
- C45 and C46 to 220pF, resulting in 100% perfect switching success

i was going to try 330pF for the trigger caps, but then i realized i'd seen most diagrams in articles about this using 220pF, and lo and behold, it was what i needed.

i don't have too much of an understanding of the roles of these particular components - despite how awesome that article i linked to is, he doesn't really cover them. but my guess is they form a high pass filter to ensure the triggering 'blip' from the switch does not contain low frequency interference, the gates needing a discrete signal within a certain fairly quick timing/frequency range in order to reliably activate. i would guess that 1000pF worked fine initially, and that the reason they'd set it that high is to try to ensure a nice strong signal that doesn't get weakened by the filter. but after time the switch's 'blip' includes additional lower frequency noise that 'confuses' the gates. the original circuit would have placed the pole for low-frequency rolloff at 796hz, whereas this one puts it at 4,021hz.

does this sort of acceptable frequency range to trigger the flip flop circuit, have to do more with the characteristics of the IC, or the value choices for the other components in the circuit? still wondering if it'd be a good idea to replace the IC. i'm a tube amp tech that's juuuuuust breaking into solid state stuff, and i'm not really sure if these suckers wear out or not, or how to tell. also wondering if you think it'd be worth it to replace the switch as well - based on my theory i'm leaning towards it, but i don't know how accurate it is. and there are still things that don't make sense to me - why increasing the switch shunt cap C47 to 18nF made the issue so much worse, for example.


one thing i am fairly certain of is that, besides them having the unimaginative aesthetics with all the charm of a puke covered trashcan, this switching weirdness must be a huge part of the reason you rarely see these Ibanez pedals around anymore, despite most folks on forums loving their actual effects. it seems like something that only set in after most folks' pedals were out of the warranty period, and while there's definitely some success stories just from replacing the switch, they seem in the minority, and just sourcing that switch alone can be tricky.

however, Boss pedals are still ubiquitous of course, and i keep hearing they used the same switching arrangement... but of the several Ibanez and Boss schematics i looked up in the process of reasearching this, i'm pretty sure i only saw Boss use 220pF trigger caps. since there's only so low you can set those shunt resistors before bleeding too much of the 'blip' to ground, i'm thinking it was fairly essential to make those caps small for the purposes of cleaning up the 'blip'.


Rob, you're a champ. can't thank you enough. love the goodwill and the high level discussion here, i have so much respect for you pedal techs. hope to branch more into that realm more over time. i know tube tech is probably scarce on this site if it's around at all, but feel free to reach out for any help on tube stuff, and if you or anyone you know ever needs a tube for some project, DM me, you've got one on the house!

Quote from: mxrshiver on August 26, 2024, 03:57:11 PMFIXED.

Really great news.   There was some hope it could be fixed.

QuoteC47 to 18nF, R49 and R50 to 180K, and, it seems most critically, C45 and C46 to 220pF.

progression of changes and results:
- C47 to 18uF, resulting in ON/OFF switching success rate changed from about 20-30% before, to 0% for ON, but 100% for OFF after
- R49 and R50 to 180K, C45 and C46 to 370pF, resulting in switching success rate of about 30-40% evenly balanced between ON and OFF
- R49 and R50 to 150K, resulting in switching success rate of about 20-30%
- C45 and C46 to 220pF, resulting in 100% perfect switching success

Thanks for summarizing.

Quotei was going to try 330pF for the trigger caps, but then i realized i'd seen most diagrams in articles about this using 220pF, and lo and behold, it was what i needed.

The Boss circuits have 470pF feedback caps and 220pF trigger caps.   The Ibanez has 1nF feedback caps so keeping the same cap ratio as the Boss means the equivalent for the Ibanez caps would be  (220/470)*1000 = 470pF.

Anyway I'm glad you played around with it.   The values I posted were only a start, with some hope of success.

I'm amazed the original circuit worked at all!   It's not even in the working region.

I noticed that there's a lot of complaints about the CF7 switching in particular.     The other pedals from the ToneLok series have an issue with the switch cap being too small, like the SoundTank series.   I had a look at the schematics of a number of other ToneLok circuits and they all use transistors, basicially the common Ibanez flip-flop.  The CF7 seems to be unique in its use of gates instead of transistors.  It appears the CF7 gate based circuit isn't quite there.

Because of that I don't have any second thoughts about modding the values.  Sometimes when these types of problems come up you ask yourself why are we changing the values when there's other pedals out there that work.  Clearly the CF7 design is marginal and there's history to prove it.

Quotei don't have too much of an understanding of the roles of these particular components - despite how awesome that article i linked to is, he doesn't really cover them. but my guess is they form a high pass filter to ensure the triggering 'blip' from the switch does not contain low frequency interference, the gates needing a discrete signal within a certain fairly quick timing/frequency range in order to reliably activate. i would guess that 1000pF worked fine initially, and that the reason they'd set it that high is to try to ensure a nice strong signal that doesn't get weakened by the filter. but after time the switch's 'blip' includes additional lower frequency noise that 'confuses' the gates. the original circuit would have placed the pole for low-frequency rolloff at 796hz, whereas this one puts it at 4,021hz.

does this sort of acceptable frequency range to trigger the flip flop circuit, have to do more with the characteristics of the IC, or the value choices for the other components in the circuit? still wondering if it'd be a good idea to replace the IC. i'm a tube amp tech that's juuuuuust breaking into solid state stuff, and i'm not really sure if these suckers wear out or not, or how to tell. also wondering if you think it'd be worth it to replace the switch as well - based on my theory i'm leaning towards it, but i don't know how accurate it is. and there are still things that don't make sense to me - why increasing the switch shunt cap C47 to 18nF made the issue so much worse, for example.

These circuit are quite tricky because there's a lot of interactions and details.   

The resistor values ensure the outputs can hold a stable state.   With gates there's some technicalities in making the 200k resistors too small and the output not holding the state.   However, if you make them too large it's hard to trigger.

If the trigger caps are too small then the blip is too small and cannot kick the state of the circuit.  However if the trigger caps are too large the trigger caps stop the circuit from changing state.  (In fact that's why you need the feedback caps in the first place.)

So there's a delicate balance choosing the "200k" resistor position to make the trigger sensitive enough that a cap can be chosen.  Then the trigger cap has the be selected to have just the right amount of blip - and then there's a chance the circuit won't work if things aren't just right.

Quoteone thing i am fairly certain of is that, besides them having the unimaginative aesthetics with all the charm of a puke covered trashcan, this switching weirdness must be a huge part of the reason you rarely see these Ibanez pedals around anymore, despite most folks on forums loving their actual effects. it seems like something that only set in after most folks' pedals were out of the warranty period, and while there's definitely some success stories just from replacing the switch, they seem in the minority, and just sourcing that switch alone can be tricky.

I guess a lot of pedals in that era aren't as common as they were.   Even the Sound Tanks aren't that common now.     I suspect it's not worth selling them.

Quotehowever, Boss pedals are still ubiquitous of course, and i keep hearing they used the same switching arrangement... but of the several Ibanez and Boss schematics i looked up in the process of reasearching this, i'm pretty sure i only saw Boss use 220pF trigger caps. since there's only so low you can set those shunt resistors before bleeding too much of the 'blip' to ground, i'm thinking it was fairly essential to make those caps small for the purposes of cleaning up the 'blip'.

Yes the Boss and Ibanez each have their transistor based circuits which are well tested over time with zillions of units out there.
(Some of the very early ones are different.)

QuoteRob, you're a champ. can't thank you enough. love the goodwill and the high level discussion here, i have so much respect for you pedal techs. hope to branch more into that realm more over time. i know tube tech is probably scarce on this site if it's around at all, but feel free to reach out for any help on tube stuff, and if you or anyone you know ever needs a tube for some project, DM me, you've got one on the house!

No problem at all.   It's nice to have a happy ending to an otherwise tricky problem.   

aaaaaghhhh should have left it well enough alone... decided i had put enough into this pedal that it was worth recapping it and replacing the switch as well, and afterwards the flip stopped flopping again! neither the new 220pF/180K circuit, nor the original 100pF/200K version, produced any results. i had to bring them both down to 150pF/150K, and now have it working every time again.

is it possible for the 74HC04AP to be wearing out? i'm surprised that the working circuit with a brand new switch was even farther from the original design.

Quote from: mxrshiver on September 03, 2024, 08:41:05 PMaaaaaghhhh should have left it well enough alone... decided i had put enough into this pedal that it was worth recapping it and replacing the switch as well, and afterwards the flip stopped flopping again! neither the new 220pF/180K circuit, nor the original 100pF/200K version, produced any results. i had to bring them both down to 150pF/150K, and now have it working every time again.

is it possible for the 74HC04AP to be wearing out? i'm surprised that the working circuit with a brand new switch was even farther from the original design

Unfortunately these circuits rely on the switch bounce time being short.   Even though you have a new switch the way it bounces can be different than before, perhaps in an undesirable way as far as the circuit is concerned.    Try changing C47 to 47n, even temporarily.   I've looked at these circuits in the past and a 47nF cap should fend off switch bounce from all but the most dodgy switches.

It's even possible the 220pF cap tweak before tuned the circuit to matched to switch to some degree.  I'd try to leave the resistor at 180k.   Then try 220pF upto say 470pF for the cap so see if you get any change.

It's unlikely the 74HC04AP would wear out.   However, there's a small chance it got damaged.

If the circuit is erratic that just means it's not triggering correctly and playing with the caps might help.  If it no longer works at all then it's possible the chip is fried.    What I would do is alternatively short the gate *inputs* to ground to see if you can force the flip-flop to the on and off state - if you can do that it means the gate is working.   For this test the flip-flop should remember the state.  If the circuit goes into one state with the input shorted but then immediately changes state with the input short removed then that could also mean the gate is fried.