Klon Centaur Clipping Diodes Pt.2: Curve Tracing

[@electronicaudioexperiments]

Interesting video. I appreciate you taking the time to unravel these mysteries for the general public in an approachable fashion. It is remarkable how much a subtle difference in the knee and slope of these curves can change the sound if you listen hard enough. Most of what makes the Centaur/KTR circuit is its unique design, and this design relies pretty heavily on the extreme compression of Ge diodes. Often, the differences between diodes are only apparent at high gain (though if you're watching this video you probably know that already). But, and this is true of any diode clipping configuration (not just Ge), it is shockingly difficult to get two diodes with different I-V curves to match up at all gain settings. Just as an aside: I think the Centaur/KTR is a creative high water mark in the world of pedal design, and the proliferation of klones says a great deal about the general laziness/risk aversion of the industry. A lot of people want a piece of that pie simply because they can't come up with anything better, while simultaneously claiming Bill is a scammer. Can't have it both ways, y'know?

[@makingmusiconline2309]

The only thing I’d have liked to have seen, in addition to what you did, would be to take, say, a half dozen of the ITT diodes and see what the variation was between them. It might be so little that it was imperceptible, or it could show a cone of values that overlapped other diode traces. Thanks for your hard work and investments to produce these videos, much appreciated!

[@Stu66orn]

Just awesome channel with pedal info. Really one of a kind where pedals get gutted and expained. Love this!

[@graybenchelec]

Thank you!

[@jeancrafterpro]

O melhor klon clone está no Brasil Collateral fx

[@lemmy8284]

After watching this video, I took out all the germanium diodes I had and checked them, and finally I chose the Mullard OA90 that showed 0.27V on my multitester. And the sound of my clone turned really hot. Thank you so much.

[@graybenchelec]

Nice!

[@Stu66orn]

@@graybenchelec Could you test diodes you have with originals in ktr if there is sonic difference with similar curves?

[@graybenchelec]

@@Stu66orn Hah that would be the test wouldn't it? The right way to do that would be to pull the original diodes of of the KTR and install sockets, then swapping in and out. Unfortunately KTRs are crazy expensive and my desoldering skills are sub-par.

[@Stu66orn]

@@graybenchelec Yes it's true. Maybe I will try a Klon board. Managed to get cheap over 1000 D9E diodes.. Just want figure does DCA 75 curve have anything to do with the sound. Maybe it just distorts sooner that's all.

[@starshineraiser6729]

This is so fun. I’ve had/have lots of Klones and so far, there are two that are miles ahead, for my liking: Decibelics Golden Horse and the Ceriatone.

[@renatonyc]

LOVE my Decibelics! I have 3 KTR (2010, 2018 and 2020), a silver Archer, a Ceriatone Centura. And had a Nordvang ‘83 Drive, a JHS Soul Food and a few other klones. Also have access to a Silver non-horsie Centaur. Decibelics is still the all around best.

[@toneelectronix]

Your final thoughts are spot on! May I add that there is another factor in play which is the innerent junction capacitance of each diode type. Still something to be aware and that I am also wondering about how it affects the overall sound. Maybe that's why equal traces does not mean equal frequency response - hence different sounding.

[@big0bad0brad]

More junction capacitance would probably kill off some of the high end

[@tiborbogi7457]

Thank you for putting down my touts. ;-)

[@bluesifyme]

Great video - amazing nerdy stuff, and the amount of effort put into it is impressive! However, from my experience with the D9 family of diodes, there’s a huge variation between them. D9E can vary from high 300s to mid 500s, so a single D9E is not quite representative; Don’t know about the others, the US produced diodes seem o be more consistent.

[@AH-zm2me]

Those PEAK Electronic testers are fantastic. Can't recommend them enough. Great video as always!

[@graybenchelec]

Thank you! Agreed, use it all the time.

[@drewg3087]

I think the new 2020/2021 KtR is using the Russian D9e. I finally got one after waiting two years and was pissed it had the new diodes. But might not be that big of a difference if I don’t almost max out the gain. Thanks so much!

[@leeroyfunk]

Thanks for this, really good to see some actual scientific rigour in your research into the magical/mythical klon diodes: It seems that the received wisdom that "authentic"/"typical characteristic" 1N34As will be close enough to the diode klon specs seems to be borne out here (I know that you've referenced/queried soulsonic's testing methods, but I seem to recall he suggested 1N34As or similar would probably do).

[@graybenchelec]

For sure, there's nothing magic here. Klon diodes measure as pretty typical germ diodes.

[@PaulPauliePabloPaolo]

I put a pair of D9E's in my JRAD Silver (Silicon) Archer. It has been revelatory. I did like the sound before and for 'clean boost' stuff it was great. But the D9E's sound fantastic to me, just the right amount of compression and clipping. Thanks for taking the time to put this together, this is still a fascinating subject however well trodden. Now I just may have to get one of those meters and a couple of 1n34a's haha... here we go again...

[@graybenchelec]

Thank you! Glad the D9E are working out, if they sound right that's all that matters.

[@edcollins9377]

Thanks for the info MrPaolo. Do you remember the Vf value for the D9E's that you used????? I've got a bunch of D9E's that range from 0.30V to 0.36V. I was planning to try a pair of D9E's with Vf= 0.30V, but I also want to try some 1N270's. I have 2 pairs pf 1N270's with Vf=0.26V, 2 pair of Vf=0.27V and 2 pair of vf=0.28V. I'm also wondering if I should use matched pairs (matched as far as Vf is concerned) or should I make them slightly asymmetrical. I also have several pair of 1N60's with Vf=0.25V. Measurements are good, but in the end, the ear is more important than the meter!!!!

[@ArdiNugrohocilacap]

In my experience, using matched Will make better ​@@edcollins9377

[@spencerleehorton]

I made my own and will check the diodes now!! Mine sounds great by the way so very interested in the results

[@chippsterstephens6800]

That peak is such a great tool. I use it for all kinds of tone tools….

[@samaitcheson7057]

Thanks Joe. Another great video. Subbed.

[@graybenchelec]

Thanks!

[@jimismith6648]

Thanks again Joe, awesome and accurate content as usual.

[@graybenchelec]

Thank you!

[@HotPink90sSun]

Enjoyed your work here -

[@graybenchelec]

Thank you!

[@recklesstoboggan]

Diode matching master! And by eye no less. Dang. And, not that an hour and a half (including the last video and this one) Klon diodes isn't enough, but it would have been cool to hear the difference between the KTR diodes and the mystery diodes that you identified by eye.

[@graybenchelec]

That may be an upcoming video, who knows. I can almost guarantee you will not hear a difference, and not because it's the "same" diode or whatever, but because the differences are so minute for most germ diodes.

[@andymouse]

Still more ! still fascinating, what's the funky noise on the traces from ? are you sure the blue trace IS a diode LOL ? I love all the nerdy dives into stuff that's a bit subjective which is 'Audio' another great and well presented video....cheers.

[@raymondcarroll58]

Geranium is a flower not an element. Sorry about the trollie comment. Great content and very well presented.

[@davidblake6889]

I found your analysis absolutely fascinating. As an ex electronic engineer, what I really cannot get my head around, is your throw-away assertion that diodes with the same curve trace can sound different! How, for Heaven's sake? If the diode has exactly the same non-linear characteristic as another diode, then the distortion it produces and the sound it makes will be the same. If it is not, what makes it different? Someone mentioned differences in junction capacitance as being a possible cause. Germanium diodes are point junction devices, so the junction capacitance is incredibly small and so any differences in capacitance between different diodes are orders of magnitude smaller still. At audio frequencies the effect of this capacitance is negligible. The phase differences would not be detectable. To find out for certain, put the output signal on a spectrum analyser and see what the amplitude of various harmonics actually are for each pair of diodes under test. A spectrum analyser cannot lie. A human ear probably can.

[@voxpathfinder15r]

I told a few guys once that the power section of my solid state Vox pathfinder 15R amplifier has the identical flat top clipping on my oscilloscope as my all tube Carvin Belair amplifier. And they read me the riot act about how an oscilloscope isn’t the entire story blah, blah, natural warm tube compression, blah, blah, feel under the fingers, blah blah blah. Nobody can tell me the piece of laboratory equipment that can show me the so called differences. I know in the field of physics that I come from - any and all phenomena that is researched and then written about has to be measured on some piece of laboratory equipment.

[@al271987]

I have an Emerson Goldenface and when I took a look inside I couldn’t easily identify what the diodes even were, yet it sounded as good as any of the other half dozen or so klones I’ve had. The Chellee Ponyboy puts a switch to go between Ge and Schottky diodes and at full gain I was flipping the switch and the differences were nearly impossible to discern to my ears. The community has made sure there’s merit in looking closer at this, but all the diode talk really seems like extreme hair splitting to me. A good klone is a good klone and at this point there are dozens available. The overall circuit is what’s important, the individual parts aren’t that big of a deal if they are designed to work within the circuit well.

[@edcollins9377]

I beg to differ. The quality of the parts and electronic components are Everything! That's why boutique pedals almost always sound much better than those clones that are massed produced or those that rely on cheap Chinese components. I've always used top grade components. Many years ago, when I first started building my own gear, I was trying to CLONE EVERYTHING. I figured if I used cheaper hardware--like jacks, pots, and switches--I could afford to build more pedals. Not a bad move at the time, because I was able to build 14 or 15 pedals for what it would have cost me to build 10 pedals had I used the more expensive, higher quality hardware. However, it didn't take too long to discover that the cheaper jacks and switches don't last as long. Furthermore, it takes both time and money to replace damaged jacks and switches. With hindsight, I wished I would have used all Switchcraft jacks ... at least on the pedals I use the most I do concur that a person can get a bit hung up on specs. When all is said and done, we can't forget that our EARS are more important than our MULTIMETERS! :) In addition to buying good hardware, here is another good tip for newbies: Be careful who you buy your IC chips from. I prefer to buy my electronics from respectable companies like Mouser, Digitech, Newark, etc. If you go to ebay, you can often find IC chips from China and other parts of Asia that are much, much cheaper, but I know for a fact that many of these sellers will often take similar, yet inferior IC chips, and sand off the markings. They will then print new markings on the chip making you think you bought an OP275 when you actually bought a cheap MC1458 or worse. These bogus IC's not only sound inferior (the difference can range from hardly noticeable to massive), they don't last as long either. (I had a friend who bought a tube of 25 bogus CA3080's and he was replacing them constantly.) Also when these fake chips get hot, the markings will often crumble and/or fade. When you go to replace them, you better hope you have a schematic, because you can no longer read the marking on the Ic chip. Sometimes the markings fade after just handling the chip a few times. So here's the deal. Once upon a time a father by the name of Mike Brady once gave his son Greg aka Johnny Bravo some excellent advice. The advice consisted of only 2 words, but those words were Latin: "Caveat Emptor"!

[@rgvt5638]

How about the 1N270 🤔 - I have a Klon kit coming my way and although it’s 1:1 with the 1N34A’s, I have some misc diodes around that might be worth experimenting with. Thanks for the info and inspiration!

[@graybenchelec]

I'm sure they'l work fine. What I hope you take away from the video is that a large variety of germ diodes, includes different packages/makers/part #'s all still fall in a pretty small band of Vf curves, and the Klon diodes are totally in line with those.

[@rgvt5638]

@@graybenchelec totally, the most notable thing for me was seeing the the D9E being a mellower slope (for lack of a technical description) and still sounding exactly like the others despite what it looks like graphically.

[@andywander]

I wonder if the KTR diodes would look any different if they were not measured in-circuit?

[@florabee9283]

I have found that some glass-cased diodes are photosensitive and behave differently when I have my bench light on vs off.

[@big0bad0brad]

Yeah they will likely be responsive to anything infrared and shorter wavelength. For example, silicon diodes actually emit and react to infrared light but they tend to be in an opaque case so you never notice it. Germanium (I have to assume) should be the same but longer wavelength. Above the silicon forward voltage you have shorter wavelength IR LEDs, then red LEDs, orange, yellow, green, blue, UV. If you hook an LED up to an opamp buffer and then measure the voltage you can actually gauge the light it's exposed to.

[@meesterdinglefritz2064]

The diodes you mentioned around 4:00 in look a lot like the 1n34a’s that StompBox currently has in stock.

[@graybenchelec]

Sure do, although the description on that product says the bands might be different color 🤔

[@adeptgopnik]

right, and the tolerances of all the parts combined in the circuit ca nsway the sound of any pedal one way or another. Which is why it is nice to build your own since you can tune it to how you like it to sound. Bill Finnegan himself even says that the tolerances in his circuits and even pots can vary from one to another. Theres no way of knowing you will for sure like a pedal without trying it yourself and feeling how it works with your playing and amp.

[@jwdory]

Great video! Would you mind sharing a link to the ebay seller with the Unknown Ge diodes? Those looked like the diodes with the closest iv characterictics. Thanks

[@graybenchelec]

I actually don't remember who it was, sorry! But I do remember asking they had more and they said no :/

[@TheTubeDude]

Since the ultimate information sought is about the guitar signal and how it's affected by the magic diodes, I would do a signal test analysis comparison of the special diodes and other Germanium diodes. (You have mentioned that there are two types of diodes. There are at least six types used in solid-state diode history; I'm old enough to have worked with Selenium and Copper-oxide diodes, but not Galena) Static voltage, current, susceptance, conductance, admittance data is nice and fun but doesn't get to the heart of the matter. What is the difference in the passed signal and how does it sound? I very much enjoy your videos.

[@graybenchelec]

Very true! I've even worked on some old tube amps that have selenium rectifiers. In the scope of clipping diodes for pedals however, silicon and germanium are far and away the two most prevalent options. The ultimate test is really the subjective experience of listening to the different diodes and picking a favorite. Unfortunately, no amount of high end test equipment can give us predictable results on that. Maybe some day :)

[@TheTubeDude]

@@graybenchelec Hi, and thank you for the very nice reply. I agree with the ultimate test. BTW I made a pedal with a Selenium clipping rectifier in the case. It takes two Roadies to transport it. -- Please keep up your fine and entertaining work.

[@themaninthesuitcase]

I have 2 builds, one with D9E and another with the "1N34A" that looks a lot like Schottky diodes. I prefer the D9E I think but need to properly A/B them still. How would the "looser" D9E present, I guess at any given gain setting you'd have slightly less clipping? I wonder if the pot tolerance would even be enough to make that not matter! Would be interesting to see the curve on the knock off diodes as a comparison and see if it's just bias.

[@EversonBernardes]

D9E would present at being slightly louder, slightly less clipping and slightly darker clipping character, as the more gradual transition into non-linear behavior would reduce the production of higher order harmonics. The differences would be _tiny_ mind you.

[@jscruffins8767]

sorry but i have to ask - what current do these pass in circuit? ive breadboarded a few circuits that dont even pass 10ma total so does this curve actually affect performance?

[@cimmerian100]

I noticed you traced the Klon diodes while in-circuit. Would that have tainted the results do you think?

[@POGFROGULPOP]

In my case it did, especially on one diode. Using my Peak DCA 75 Pro at 5 mA, the readings I got were 0.437, and 0.435 volts whilst in circuit. Then once removed out of circuit, I got 0.435, and 0.367 volts.

[@sgtrutters5892]

Great video. I want to get one of those DCA 75s! Just curious and it may not even be an issue, or if it works that way, but do we know how much current is being driven through the diodes in the circuit?

[@MrCacciLLo]

graphs are Current vs Forward Voltage

[@Annunaki_0517]

Did you test more than one of the grey band diodes? In case that one you checked was faulty and leaking voltage?

[@graybenchelec]

I have yes, they're all similar curves.

[@Route66Wanderer]

Unfortunately your DC test of the diode’s curves meant nothing. The diodes in the Klon are used as AC signal clipping diodes. The function you want to see is the AC signal starting to distort as it starts nearing clipping.

[@iblesbosuok]

Come on guys! It's about personality. Be yourselves. Choose the gears that best suit your natural personality. No need to be fancy or expensive. Choose gears that make your heart comfortable. Choose the one that suits your soul. Once you get the gears that match your natural personality, customize them to suit your spirit even more. No need complicated knowledge. No need for confusing scientific criteria. No matter how high a scholar is, it is impossible to describe your personality. Just rely on your feelings. Whenever you find a component that you feel good about, grab it. Try and try the components. If necessary, also try other types of components such as N-ch JFET or N-ch enhancement MOSFET to replace it. Those ones still potential to get the result of customization that is in harmony with your soul. Art for art. Prioritize your natural personality. Rely on your heart. Rely on your soul.

[@RobertNolan]

I do try to check to make sure it reads around .35 but also I've always felt that to be an incredibly wasteful process - both in time and of materials. Most will sound exceptional, but you end up getting a bad sounding one here and there. The ear test is very important.

[@geezberry8889]

where can i buy your unknown diodes?

[@carsonyuh]

Give us the link to the unknown ge plz

[@ЛисёнокФенек-ч1ч]

Hi! To match two clipping diodes, is it possible to judge by the resistance of the pn junction, or do you need to look at their volt ampere characteristics with a special tester?

[@graybenchelec]

Don't forget Ohm's law. Resistance, current, and voltage are intimately related. By looking at the volt ampere characteristics, you are indirectly looking at resistance of the junction. You can use a simple multimeter to test the forward voltage, which will give you a "snapshot" of the junction resistance at a certain test current. The curve traces are just a bunch of snapshots lined up and put on a chart.

[@chrisquinn9104]

Have you compared these to BAT diodes. They seem to be used in a lot of great pedals for low forward voltage, including the Greer Southland. I’ve been using them in builds as well. I just have no experience doing an apples to apples comparison with germanium with the same forward voltage Thoughts?

[@graybenchelec]

I've been adding different diodes to that google sheet in the description. I should do the BAT41, good suggestion.

[@chrisquinn9104]

@@graybenchelec I haven’t been shown any empirical evidence or a good example of why one would use a specific diode other than for the forward voltage number. It leads me to think the Klon “special diodes” are hype, but I’m still open to seeing some evidence to the contrary. I mostly want to know because I built pedals and it would help me select diodes. A few BATs would be good

[@big0bad0brad]

@@chrisquinn9104 The only other parameters that in theory could matter are parasitic inductance (this will be so low that it doesn't matter at all), or junction capacitance (intuitively this feels like it should be too low to matter for audio circuits but if you find that some diodes lose too much high end, this would be your culprit).

[@edisonmanfroi4957]

I had a KTR. I thought it was a pretty good pedal. I made several KLONES that sounded identical to the KTR. However, I am not satisfied. For me, the diodes in the KTR are no longer the same as those in the old KLON CENTAURs. Since I never had access to the old KLON CENTAUR pedal, I still believe that I have not reached the end of this analysis. I have not yet seen anyone present the DCA75 curve for the original diodes of a KLON CENTAUR. If anyone has a KLON CENTAUR, please find a friend who has a DCA75 and draw up the curve and show it to everyone, please. I believe that this could put the end to a series of speculations.

[@francisco.amorim]

Waiting for your klon clone offer for a more affordable price than a ktr or a ceriatone clone 😅

[@graybenchelec]

EHX Soul Food is a good pedal :)

[@TomandTheBrokenHearts]

Hi, are you selling the ge diodes that test like a Klon?

[@graybenchelec]

I'm not no, just keep an eye out on ebay for germanium diodes 👍

[@Evan-hm7tz]

if it looks like a duck and quacks like a duck, its a fuckin duck.

[@yvanflodin]

aren't these diodes supposed to be 1n34a's? you can still find a lot of them , i don't understand how one guy could own all the remaining stock?

[@graybenchelec]

That's sorta the trick, Bill believes that the specific 1N34A he has/had are different than other 1N34A, and are crucial for getting the sound he wants from the klon.

[@yvanflodin]

@@graybenchelec not impossible since 1n34a's can vary, at least they do nowadays. I myself bought a bunch over the years and not all are the same, some have different colour ring, etc. No idea if that was always the case or if sellers just understood it's a very desirable reference and sell anything under the same name? ... Still, the fact that one guy owns them all sounds like desparate gatekeeping to me haha

[@woosix7735]

That box of diodes is probably worth, … I don’t even what to know

[@Zardox-The-Heretic-Slayer]

if you think that's impressive, you should see the amount of diodes I have!

[@qddk9545]

Snake oil, fairy dust, marketing BS. He needed to have something magic to hang the much to high price tack on. I suppose that different caps in the signal path would make a much bigger difference. And what about different brands of TL072, they probably don´t sound the same either.

[@1pcfred]

The design is what justifies the cost. R&D isn't free.

[@A.P.O.L.L.O.1.3]

DUDE what happened to your fingernails ?

[@steelydan146]

This klon business is a total scam. All the good copy pedals like from J. Rockett, Wampler, Way Huge, MXR, etc.... they all sound the same!

[@ciaranjmurray]

Do they? I've only ever had the MXR Sugar Drive and I like it fine. But I have heard it said more than once that it goes into higher gain territory than the Klon/most klones. Any opinions on this?

[@MrCacciLLo]

@@ciaranjmurray I only tested the EHX Soul Food and the J Rockett Archer side by side, and they both sounded the same to me and my friend. We couldn't find any difference, or better, any relevant difference.

[@1pcfred]

@@ciaranjmurray generally what we are most sensitive to are changes in volume. With more being perceived as being better. It goes to 11.

[@das250250]

All great nerd food but if you can hear differences between these diodes you are simply fooling yourself .. then add that you use this through a bunch of pedals and amps and rooms and crowds and pa mics ...

[@Zardox-The-Heretic-Slayer]

what floors me about this whole thing is that he spends all that time designing a really nice overdrive with a reasonably rare addition (for its time) of a voltage doubler, he also used a novel gain stage which was only really used in amateur radio and other niche areas and followed it up with treble stage that almost acts like an exciter which again, no one else used in overdrive pedals and what does everyone (including Bill) focus on? the tacked on diode clipping stage which people can even barely tell is missing when you remove it