Messages

Hello all.

Hoping to find some builders of the Madbean Setback who have experience with the build. 

In particular, I'm looking for some guidance on how I can troubleshoot/reduce hiss in my Setback build. It may be normal or may be that I have a problem.

Unlike a few other short analog delays I have or have built (a GGG AD3208, BYOC Analog Delay and an old DOD FX90 (later 4007UBE equipped version), the Setback has too much inherent hiss / noise floor to be used in front of a distortion device (the aforementioned delays impart completely negligible degree of hiss which is not enough to be amplified noticeably by the following distortion device when they're set to very low levels of effect mix/blend (within the first quarter turn of the control which adds in in the delayed signal). 

Sure, I could use the Setback after my distortion, but I love the "ducking" effect of putting delay before gain pedals for lead playing, so my intended use of this build is dedicated pre-distortion with the MIX set very low (typically feeding into a Madbean Archibald, a Barber Burn Unit or Tube Screamer).  Of course, the distortion boxes will amplify any hiss they receive, so I'm looking to minimize it, even at the expense of clarity in the delayed sound if need be. It's just being used as a little cushion.

My Setback's inherent hiss/noise floor is more in line with PT2399 delays which I've always considered "noisy" and which I'd only use for their longer delay times and then, only after distortion devices. My Setback is actually hissier than two of these, quieter than two of them.  It has more hiss than a CE-2 clone (which I can use in front of a distortion), but also has more hiss more than an XVive Echoman and Madbean Skoolie (both of which I can't use in front of a distortion either).

So, for this discussion and since it exhibits the behavior from the get-go I'm primarily focusing on the analog bypass sound itself (with the MIX knob turned ALL the way counter-clockwise) which seems to me to be the simplest scenario where the symptom presents itself.  There is substantial amount of constant hiss even at this setting (and yes, this constant noise floor changes - it deepens and get a little louder as I move the MIX control clockwise (more of the effect equating to more of the noise is certainly expected).

Note that this is not necessarily an unusable degree of hiss for normal post-distortion pedal usage.  In fact, I played on it into a clean amp today (with no overdrive/distortion device turned on) for 30 minutes without being bothered by the amount of hiss at all.  It's a very nice sounding delay.  It is calibrated to the best of my ability (I used a scope the first time - but I'm a novice with a scope, so I did it by ear the second time) and I have no clock artifacts when I do turn the delay up to more significant MIX settings (something I cannot say of my real DOD FX90, but it is absolutely hiss-free when used at low MIX settings into a distortion box).

Using an audio probe, I can hear plenty of hiss is present at Pin 1 of the opamp (IC1_A, LM1458), but that is less than what is present on the tip of the output jack, so there is certainly some hiss reduction going on during it's path throughout the circuit.

Any ideas where to begin either taming it with caps, troubleshooting if it is not the expected degree of noise from the analog through component of this circuit or simply modding it?  In a perfect world, my input signal would be untouched when the MIX knob is at minimum, and remain hiss-free and strong enough on its own to dominate the wet signal as it is blended in (again, given that I'll just be using it within that first quarter or less of the MIX control's rotation). 


Perhaps one of my components is bad (or simply of lesser quality).  In case any of these makes have a reputation for noisy behavior, I'll list what I used:

ICs used:


National 1458P  (I've tried three different ones, two from different suppliers plus a few other opamps I had lying around - NE5532 and a LF353 - all with no significant reduction in hiss)

XVive MN3005  (I've tried two different ones, both purchased at different times, so likely not from the same production run)

Cool Audio V571D (I've tried two different ones, but purchased together, possibly from the same production run)

Cabintech CT3101  in place of MN3101  (I've only tried the one device, but if this component's job is one that could introduce hiss with the MIX control fully counter-clockwise, I can dig through my builds and pull out a proper MN3101 to try out.  I'm sure I have something around here with one).

 
* all parts are as specified in the build doc (ex: I used 2n5088 where specified, not subbed for a similar part)
* used 2.2M for R40 (using it with 12v power supply purchased from Mouser)
* used tantalum for C12 and C19
* used electrolytics for all 1uf caps
* all caps are 16v or higher
* As for the volume drop: I used the stock values for R4 and R48
* voltages are all very close what is specified in the build docs


Here's a link to the Setback build doc:

https://www.madbeanpedals.com/projects/_folders/Delay/pdf/Setback.pdf


Thanks in advance for any assistance and apologies for the long-winded description.


Greg

Yes - and pin 5 turned out to be an easy place for me to ground that 68K resistor without flying its lead across other circuitry or working under the board.  So the cap and the resistor just rest on top of IC4 (whose pin 2 was bent up 45 degrees and then clipped short giving me a flat spot to mount the other side of the 68K and that leg of the capacitor.  Ended up looking pretty clean, I'm happy to say.

Thanks again for the tip, Dan

Dan -

Ah - I wired it up wrong the first time - never lifted pin 2 out of the socket (among other mistakes with the resistor and at pin 5).   Thanks for the explanation and pic - that helped.   I connected it up properly and the result is as you expected - the tick tock is now completely gone!!   I'm not going to put as many exclamation points as this deserves, but believe me - I am ecstatic.   I've been just sitting here listing to it swoosh through a loud amp with the guitar turned off for the last 15 minutes and it's lulled me into a relaxed out state of mind - a sure sign of clean, smooth phasing.

Thanks to all for the help, the lessons and for the project itself Mr. Bean.     

whoops - I posted before completing my final sentence:

So assuming that you are correct that IC4-A in the Glasshole is actual the target for this fix, it unfortunately didn't work for me.

Zerro - I stripped the new diode out of the equation (the one I tacked on to serve pin 11) leaving just the added 120 ohm resistor and 10uf cap.  The other diode D9 is board mounted (and tied to the 120 ohm resistor I added in line) I've not tried removing it, but I did jumper it.    Neither of these changes affected the ticking.

danfrank - Thanks for the suggestion.  I gave this a try as well on IC4-1 - which I'm taking as being IC4-A (tying a 33n between pins 1 and 2, then a 68K resistor between pins 2 and 5), then unfortunately, this didn't change the behavior.

Given that on this project, IC4-A's pins 2 and 5 are tied to ground, yet the examples shown of this 'soft square' trick from the .pdf you linked don't have these pins grounded, I'm wondering if to implement it here means that I need to lift those pins from ground a bit or something.  I tried disconnecting them completely from ground (again, with a stunted row of sockets which allowed me to not make that connection to those pins of the original board-mounted socket) and that "fixed" the tick - by stopping the oscillation altogether 

Also, here is another link to that PDF which doesn't require an account with scribd.com (it's a good article - thanks for that)  -  http://www.moosapotamus.net/files/stompboxology-mo-tremlo.pdf

So assuming that you are correct that IC4-A

I finally had some time to sit down with this again and tried what you (Zerro) suggested in your last post.   

I lifted the cathode side of D9, attached that to one end of a 120ohm resistor, then attached that where D9's cathode's pad on the board.  That connects to the + of C23 and to pin 4 of IC4. 

To take care of decoupling and creating a reservoir for IC4's pin 11, I added a second 14 pin socket with the pin 11 section removed on top of my original socket.  I soldered a 120 ohm resistor and the (-) leg of a 10uf cap to IC4 pin 11, then soldered the (+) leg of the cap to ground.  Then soldered a 1n914's anode to the other end of the resistor and inserted its cathode into the original socket for IC4 pin 11.  Looked  pretty questionable, but the voltage readings told me that all was well in terms of connectivity.

The result, unfortunately, was the same - tick tock still present at the same volume level it was prior.   

I took a last shot at re-calibrating it and doing the chopstick trick (moving around the now shielded Input/Out leads which run from jack to switch) just for grins, but no go.   I think I'll keep this thing as a great sounding 2 and 4 stage phaser and maybe try building another and see how that goes.

Many thanks for the suggestions and although I wasn't able to resolve it, I definitely have learned a thing or three in the process.  One eye opener was how all of these additional components I've added throughout this process have had so little effect on the core sound - it still sounds just as rich/full as it did originally.

Got it.   I'll reply back in a week or so once I've had time to order what I need.     Thanks again.

Zerro - you rock.   Many thanks for the detailed information and know that I appreciate the lessons you're providing.

As for the caps, are you suggessing small MLCC or actual ceramic discs?

Update:


Zerro,

I've tried the following but unfortunately there has been no reduction in the ticking (which again is present only when in non-bypass mode and only present on the Six and Eight Stage settings).   Granted, I'm using just my ears for this test, not an oscilloscope, but the fact remains that the ticking remains loud enough to be audible after about 4 seconds into the decay of the ringing out of a struck chord played, with no other FX in line and into a non-overdriven amplifier.  There is no other unwanted noise - a nice, quiet effect other than the ticking.  I still have dead silence (in terms of LFO tick) in the Two and Four Stage settings.



a)  Replaced the LT1054 with two 9v batteries.    This did not reduce/eliminate the ticking, so I reverted to using the LT1054/OneSpot power supply.



b)  Soldered two small-sized 47nf poly caps from pins 4 and 11 to ground on IC4.  This did not reduce/eliminate the ticking, so I left this in place and then:



c)  Soldered two small-sized 47nf poly caps from pins 4 and 11 to ground on IC1.  This did not reduce/eliminate the ticking, so I left this in place and then:



d)  Soldered a small-sized 47nf poly cap in parallel with R40.  This did not reduce/eliminate the ticking, so I left this in place and then:



e)  Harvested a ferrite bead from an old computer keyboard and replaced my shielded output wire (running from switch to output jack) with a longer, non-shielded wire and looped it three times through the bead (the max that the bead's diameter will allow).    This did not reduce/eliminate the ticking.



f)   I re-flowed the solder on all of the LDRs, this did not reduce/eliminate the ticking, so I then:



g)  I went back and double-checked my voltages.   Everything still looks good and matches with a few mv to what Brian's build doc specifies with the exception of IC3's pin 7, which he has listed as being -10mv, but in my case, I read +12mv.  Is this significant?



Note that I've not tried using ceramic capacitors for this test as you suggested.  I can certainly order some if you feel that will be the magic bullet.

Given that none of this has had any audible impact and that (as I understand it) your suggestions are aimed at reducing an induced or indirectly picked up (same thing?) noise, then perhaps the problem is rather something directly injecting it or not filtering out as it is supposed to do?   Is there any particular set of components whose job it is to do so?    Perhaps I have a bad component or component with incorrect value?  Two to four bad LDRs?   A bad solder connection?  Perhaps an incorrectly oriented polarized component?   Given that the thing works and the voltages are all in pretty much in order tell me that the latter is not the case, but at this point I'm open to trying anything and also open to assuming that I could have done with component choices or build/soldering technique could be the culprit.   Any further theories are welcome and will be much closer to the mark than what I could come up with, so bring it on!


Thanks again

Gotcha - I'll give it all a try at my next opportunity and will report back in a few days.

Many thanks and enjoy your weekend

I just tried a few more shots in the dark one at a time, all without success.

Repositioned the power lines away from the rate pot temporarily

Tried a different LT1054

Swapped the positions of the TL074s in IC1 and IC4

Re-flowed the solder on the components in the neighborhood of IC1 and IC4 and the LDRs

Played with the trimmers - sweeping these around had no effect on the amount/level/quality of the ticking.

Putting my finger on the Rate pot lug 3 (connects to IC4 pin 1), the quality of the ticking does not go away or even really drop in volume, but it does change in quality (less full-frequencied, cuts some low end out of the signal).


Admittedly, that's some willy nilly troubleshooting and definitely where I draw the line today or I'll end up breaking something irreparably.


Also possibly of note:  I used all 1/4 watt metal film resistors, so I have very few resistors lying flat to the board.   All of the resistors running alongside IC4 have their tall end up in the air right next to IC4 and the flush-to-board end is therefore on the side by the  Depth pot.

Thanks for the reply Zerro

I tried what you suggested with a pair of 100nf caps and it didn't help, unfortunately.

I also tried a few other things:

a) Disconnected the rate status LEDs from the offending stages (6 and .  No difference.

b) Replaced my large size C18 with a smaller, lower voltage version of the same (thinking that the large one was acting as an antenna).  No difference.

c)  Wondering if my plastic light shield's tie-down harness (thin metal wire) was acting as an antenna, I removed it and held the light shield in place.   No difference.   

d)  I then removed the light shield and the ticking stopped.  Of course, my swooshing also stopped.    I can make it fade in/out by blocking the light that reaches the LDRs and if I do this after first strumming the guitar, the instrument's signal is completely audible, just without any phasing (or ticking) until I start bringing in the light shield.   

Also, for what it's worth, I ended up using a diffused green LED for the lamp as it was my closest in value to the 520 nm mentioned in the build doc.

So the ticking is audible only when Phasing is audible.   Does that narrow it down to any particular set of components?


And again, my issue in is non-bypass mode only.  I clarify that again as it doesn't match the Subject line of this thread.  Let me know if it is more appropriate to open a new thread rather than convoluting this one further.

Thanks

Bump   

I took another listen to mine again today and found that with the amp cranked fully (low watt amp) I could indeed hear some clock noise in bypass after all (mine already has straight runs from the tip of the input/output jack straight to the switch with shielded cable) but this was only when it was in the middle of a long chain of DIY effects powered by a single One Spot, two of which have charge pumps in them.  Asking a lot from the One Spot for sure.  When I took half of the downstream effects out of the picture the clock noise in bypass became very negligible but still there.  When I reduced the chain down to where the Glasshole was the only thing being powered by the One Spot and it was the only thing between my guitar and the cranked amp there was no clock noise at all on any of the various "stages" settings (again, in bypass).   So in that way I'm good to go.

However, I still do have the issue of the clock noise in non-bypass mode on the 6 and 8 stage settings.  Testing for this issue has primarily been done with the Glasshole isolated - alone on the One Spot, no other FX in the chain, but of course it is also present when it was in the chain as well.  The clock noise present across multiple amps, multiple guitar cables.  Also present when powered by a 9v battery rather than the One Spot I mentioned earlier.  I'm sure that some of my tests are no more than shots in the dark and have nothing to do with the actual issue, but I figured that I'd rule out what I could before posting again.   Let me know if you have any recommendations of what I should try in order to track it down and lessen or eliminate it.

I just reread my original post and I mentioned that it was a "continual" ticking and should clarify.    That is only the case at higher speeds.  When at slow speeds I hear two tick/thumps - one at the peak of the sweep (peak of the indicator LED brightness) and once at the bottom of the sweep (well after the indicator LED has turned off).


I've also spent some more time today listening to it with the amp up loud but without a guitar in hand - just listening.   The tick/thump is absolutely not present at all on the 2 and 4 stage settings.

The volume of the tick/thump is a little less than what I get when switching between the different stages with the rotary with the effect active (neither are sounds that worry my in terms of speaker health - unlike the pop from engaging either of the Phase 45s I've built, but that's another discussion altogether). 

The overall hiss (and its low end content) and the click/pop when engaging the effect are all very negligible.   In other words, it is hard to believe that I've miswired something, used an incorrect part value or that I have a bad solder joint as everything really seems to be working as expected other than the tick at the top and bottom of the cycle of stages 6 and 8.

If there is any reason to think that the ICs are to blame I can swap them around but I'd prefer not to pry them out of the socket arbitrarily.