Setback / DOD FX-90 project - noise floor / hiss reduction assistance

[geetarm@n1]

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