Aion Refractor (Klon Centaur Klone) Unique Issue?

[blackieNYC]

since there are successful builds here, how does this thing compare sonically to an MXR D+?

[EBK]

since there are successful builds here, how does this thing compare sonically to an MXR D+?

Never experienced a MXR D+, sorry....  I guess I would describe this circuit as a boost-plus-flavor, if that makes sense.  It's not crazily unmusical at extreme settings, nor is it a take-charge-guitar-solo type of thing, in my opinion.  I think it sounds rather "nice".

[Off topic] Please tell me what that beautifully complicated box of insanity is in your avatar pic. [/Off topic]

[uviteru]

I've finished soldering my Refractor, with the part substitutions I mentioned previously (turns out my inventory of TL072s has been depleted, so I can't report on using those  ).  Works perfectly, as far as I can tell.  I do not get reverse gating at high gain, but there is some sort of very noticeable compression artifact when gain is maxed, which I suspect is from the diodes. 

What exactly do you mean by compression artifact? That could be similar to what I'm getting. I actually don't have the pedal anymore, someone wanted to borrow it, so I lent it to them--I plan to build another one sometime soon too see if i get the same behavior.

As a wild guess at what could go wrong while building one of these, I'd suggest maybe a dead spot in one half of the dual gang pot could be the culprit.  The gain control trades clean for saturated over its travel, so in the event of a dead spot, the affected component would drop out or operate in an unexpected manner.

Thanks for the tip. I could definitely see that happening, because it seems to happen when the gain is cranked no matter what the volume and tone.

[banjerpickin]

I've built three of these boards over the years, all three being the older layout version, and never had this problem (or cnspedalbuilder's clean boost issue).  One of them, the one I kept for myself, does have an issue with the tone pot - you have to crank it to get any treble out of it, but none of the others suffer that issue.  Must have put a wrong resistor in somewhere.  I grabbed some NOS D9E's for the diodes, but found that BAT41's actually sounded a little cleaner and less spongy.  Go figure.  I built toggle switches so you could swap the two sets.  At max gain both pairs get a bit fuzzy but never any dropoff or unpleasant compression artifacts (regardless of volume and playing intensity).  What diodes did you use?  I wonder if there's some QA issue with your particular diodes.  I'd try swapping them, even to a different kind altogether and see if you still have the same issue. 

If that gets you nowhere, I'd take an audio probe and follow the general path of the signal just below the settings where you get the dropoff, and then at the settings you get the dropoff.  Perhaps you will discover the culprit part or section that way.  Looking at your photo, I'd reflow all 6 leads on the gain pot just to be safe, or maybe even try a new pot.  Great design and clean build, that's frustrating it's not working!

If you could post a video or audio clip, that would be very helpful. 

It's hard to describe and hard to capture on a cell phone recording

This has me very curious as to how perceivable the issue is     

[banjerpickin]

One last thought:  my Klone operates best at about 11 o'clock gain with the BAT41's and volume set to push any clean tube amp on the edge of breaking up into overdriven sound, as our lord and savior Bill intended  .  I've also been known to use it as an "always on" with lower gain settings before other overdrive pedals to drive them harder.  I've never used the gain past 12 o'clock in any performance or recording setting, it just doesn't sound good.

So, if it's working below 12 o'clock...maybe just let it ride.  Although I know that I probably wouldn't be able to stop messing with with a pedal that I knew wasn't 100% functional.

[EBK]

By "compression artifacts", I think I was maybe referring to that same "spongy" quality.  Not necessarily objectively unpleasant compression, just altered the feel of the pedal in a perceptible way.  Kind of hard to describe adequately.
Using BAT41s, but I have some BAT43s I could try as well as some mystery germaniums.

[banjerpickin]

By "compression artifacts", I think I was maybe referring to that same "spongy" quality.  Not necessarily objectively unpleasant compression, just altered the feel of the pedal in a perceptible way.  Kind of hard to describe adequately.
Using BAT41s, but I have some BAT43s I could try as well as some mystery germaniums.

Yup I know exactly what you're talking about.  Spongy is the perfect word to describe it.  Or maybe splatty/blatty.  Once you get north of 3/4 o'clock it just seems to lose all focus, like a RAT (which I don't prefer) except even less desirable.  The circuit wasn't meant to be used as an overdrive/distortion, it's really a complicated and fancy booster.  This is supposedly why Bill only took orders over the phone, so he could talk you out of buying one if weren't going to use it "correctly" in his mind.  And definitely try out those other diodes.  It feels kind of dirty as Bill was painstakingly meticulous with his mystery diode selection...but since he bought them all up he leaves us no choice but to experiment 

If you really want to dial it in, work the tone and volume knobs until your amp starts breaking up and then dial up the gain to taste, as opposed to most other OD pedals where you start with gain and then use volume to normalize.  This is not a good bedroom volume pedal, you're gonna have to crank your amp at times to take full advantage of the effect.  Vox amps, especially my AC30, can get there at lower volumes because of the relatively low headroom.  My Twin Reverb has to get brutally loud before the Klone can help it out.

But, none of this describes the gating/dropoff problem that uvitero is having.

[midwayfair]

It feels kind of dirty as Bill was painstakingly meticulous with his mystery diode selection...but since he bought them all up he leaves us no choice but to experiment 

It's not a mystery -- they're 1n34A, which is exactly what they were speculated to be based on measurements when it was traced. He flat out told people and confirmed it. I don't feel like finding the post where he did but it was on TGP and by that point people barely cared because it doesn't make nearly the difference people think it does.

The diodes are hard clipping. The only effect they might have on oscillation being produced in this pedal is if you remove them entirely and the gain becomes high enough to create motorboating. Whatever's causing the op's problem is very, very unlikely to have anything to do with the diodes (unless somehow they aren't connected to the circuit), and the circuit dipping and compressing is not normal. The only compression in this design should only come from the fact that the signal is clipped when the gain knob is turned up.

[banjerpickin]

Roger that on the 1n34A's  - I realize I probably look like a Klon fanboy but I actually think it's just alright in terms of usability.  The controversy surrounding the pedal, however, is fascinating and I've read quite a bit on it...not sure how I never stumbled across a thread where he admitted to the 1n34A's.  Maybe he ran out of his original mystery diode and had to make a substitution   
 
But, I do hear a very noticeable difference switching between the D9E's (GE) and BAT41's (SI).  Is it the difference in material or forward voltage?  I find the more I learn about electrical engineering the less I actually know, but even in a hard clipping situation wouldn't the forward voltage of the diodes matter?  I've looked at spec sheets and the D9E's come in right at .35v, just like the 1n34A's.  The BAT41's come in higher at .38~4v depending on what you look at.  Is that enough to implicate the tonal the difference?  The BAT41's are slightly louder, which makes sense, if they are letting a little more signal through before clipping.

I've also read somewhere that GE diodes have a less linear response to frequencies (high frequencies pass threshold faster) which might give the qualitative quantity of "spongy/squishy" with the low end not getting clipped as much.   And SI diodes have a faster forward/reverse bias change over, plus a flat response, hence the qualitative quantity of "tighter" sound. 

So, take me to school midwayfair. 

[banjerpickin]

.

[midwayfair]

Roger that on the 1n34A's  - I realize I probably look like a Klon fanboy but I actually think it's just alright in terms of usability.  The controversy surrounding the pedal, however, is fascinating and I've read quite a bit on it...not sure how I never stumbled across a thread where he admitted to the 1n34A's.  Maybe he ran out of his original mystery diode and had to make a substitution   
 
But, I do hear a very noticeable difference switching between the D9E's (GE) and BAT41's (SI).  Is it the difference in material or forward voltage?  I find the more I learn about electrical engineering the less I actually know, but even in a hard clipping situation wouldn't the forward voltage of the diodes matter?  I've looked at spec sheets and the D9E's come in right at .35v, just like the 1n34A's.  The BAT41's come in higher at .38~4v depending on what you look at.  Is that enough to implicate the tonal the difference?  The BAT41's are slightly louder, which makes sense, if they are letting a little more signal through before clipping.

I've also read somewhere that GE diodes have a less linear response to frequencies (high frequencies pass threshold faster) which might give the qualitative quantity of "spongy/squishy" with the low end not getting clipped as much.   And SI diodes have a faster forward/reverse bias change over, plus a flat response, hence the qualitative quantity of "tighter" sound. 

So, take me to school midwayfair.

To the extent I am qualified to take anyone to school:

Bill said they have always been 1N34A, but that he bought the entire stash of the particular 1N34A that existed. The 1N34A does have a slightly unusually high Fv when most of us measure them in person, but they are relatively consistent among manufacturers, for a Ge part anyway. For comparison, my D9Es (and actually ALL my Russian Ge diodes) are marked .27V while most 1N34As measure .35V.

Yes, the Fv matters, because of course it does. Two different forward voltages will not be identical at all points of travel on the knob except when the gain is at 0 (because then the diodes get no signal). But the dirt stage op amp runs on +9V, not +18V/-9V like the second one, and the gain is absolutely high enough to clip a raw guitar signal starting pretty early in the pot's travel. For the most part, most people would agree that you can hear a difference between drastically different diodes between about 10:00 and 2:00 on the dial. Before 10:00, the dirt signal could be small enough to not make a difference, and the difference is masked by the clean signal. After 2:00 and all the way up (remember it's a linear control, not an audio control, so there's not a huge difference there), the clean signal's almost gone and the op amp is probably clipping all or most of the time anyway and maybe you're getting a bit more volume with a higher Fv but not too much more clipping. I mean, what's .1V of difference in the Fv of the diodes when your signal's 9V coming out of that op amp stage? Go ahead and run the calculation to convert that to a decibel difference and I'm not even sure humans can detect a difference that small! [just to be clear, that .1V difference IS amplified 200x, so it's noticeable at the output of the pedal, but I'm just trying to convey how small a difference it is to begin with.) So the fuss over the diodes has ALWAYS seemed a little overly fussy to me. I do think there is a reasonably cool difference between twice the Fv (like silicon) and the 1N34As at moderate gain levels, where you are still getting plenty of dynamics.

I agree too that there seems to be a clear sonic difference between germanium diodes and silicon diodes. Just not mush difference between two pairs of the same substrate.

So just thinking of some things that diodes can differ in that might affect that? Capacitance, leakage, resistance across the diode, speed (some diodes ARE faster, like the 1N5363BG are super quick, for instance). Some of these might make a difference in the sound, maybe in a blind A/B test even.

What they DON'T account for is some sort of "dipping" compression like what the op is experiencing. Unless I'm completely misunderstanding what the op's talking about, that sounds like something's not working correctly. Since the op is also having some clean boost issues, there's something wrong. I'm usually pretty good about being able to help people, but I'm a little stumped. At this point I'd probably start replacing ICs and reflowing solder joints with new solder and measuring pots and continuity to find something that's not otherwise observable with voltages and such.

[EBK]

Just to clarify, when you say "dipping" compression, are you referring to the spongy/squishy feel I get at high gain or the blocking distortion (reverse gating) uviteru is getting?

[banjerpickin]

Just to clarify, when you say "dipping" compression, are you referring to the spongy/squishy feel I get at high gain or the blocking distortion (reverse gating) uviteru is getting?

Dipping compression = reverse gating. 

The spongy/squishy feel is just how the circuit sounds at higher gain leveles, nothing wrong per-say, just not very useable! 

To the extent I am qualified to take anyone to school:

Thank you for the in depth explanation!  It looks like I was at least on the right track.  That's interesting your D9E's are at .27v because I just measured what's left in my bag out of curiosity and got .35v.  They were originally recommended to me by Aion's site as a good 1:1 replacement for the mystery Klon diodes, can't even remember where I bought them from, being a few years ago. 

At this point, for uvitero, it certainly sounds like a faulty, damaged, or not properly connected part.  We shall see... 

[midwayfair]

That's interesting your D9E's are at .27v because I just measured what's left in my bag out of curiosity and got .35v.

Fv measurements vary between multimeters (it has to do with current or impedance or something that someone smarter than me can chime in to explain). It's annoying -- I had to relabel a lot of stuff in my bins when I switched to a new meter.

[PRR]

> Fv measurements vary between multimeters

Forward Voltage varies with current (as you suspect).

There is no standard test current among meters which include a Fv test feature.

You can test the tester to find the current it uses.

You can build your own tester and *know* what you are testing. A 9V battery through a 8.4K resistor will be pretty-near 1mA for all dark (non-LED) diodes. And 9V through 85K will be 0.1mA, which is pretty near where we probably begin to get rounding in clippers. For most materials, diode drop changes about 20mV (0.020V) for every 1:10 change of current.

[Les Turnbull]

Would this build work with TL071CP .

[EBK]

Would this build work with TL071CP .

No.  That is a single op amp package.  You need a dual op amp.
TL071 = 1 op amp
TL072 = 2 op amps
TL074 = 4 op amps

[Les Turnbull]

Cheers Eric , Have a few TL071 but will have to order some TL072 .

[uviteru]

So I just got the pedal back from the guy that borrowed it, and he said it wasn't working. So I opened it up and the power/ground connection was shorted! I did some troubleshooting and identified the zener diode that bridges the ground/power connection had shorted causing the issue. Does anyone know what would cause the diode to short? I suspect he might have used a center positive supply instead of center negative, but he hasn't responded to me yet...

Also, what would be a good replacement for a 1N4742 zener? Could I used something else? I'm fresh out of those unfortunately....

[EBK]

but he hasn't responded to me yet...

Hmm... 

Also, what would be a good replacement for a 1N4742 zener? Could I used something else? I'm fresh out of those unfortunately....

Just remove it.  Overvoltage protection only.  Just keep your DC input below 12V (actual volts, not nominal volts), and you'll be fine as long as you don't loan it out again, assuming it comes back to life.