The Klon’s Mid Sweetener and Volume

I’m constantly A/Bing klones I’ve built with my Centaur, and sometimes each other. Usually, I can dial them in to sound identical, regardless of their specs. Sometime I can’t, and a day later I’ll come back and I can.

Lately I’ve been finding that when one unit sounds different from another, the cause is almost always the Output knob. Just the subtlest difference seems to sound not like one unit being louder than another, but fuller and ballsier than the other. If I have two units built to the same specs, and one seems to have more bass/lower mids, all I have to do is turn down that unit’s vol just the tiniest bit and they sound the same. Or increase the other unit just ever so carefully. The effect is the same- with the “louder” unit, I hear more of the frequency spectrum rather then just an increase in amplitude of the frequencies that are already present. This seems to happen at any volume range- I have this issue at room volume and when cranking it. I think this is a very large part of why we perceive unit-to-unit differences. Well, aside from hearing with our eyes. I’ve also tried it with old TS builds and it also accounts for the differences I used to hear.

The volume issue has been masking the impact of the Klon’s mid-sweetening resistor for me.

The most tonally significant difference between the first and last versions of the Centaur is a resistor labeled R15B that Bill added ” to give the circuit a very small amount of additional low-mid response”. We now know that this was added between the 2nd and 3rd revisions of the PCB.

I’ve know about this resistor for over 3 years now. Until a week ago, I didn’t think I could hear a difference whether it was in the circuit or not. I had even put it on an on/off/on switch that could swap it out with a pot to mess around with lowering the resistance. The effect is very subtle until you start getting lower.

Last week, I was messing around with a test unit that has this resistor on a plain old on/off switch. I noticed that with the resistor switched out of the circuit, the test unit was actually ballsier sounding than the real thing. That can’t be right. While barely perceptible, that resistor is supposed to add lower-mids. If anything, the real Klon (a v3) should be ballsier. The units’ volumes sounded identical, but the real thing was lacking on the lower end. So with the volume issue in mind, I switched the resistor back into the circuit. The test unit was still ballsier, just enough that I could identify which unit I was playing through while having a friend switch them in and out of the loop. Then I tweaked the volumes to where I couldn’t tell he was stepping on the looper anymore. The volumes still sounded identical, but they had equal fullness as well. Both pedals had the resistor in play and were indistinguishable.

From there, I switched the resistor out again on the test unit and compared. The difference was only apparent on the lower strings, at least in standard tuning, particularly when palm muting. There’s just a very small amount of extra meatiness there.

The problem was that the resistor adds so little that the minutest difference in volume can completely hide it. Without actually seeming louder, the “louder” unit will sound ballsier than a unit with or without that resistor.  So in the past, when I was comparing a unit with R15B to one without it, I was compensating for the subtle loss in low-mids by tweaking the other unit’s volume and not realizing what I was doing. R15B’s toneshaping can be completely drowned out by just the lightest touch on the Output pot.

Timmy Versions

IPaulC has been pretty open about the Timmy circuit and its history. But all the info on it is spread out between various threads on TGP and the DIY forums, and some circuit changes seem unaccounted for. So here is my attempt to make sense of it.

The earliest version of the Timmy is built on protoboard. The most commonly posted schematic seems to match this version.

 

This version has 6 visible diodes with two of them on a DIP switch and most of the parts from the schematic accounted for.

When PaulC switch to fabbed PCBs, he retained the same basic component layout. Let’s call this V2.

• The silver mica cap is now polystyrene
• where is the 6th diode?
• Standing resistors are all now flat, 2m2 drop resistor looks like it’s been moved onto the PCB instead of hanging between the 3PDT and sleeve of the input jack

 

Next, the layout changed. We’ll call this V3.

• Diodes are now on a SPDT toggle switch, 6th diode is still unaccounted for
• Additional 4n7 cap to right of IC
• Addition 33k resistor under 1uF cap
• This unit has an electrolytic 1uf instead of film

 

 

Using the same basic layout as V3, V4 adds more holes for components and some labeling.

• 33k resistor has been changed to 12k
• Two of the 3k3s have been changed to 10k
• 1k5 is now 1k
• 4n7 has now been changed to 2n2

 

Later V4s have…

• LM1458N op-amp
• Small cap from input to ground, Paul says this is 47pF (http://www.madbeanpedals.com/forum/index.php?topic=23478.0)
• Input cap changed from 47n to 39n

 

 

According to Paul:

“The brown cap is a 47pf to ground on the input.  That was put in due to some guys getting radio.  Small enough to not really effect the audible high end, but large enough to cut down on radio noise.  The input cap was changed from 47nf to 39nf. That changes the low end roll off from -3db@6.3hz to -3dB@8hz which really is no change.  That was done years ago when there was a shortage of 47nf for a few weeks from all my vendors.  Since it made no difference in performance I just stuck with the 39nf which I used in another spot.   The output stage had it’s values changed from the stock 2x 3k3 w/4n7f setup to 2x 10k w/2n2f.  This was done to get a little less loading on the 2nd stage to help the headroom a tiny bit with a weak battery.  The 1k5 preset resistor in the treble control was changed to 1k to bring the brightest setting back to where it was with the 3k3 setup.  You would have to squint your ears to really notice any of these changes.  What can’t be seen is the treble control was changed to linear to cut back on the dead spot between 7-10 with the audio taper.  That one had a big effect on knob settings.  I’m not sure, but it probably has the volume pot tweak which had the resistor soldered on the pot, and then it was added later to the board which came after the version shown.  I never promoted those pot changes because they don’t effect the sound – just smoother knob settings and I didn’t want it to look like I was trying to resell a new version to people.” (http://www.madbeanpedals.com/forum/index.php?topic=23478.0)

• So at some point in the V4s, the treble pot changes from 50kA to 50kB.
• I can’t account for the resistor he says was on the volume pot but moved to the PCB. It could be one of the 3k3s, if you have it connected after the 1uF output cap instead of before.
• Paul doesn’t mention the 33k that is added and then later changed to 12k. Maybe this is the one that was on the volume pot.

 

This post:

“Well – there’s even a newer version… For about the past year I changed the volume pot value and soldered a resistor on it to smooth it’s taper. In the old style unity gain was about 2 o’clock with all the volume boost going on in the small area between 2 and 5 o’clock. That always bugged me, but using a linear pot was worse – that would put unity at around 9 o’clock making it a useless control imho. With the gain at zero the pedal puts out 12dB, so changing the volume to 25ka and adding an 18k resistor from input to output fixed this for me. I calculated the 18k value to put the control -12dB at 12 o’clock making it unity at that setting. Like the treble pot change this was about smoothing the taper and not about changing the sound. It doesn’t sound different – the min and max settings are the same. Just the settings in the middle are tweaked to not be so touchy.

“About two pcb batches ago I added this resistor to the board because it was a pain to solder them on to the pot. I didn’t call this a new version because I don’t want people to think it sounds different enough to get them to buy it again. The sounds are all there – just at a little different knob setting. These are all pretty much mechanical tweaks and not tone tweaks

“Another change that ALWAYS happens is the resistor that feeds the LED. Those blue leds are all over the map when it comes to brightness, so in every pedal I have to adjust the value. I have a jig made with a dozen or so resistors in it. I plug in an LED and move it around to find the value that’s not to bright or to dark. Then I put that value on the pcb with that led. This does not effect the tone – just the brightness of the led.” (https://www.thegearpage.net/board/index.php?threads/timmy-version-question.1444785/)

• Based on the above, I would bet that the 33k/12k resistor in the photos is in the same place as the 18k Paul mentions here.

 

 

Klone Models

I get a lot of questions about the differences between the klones I build, and whether or not I’m the same person here as I am on Reverb. So here is a handy run-down of each klone I make.

Clontarf Lapith

Specs: 1994

Enclosure: 1590XX (aka 1790NS),  145 x 121 x 40mm or 5.7″ x 4.7″ x 1.5″

Power Supply: 9v DC, Center-negative

Battery: Includes 9v battery holder

Footswitch: Carling DPDT (sometimes BLMS 3PDT)

Resistors: 6 1% metal film, rest 5% carbon film

Capacitors: Film for pF and nF values, electrolytic for 1uF+

Pots: Wired

PCB: Reproduction of the Centaur’s PCB

Diodes: Magic

 

Hylonome Klone

Specs: Post-1994

Enclosure: 1590BB, 119.5 x 94 x 34mm or 4.72″ x 3.71″ x 1.33″

Power Supply: 9v DC, Center-negative

Battery: Includes 9v battery snap

Footswitch: BMLS 3PDT

Resistors: 6 1% metal film, rest 5% carbon film

Capacitors: Film for pF and nF values, electrolytic for 1uF+

Pots: PCB-Mounted

PCB: Reproduction of the Centaur’s PCB with the daughterboard removed

Diodes: Magic

 

Klone (Old Model)

Specs: 1994 (Sometimes Post-1994, check listing for details)

Enclosure: 1590BB, 119.5 x 94 x 34mm or 4.72″ x 3.71″ x 1.33″

Power Supply: 9v DC, Center-negative

Battery: None, requires external power supply

Footswitch: BMLS 3PDT or DPDT

Resistors: All 1% metal film

Capacitors: Film for nF values, electrolytic for 1uF+, cermanic for pF due to size constraints

Pots: PCB-Mounted

PCB: Heavily modded version of Madbean SunKing

Diodes: Magic

 

—

The post-1994 specs are supposed to add lower mids, but I don’t hear any difference. I think a difference in resistor value could have been greater, so my own advice is to not worry about it. The post-1994 circuit also includes a little bit of extra switch popping protection, but it’s mostly overkill.

The Lapith is physically larger than the Hylonome and Old Modem Klone. It’s about the same size as an actually Centaur, while the latter two are the same size as a KTR oriented sideways.

All three pedals take a typical power supply (9v DC, center negative). The Lapith has enough space inside for a nice littler battery holder. The Hylonome has a baterry snap, but no holder. It might rattle around in there a little, so I recommend a bit of foam or tape to hold it in place if you’re going to use one. The Old Model Klone wasn’t designed to take a battery, and older units don’t have to correct type of jacks to allow a battery snap. More recent ones do, but I don’t add the snap unless someone asks for it. A battery may fit, bit risks damaging other components or shorting things.

The Lapith usually has a Carling DPDT footswitch, the same model the Centaur uses. Some units have a BMLS (Bitches Love My Switches) supplied 3PDT when I couldn’t find a reliable source for the Carlings. The switches can easily be swapped out on the Lapith if you want to upgrade, but not necessarily on the two smaller klones (not enough room inside). The Hylonome always has a BLMS 3PDT. The Old Model Klone used to have an X-Wing style DPDT for neater wiring, but I found that they had a high failure rate and now I use 3DPTs from BLMS in these too.

The real Centaur has a combination of 5% carbon film and 1% metal film resistors. I duplicate this in the Lapith and Hylonome. For the Old Model Klone, I just use all 1% metal film. In theory, metal film resistors are more consistently in-spec and less susceptible to noise. In practice, there are no audible differences with stompbox voltages.

For capacitors, the real Centaur does not use ceramic caps for pF values like most pedals do. Instead, the lowest values are film. The Lapith and Hylonome also use film. The Old Model Klone’s PCB doesn’t have space for larger film caps, so I use ceramic. I don’t think they sound different.

The Lapith’s potentiometers are connected via wires. This also allows the PCB to be mounted to the enclosure with standoffs. The Hylonome and Old Model Klone have PCB-mounted pots due to the tighter spacing inside the enclosure. Wiring the pots takes longer, but is less likely to break and easier to repair. Board-mounted pots look neater and are quicker to assemble, but would take more work to fix if they started to break at the solder joint.

The Lapith’s PCB is a reproduction of the Centaur’s. The first few Lapith units made use of the daughterboard that attaches to the Gain pot, but I thought running wires directly to the pot and using the daughterboard as a place to add an extra standoff made for a more stable build. All Lapiths after unit #010 are built this way. The Hylonome’s PCB omits the daughterboard to make it small enough to fit inside a 1590BB enclosure. Anything to the left of the daughterboard has been left intact. The Old Model Klone is built on a PCB that I based on the Madbean SunKing. I had built two klones with the SunKing PCB, but was not happy with the location of a few solder pads and traces, so I redrew everything to fix these issues. After a few dozen builds, I made a few more changes to fix common issues (potential solder bridges), and adapted it for PCB-mounting the pots.

All three klones come with the “magic” diodes. These are the most important component. Varying diodes are why so many different klones on the market sound different, particularly when they’re using current-production germanium diodes from China which are pretty inconsistent.  I bought as many of the magic diodes as I can find and should have enough for another year or two of building. Until I run out, all of my klones will come with these stock.

---

 

The TL;DR version:

The Lapith is the most accurate klone on the market, but is a bit too large for some guys’ pedal boards.

The Hylonome is smaller and only less accurate in that the PCB needed to be altered on one side to fit the enclosure.

The Old Model Klone is less accurate in that it’s built on a PCB that ISN’T a repro of the Centaur’s with less emphasis on choice of component type, though the circuit is the same. Same size as the Hylonome. The build style lets me put it together much more quickly, so I can offer it for a lower price.

 

 

 

 

 

 

 

 

A ManticoreFX Website That Is Not Mine

http://www.manticorefx.com/

That is not my website. I have no affiliation with this fellow.

Klon Buffers: They’re All Wrong

Why is it that, when it comes to the Klon circuit, misinformation is the norm? Every so-called Klon buffer schematic and layout I have seen on the net is wrong. And every single stand-alone Klon buffer seems to be built using those schematics and layouts. Did it never occur to anyone to cross reference the build docs with the actual Klon schematic?

Let’s talk about the real Klon buffer. It runs off the Klon’s 9v supply, which has a 47uF filter cap and a polarity protection diode to ground. Vref is made with two 27k resistors that form a voltage divider. The output cap is 4.7uF. All the supposed Klon buffer docs and builds use 100ks for the voltage divider, a 1uF output cap, and omit the filter cap and diode from the power supply. Ok, so the end result is the same. You get the same Vref, and the output cap is still large enough to make no real difference. We can live without the extra stuff in the 9v section of the circuit (but it’s a good idea to have them when the circuit is built on its own). Without the specific values, it’s not the Klon buffer. It’s a regular old op-amp buffer.

The first time I ever saw a schematic for a standalone Klon buffer was on Beavis Audio’s site (no longer active, but available cached here: http://web.archive.org/web/20110713081503/http://www.beavisaudio.com/techpages/Buffers/)

I have to assume that either that schematic swapped some parts’ values for more common ones to make it easier to build or that Beavis was looking at another schematic that had. Also, since buffers are typically incorporated into other builds, there’s no reason to include the extra parts stuff on the power section, since whatever you’re building the buffer into would already have that. But all it takes is one incorrect schematic and people are making layouts to share with other DIYers, then enterprising builders take that information and use the word Klon to market their builds, built exactly as the incorrect schematic shows. This tells us that they don’t really know what they’re doing and are more interested in the word ‘Klon’ than finding out if they’re building the right thing.

 

Here are more Klon buffer docs. All wrong. If you’re going to call it a Klon buffer, at least get the 27k’s and 4.7uF in there.

http://tagboardeffects.blogspot.com/2011/04/klon-buffer.html

http://turretboard.knucklehead.dk/2011/10/13/stripboard-klon-buffer/

http://revolutiondeux.blogspot.com/2012/02/klon-buffer.html

http://shop.pedalparts.co.uk/Klon_Buffer/p847124_6346591.aspx

http://147pedals.blogspot.com/2010/11/plon-fentaur-buffer.html

http://effectslayouts.blogspot.com/2014/11/klon-buffer.html

Bill reveals his diodes, and the myths won’t die.

Rockett and PGS release the Ikon, a klone supposedly based on the Gold Klon instead of the Silver. Someone posts gutshots and we see the the Ikon has D9Es. Bill chimes in and mentions that his diodes are 1N34As. Barely anyone notices. People keep repeating the same BS.

http://www.thegearpage.net/board/index.php?threads/rockett-archer-ikon.1587262/page-5

The diode I have always used is a germanium diode with the part number 1N34A, but you should understand that this particular part has since the 1950s or so been manufactured by literally hundreds of different companies, and having listened to as many different ones as I have, I can say with confidence that they all sound somewhat different in my circuit, and often they sound VERY different.

This is surprising, because 1N34As are the most commonly used diode when germanium is called for in a pedal. No one expected it to be what everyone is already using. But there are some important points to keep in mind.

• When sousonic listed the diodes as 1N34A in his schematic of S 698, he has no way of knowing that that’s what they actually were. He was just suggesting the most readily available germaniums for clippers.
• For any part other diodes, I would probably say that brand doesn’t matter. The same parts from different manufacturers should be electronically equivalent. ‘Electrons aren’t bullshit dependent,’ as I’ve seen it written somewhere.
• Most DIYers and buyers of klones with 1N34As have reported that they sounded different than the real Centaur, so we had all reasoned that Bill’s diodes ARE NOT 1N34As.
• My experience is that germaniums of the same part number can sound a bit different not only across brands, but also within the same batch (presumably made in the same production run). The fV can even be about the same. I deliberately stock up on cheap Chinese-made 1N34As that are particularly inconsistent so that I can pick out the ones that match my Centaur and KTR. Batches that are more consistent seem to fall into the ‘less Klon-like’ camp.
• It’s possible Bill is lying to throw us off, but even if that’s the case, he’s still correct about brands sounding different. I do think they don’t sound AS different as he’s making them out to be, and with some auditioning, it’s not hard to find a pair of diodes that nails the same kind of clipping as the Centaur.
• Tolerance and drift are still a thing. Always remember.

The Rockett Archer Ikon

What a bunch of crap: “Features actual Klon “magic” diodes. That’s right—THE part”

Nope, the Ikon has D9Es. Just look inside one and see. This isn’t marketing BS, this is a flat out lie. D9Es are old soviet diodes. There’s no way Bill had access to them in 1990. And they’re easily identifiable by their blue and red bands. Bills diodes have a single black band marking the cathode (soviet manufacturers marked the anode, or both).

BYOC used the D9Es in the Silver Pony kit that came out last year. There was a lot of hullabaloo over whether or not they were the same as Bill’s, and it was immediately obvious that they weren’t. The PGS and Rockett guys are a year late to this.

Gold vs Silver

Here’s an interview with CVT from Rockett: http://tonereport.com/blogs/interviews/prepare-for-launch-an-interview-with-chris-van-tassel-of-j.rockett-audio-de

“Even yesterday, as we were working on the gold version of the Archer with Pro Guitar Shop, we wanted to test it against a gold Klon. And the gold and silver Klons were different.”

You’d expect a manufacturer to understand component tolerance. You’d also expect someone who has worked with Bill to know that the guts of the Silvers and Golds are identical. Same PCB, same components, same diodes. The first few hundred (out of ~< 4000) Golds left out a few resistors, and maybe the first hundred or so Silvers did as well, but beyond that they all have the exact same specs. This is a myth that just won’t die, and I suspect CVT is using it to market the Ikon.

More Tube Screamer Stuff

I’ve been doing a lot of looking into Tube Screamers in preparation for a project thats’ tentatively titled The Omni Screamer. An expansion on the old MohoMods MultiScreamer. Anyway, here’s a collection of my notes in a somewhat coherent format.

 

I don’t know how official this schematic is, but it looks like a factory schematic and not one some internet dudes drew, so I’ll take it as canon.

 

The Op-Amps

Ibanez/Maxon hasn’t always been consistent with what op-amps they use in all the TS variants. The factory schematic for the TS808 seems to call for an RC4558 (Texas Instruments?), but they mostly shipped with JRC4558s. Some other op-amps show up too, mostly 4558s from other manufacturers. The TS9 was all over the place, ‘seemingly random’ according to Analogman. For the reissues, Ibanez seem to have settled on the JRC4558D for the TS808, and the TA75558P for the TS9. This is likely to justify the higher price point of the TS808 and keep it mojotastic. Ultimately, the JRC4558 seem to be THE Tube Screamer IC. Yeah, there were other 4558s, but current production models only ship with the JRC or TA75558P.

The Output Resistors.

The TS808 has a 100R series resistor and 10k shunt resistor at the output. The TS9 and pretty much every other TS variant have a 470R series resistor and 100k shunt resistor. Does it matter? Honestly, I’ve played with these on switches and hear no difference. I’ll take either pair. But this does help us classify the circuits. Since the op-amps do vary, we can define a TS808 circuit as one with 100R/10k and a TS9 circuit as one with 470R/100k. Because those are the only parts that differ.

This thread has some discussion on whether or not they matter: http://www.diystompboxes.com/smfforum/index.php?topic=108501.0

The Clipping Diodes

The diodes don’t get a whole lot of discussion, as everyone is focused on the op-amps and output resistors when they do their 808-spec mods. Most DIY clones use 1N4148s or 1N914s, which should be similar to what come stock in real TS’s. Early TS808s seem to have has 1S1588s (orange with a blue stripe marking the cathode), though the schematic I posted above calls for MA150s. Most of the other TS variants, including Maxon’s current offerings, use MA150s or something that looks the same, orange with a white stripe. The TS5 is notable for having 1N4148s. The diodes are hard to identify because they’re not labeled on the body. We have to go by what the schematic calls for, and we know that they didn’t always use the specified parts when populating the PCBs at the factory. Or we go by appearance of the diode and take an educated guess. So I’m going to say the definitive TS diodes are 1S1588 and MA150.

Does this matter? I’d expect 1N4148s to have similar, if not identical, clipping characteristics. This is something I’ll have to experiment with.

The Input and Output Buffers

All official TS variants have input and output buffers. Even the OD9, which is true bypass, maintains the buffers. Some clones strip these off, because they’re unnecessary with TBP. The TS808 and TS9 buffer configuration is made up of two discrete transistors. That means that they’re individual parts. A very early version of the TS808 (the narrow box) used both sides of a dual op-amp for the buffers instead of individual transistors. The TS10 has an extra buffer coming off the input buffer. The ST9 and STL both use one half of a dual op-amp for the input buffer and the other half for the mid boost part of the circuit, then the output buffer is a discrete transistor. For the most part, all official TS variants have the same stuff between the buffers. Even the TS808/9 output resistors are part of the output buffer.

1979 TS808 Narrow Box / Original Maxon OD808 / TS808 35th Anniversary Reissue

These three pedals are all pretty much the same. They even share the same PCB layout. The input and output buffers are both halves of a dual op-amp instead of discrete transistors. If Matsumin’s schematic for the OD808 is to be believed (and expanded to include both TS808s), they have a 220R series resistor where the later 808s have 100R, but they do have the canonical 10k shunt resistor. Everything inbetween the buffers is the same as every other official TS, save for the op-amps. The OD808 and ’79 TS808 had two MC1458Ps, and the 35th Anniversary has two JRC4558Ds.

TS9 / TS9DX / TS7 / TS5

The things that define a TS9 in comparison with the TS808 are output resistors. Other TS versions are essentially TS9 variants, sharing the same output resistor values. I consider the TS9 the branching point for most of the other variants.

The TS9DX is exactly like a TS9 when the mode switch is set to the TS9 position. The earliest run of these pedals had the drive pot connected incorrectly, resulting in less gain than a typical stock TS9, but this was fixed on previous production runs. It uses the same PCB layout as the TS808 and TS9, with an additional daughter board for the parts that the switch controls. It seems to ship with the TA75558P most of the time, but units with JRC4558Ds have been known.

The TS7 is a TS9 circuit with a Hot switch. Despite being lower-end gear than the TS9, they usually come with JRC4558Ds straight from the factory. It has a B20k tone pot instead of the usual W20k or G20k (same thing). The PCB is unique to the TS family, as the ToneLok series of pedals have lots of daughter boards crammed into the box, with way more solder pads than are necessary so multiple circuits can be built on the same boards.

The TS5 comes with the TS75558P. While most of the other TS’s have what look like MA150s for clippers, this guy has 1N4148s. The tone pot is G20k, surprising for a cheaper pedal. I also found that the 51pF cap is 47pF. Not a big deal, it was probably just what the factory had on hand. I have gone over it enough to find any other differences, but it’s pretty much a cheaper TS9 in a crappy plastic enclosure. Some very early runs had metal enclosures.

TS10

The TS10 has always interested me because, for years, believe my OD808 reissue was actually a TS10 circuit. Well, the TS10 has the TS9 output resistors. Sometimes it shipped with an MC4558 and sometime with a JRC4558. It also has an extra transistor, an emitter follower off the input buffer going to the bypass. In the effect portion of the circuit, the only difference between a TS9 and TS10 is that the TS10 has a 220R resistor in series with the input of the op-amp.

Maxon OD808 Reissue / OD9

The OD808 reissue has the TS9 output resistors, not those of the TS808. Maxon claims they also reversed their positions to reduce noise levels. The IC is a JRC4558. There is no 220R from the TS10 circuit, though the input buffer does have the TS10 emitter follower tacked onto it. So we can call it a hybrid TS9/TS10, but the effect portion of the circuit is just a TS9.

The OD9 is a TS9 with a JRC4558D and true bypass. It shares its PCB layout with the TS808 and TS9, but leaves out all the switching circuitry.

When the pedals are on, the OD808 and OD9 are exactly the same. The only difference is in the bypass.

ST9 /STL / Maxon ST-01

Theses guys all have the same mid boost circuit spliced into the regular TS circuit. The mid boost includes one half of a dual op-amp, and the other half is used for the input buffer. A transistor is then used for the output buffer.

The ST9 has its Mid Boost circuitry in front of the clipping section of the TS circuit. On the STL, its after the clipping section. But the Mid Boost stuff is otherwise identical.

Both the ST9 and STL seem to mix the 808 and 9 output resistor values. They have a 470R series resistor and a 10k shunt resistor. The positions are reversed like in the OD808.

The ST9 has a G100k tone pot, according to Dirk Hendrik’s tracing. The STLs is the typical TS tone pot, G20k, according to the factory schematic.

I have no idea whether the Maxon ST-01 is an STL or ST9, but it seems to be one of the two.

OD9 Pro Plus / TS9B / TS808DX

I’m grouping these together because I don’t know what’s in ’em. The OD9 Pro seems to have some boost circuitry, and it sounds like there’s an internal charge pump. Rumor has it that the TS9B is not a true TS circuit, but I haven’t seen gutshots yet. The TS808DX also seems to have some charge pump stuff going on in it, in addition to boost circuitry.

 

Tube Screamers

I spent last night looking over the guts of my OD808, and after finding out that it’s a TS9 circuit with the TS10’s emitter follower after the input buffer, the question that follows is this: if the OD808 is a TS9, then what is the OD9?

I don’t trust most of the crap I read on the internet. Let’s find out. We can start with by comparing what Maxon says with the PCB. Let’s look at Maxon’s Tech Talk for the OD9.

“Input is fed into a 2SC1815 emitter/follower transistor to optimize the signal’s impedance structure.  From here it’s on to the first section of an NJM 4558D Op Amp – NJM is the company formerly known as the Japan Radio Corporation (JRC) – the legendary manufacturer that designed and built tons of these components in the late 70’s/early 80’s.

“In this non-inverting Op Amp there are diodes in the feedback loop that clip the signal gradually as well as a High-Pass filter that keeps low frequencies from being distorted as much as high frequencies – the classic “808” configuration that creates the tone we all know and love.

“After this is a passive LP filter and then an active tone stage that uses the other half of the Op Amp to create a HP filter that brings back frequencies cut by the inter-stage LP filter.  After the tone stage it’s through the volume pot, another E/F 2SC1815 to provide a low impedance output and then the true bypass switch.

“The OD-9 features Mechanical True Bypass switching via a Fujisoku 4PDT switch – input signal is sent to ground when bypassed.”

I sounds like they’re claiming it’s a true bypass TS9 with the input and output buffers intact. Let’s look at the actual PCB. Here’s a good comparison: http://strangedesign.org/forums/index.php?topic=1426.0

We can see that the OD9 leaves out all the bypass switching junk that the TS9 has, even though both pedals still use the same PCB. I count six transistors on the TS9, which are all accounted for by this schematic: http://www.freeinfosociety.com/electronics/schematics/audio/pictures/ibanezts9.gif. The TS10 has seven, with the extra emitter follower. This TS9 RI has it’s opamp swapped for a 4558, but they normally come with a TA75558P.

So compared to the TS9 reissue, the OD9 has true bypass and the correct op amp. That’s it. Same board. Same enclosure. Same effect circuit, less the bypass junk (to be fair, Screamers have a good buffered bypass). The OD808 has the same effect circuit as well, with the TS9/OD9 output resistors, no 220R resistor in series with the op amp’s input, and the TS10 bypass emitter follower thingy. I call it a thingy because I have no idea what effect it has on the sound, or why it’s even there. But the OD808 has its own PCB, while the TS808, TS9, and OD9 all use the exact same PCB layout. Even when the PCBs have Ibanez screened on them and not Maxon. The TS10 has its own PCB, which is not the same as the OD808 PCB.

We can summarize the important differences (or at least the ones I care about) as follows:

TS808 – Original Run

– various 4558s, mostly JRC but sometimes other brands

– 100R series resistor, 10k shunt resistor

TS9 – Original Run

– “seemingly random op-amp chips” -Analogman

– 470R series resistor, 100k shunt resistor

TS10

-Some have MC4558, others JRC4558

-470R series resistor, 100k shunt resistor

-220R in series with opamp input

-emitter follower after input buffer, before bypass

-according to RG Keen, input buffer runs of bias voltage and not 9v

TS5

– TA75558P

– 470R series resistor, 100k shunt resistor

TS7

– JRC4558D

– 470R series resistor, 100k shunt resistor

TS9 – Reissue

– TA75558P

– 470R series resistor, 100k shunt resistor

TS808 – Reissue

– JRC4558D

– 100R series resistor, 10k shunt resistor

Maxon OD808

– JRC4558D

– 470R series resistor, 100k shunt resistor; series resistor is placed after 10uF cap and shunt resistor instead of before it

– No 220R as seen in the TS10

– Include extra emitter follower of TS10

– Input buffer runs on 9v, not bias voltage

Maxon OD9

– JRC4558D

– 470R series resistor, 100k shunt resistor

– True Bypass, no switching circuitry

 

There may be some minor differences in the TS5 and TS7, but who cares about those? I’d also like to collect info on the clipping diodes, but that’s a bit harder.

 

——

Just took a look inside of my TS7 and TS5. The TS7 has a linear tone pot. It’s marked 203B. The OD808 has 203W and the TS5 is marked 203G. I know that W is S-taper, but I don’t know what G stands for. The TS7’s clippers look very similar to those in the OD808, and according to the factory schematic, they’re MA150s (http://www.ibanez.com/parts/2004_PARTS/electronics/tonelok/circuit%20diagram/TS7-01.pdf). The TS5 has 1N4148s.

 

 

 

Pearl OD-05 and Maxon OD808

Fun fact- both of these pedals have OD in their names.

Maxon OD808

The OD808 was my first overdrive pedal. Of my many TS builds, on one (a Son of Screamer) ever got 100% there. Others were close, but the OD808 just had a little bit more creaminess to it. Having read all sorts of BS on the interwebs about TS versions, I triehd to trace it a few years ago, copying the PCB, and cloning it part for part. I never got around to figuring out what parts on the main PCB connected to the daughter board, and it’s sat on my shelf ever since. I just dug it out, and this is what I found:

– Clipping diodes are orange with a white stripe marking the cathode. Maxon’s marketing claims they’re Panasonic MA150s, which I’ve never seen for sale anywhere.

– Opamp is a JRC4558D. As expected.

– The input cap (C1) is 20nF. Not 22nF. True to the original TS808. In fact, all the component values are true to RG Keen’s schematic in his Technology of the Tube Screamer article. 51pf, not 47pf. 51k, not 47k. The two .22uF caps are tantalum.

– Analogman’s article is incorrect, kind of. The OD808 is not a TS10 circuit (at least the part that matters). It’s a TS9 circuit. There is no Ra (220R in series with opamp). It DOES, however, have the extra emitter follow in the input buffer. Not sure how much that matters since it’s for the bypass. The input buffer transistor is connected to 9v, not some bias voltage as RG’s article says the TS10’s is. The shunt resistor in the output buffer is 100k and the series resistor is 470R, same as the TS9.

– TR1-4 are C1815s. Q1-2 are K246.

– “It should be noted that the Reissue Series OD808 uses different output resistors than the original version.  Vintage TS808’s are susceptible to microphonic noise due to static electricity buildup on the chassis.  To improve performance, the positions of the output resistors on the reissue OD808 were reversed and their values increased in order to reduce noise levels.  This circuit change has no audible affect on the sound of the unit, save for reducing the noise levels.”

That’s a lot of words for trying to explain that it has the TS9 output resistors. But this is the one area where the positions of the components differ from RG’s schematic. RG has RB -> 10uF -> Rc-to-ground/Output. The OD808 has 10uF->Rc-to-ground/Rb->Output. Is this what Maxon meant by reversing the positions of the resistors? They put the series resistors after the output cap and shunt resistor instead of in front of them. I don’t think this should matter electronically, but what do I know?

Pearl OD-05

This is the coolest disappointment ever.

The Pearl Over Drive (the space is part of it’s name), is such a great idea. It’s a TS style overdrive, with clipping diodes in the neg feedback look. It’s one of the smoother sounding overdrives I’ve heard, probably due to the 1k resistors in series with the clippers. With the EQ Gain at noon, it sounds pretty transparent. No crazy mid hump. But the parametric EQ lets you give it any hump or scoop you want.

The disappointment: It can’t get clean. Even with the Over Drive knob turned all the way down, it’s still dirty. And doesn’t’ boost very much at all. For whatever reason, there’s a 10k resistor in series with the pot. That’s too much. Swapping it for 1k per BYOC and Aion’s suggestions allows it to get clean. And that doesn’t seem to reduce your max gain. The range of drive kind of sucks. It’s goes from a little dirty to entirely saturated pretty early on the knob’s rotation. Reducing that 10k is a must.

The cool: The parametric EQ is friggin’ sweet. Boosting 600hz makes it a great boost for my Dual Rec. Really, whatever frequency you’re lacking or have too much of, you can deal with. It’s incredibly versatile in that regard. And dealing with that 10k resistor makes it even more versatile. Set clean, it can be a treble boost, or whatever kind of boost you want. The enclosure is pretty cool too.

Every schematic for this thing has issues. After discussing it with BYOC, Aion, building two of them, and then getting my own unit, I can safely say that I know where to run the wires. They don’t necessarily go where the schematic says they do. That doesn’t mean it won’t work the same way, just that what’s on paper isn’t how it was assembled. Aion’s Fractal project is the only place to get a PCB that I know of, and BYOC’s Parametric Overdrive is a great kit for those who don’t want to source their own parts. Their PCB should be good, but don’t trust the BYOC schematic. It’s wrong.

For what it’s worth, my cheapo China-sourced 1S1558s that are probably not really 1S1588s seem to sound about the same as the real things in the OD-05. I need to turn this circuit into a lab unit and figure out what mods make it all the cooler.

Here’s the schematic that shows  exactly what’s in my unit:

Centaur Versions

The other day I stumbled on a revision that Bill made to the Centaur PCB after the changes he mentioned in the Premier Guitar interview. In the interview, the changes he talks about can easily be identified with R1B and R14B on the PCB. But he never addresses R1C, a 100k shunt resistor at the output. I had assumed that he had added R1C in the ’95 revision and he either forgot about it, or didn’t think it was significant enough to warrant mentioning. Going through some gutshots, I noticed that the PCBs immediately after his 95′ changes had vertically oriented output resistors, where most of the later PCBs have them horizontal. This means he made a third revision later on, most likely to add R1C. So we have a 3rd version of the Centaur PCB, with R1B and R14B, but lacking R1C.

Going over known units, we can get a general idea of what serials have what PCB (these serial ranges are not exact):

PCB Rev	Gold	Silver
Rev1 – ’94	001-300	?
Rev2 – adds R1B and R14B	300-800	S001-S200
Rev3 – adds R1C	800-3858	S200-S3922

These serials are rounded off to the nearest hundred, except for the highest ones that are known. I don’t know of a single Silver with a rev1 PCB, but they may exist, as S001 isn’t accounted for. The fact the Silvers with rev2 PCBs exist show that Silvers were being produced alongside Golds in the early days, though in much smaller numbers. My data on Silvers is much more tentative, as most are horsieless with rev3 PCBs, and the fact that the changes are bunched up in the earlier serial numbers means I need more info on those units. Changes in Golds are more spread out, and so a nice spread of data gives us a clearer picture than what we can see with the Silvers. Assuming a total of 8000 units (in reality, it’s most likely a little less with no 4000 range serials attested), we can get an idea of how common the various Centaur versions are:

PCB Rev	Gold	Silver
Rev1 – ’94	3.75%	?
Rev2 – adds R1B and R14B	6.25%	2.50%
Rev3 – adds R1C	40%	47.00%

 

We can do the same thing with Horseyman:

Horsie	Gold	Silver
Closed Tail	1-1800	S001 – S400
Open Tail	1800-3000	? – ?
None	3000-3858	S400 – S3922

 

Horsie	Gold	Silver
Closed Tail	22.50%	5%
Open Tail	15%	?
None	12.50%	45%

 

Horsieless Silvers are the most common. I know Open/Short Tail Silvers exist, but don’t know any serials for them, hence the question mark. Again, these numbers are very very tentative and have a lot of rounding going on.

The important thing to keep in mind is that whether or not a Centaur is Silver or Gold has little to do with what’s inside. There are no Silver/Gold circuit differences. But there are three different versions of the circuit, and 87% of all Centaurs have the rev3 PCB, which all klones are based on as well. The differences are minor, and only R14B should be audible. My experience is that swapping it in and out of the circuit doesn’t produce any audible effect, despite what Bill was going for.