DIY "Keeley"-TS9 mod questions

[red]

A while back I bought a new Ibanez TS9 reissue pedal, but I ended up not liking it that much - it was OK with my humbucker Les Paul guitar (which had a lot of low and high end), but it didn't do much with my Telecaster or Les Paul Jr (which are already mid-heavy). The pedal also didn't go so well with my tube amps, and after selling the Les Paul with humbuckers, I had no place for the TS9 pedal.

That is, until a week or so ago, when, lucky for me, a potential buyer tried the pedal out, didn't like it, and said if he'd ever buy one he'd have it modified Keeley-style. And then it hit me: I know how to solder, I know what a capacitor is, what a resistor is, and so on! So I got all the information I could from the Internet (I live in Romania so sending the pedal over to Keeley is really not an option, it would more than double the price of the pedal in shipping and customs taxes alone), and what I did was:

1. changed the 0.047uF cap to 0.1uF for more low-end.
2. swapped the JRC4558D IC I had in the pedal for a TI RC4558P.
3. replaced a couple of 1uF electrolytic capacitors to WIMA metal film.
4. replaced the 0.22uF tantalum capacitors to WIMA metal film (I'll change the one in the tone section to 0.18uF later).
5. changed the 4k7 resistor to 2k4 and the 51k to 20k (Keeley's "more/less" distortion).
6. the TS9 to TS808 mod: changed the 470 resistor to 100 and the 100k to 10k in the output section.

The pedal is very different, it's great-sounding and needless to say it's off the market now.

Now, my questions are:

1. especially when I turn the DRIVE knob almost all the way up I get pretty audible noise. It's not so bad that it doesn't get drowned in the actual playing once that starts (kind of like the 60-cycle hum of single coils - or 50-cycle hum, like we have in Europe). I've seen suggestions on the net that I should swap the input transistor (2sc1815) to a MPSA18. However, from the data sheets I found on the Internet, the 2sc1815 is "E C B", and the MPSA18 is "E B C". Should I just try to bend the legs on the MPSA18 until I can reposition "B" and "C" in each other's place?

2. which one is the input transistor exactly? :confused: I can solder and pick up a capacitor, but I'm not really competent enough to know just by looking at the board, if there's more than one component with the same specs on it. I'm assuming that the input transistor is the one in the upper-left corner of the board, near the IC:

http://www.planeteleven.net/tubescrmr/images/ts9mods_graphic.jpg

Is that correct?

3. I managed to find a picture of a Keeley-modified TS808 on the Internet, and it looks like he also changed the 0.02 input capacitor to something else (I can tell because there's a red non-polarized film-type capacitor there instead of the original see-through plastic greenish-ones that you can still see a couple of in the picture). I'm talking about the cap in the lower-left side of the picture, under the IC chip. Now, I didn't change that cap, and Keeley doesn't mention anything about changing that cap in the promotional talk about the TS808 Mod Plus on his website. So what value is that cap, and why was it changed?

http://www.stevesmusiccenter.com/KeeleyTS808Board.jpg

4. Also in that picture, you can see that he changed 5 resistors to the metal-film type. I've only changed 4 of them: 2 for the more/less mod, and 2 for the TS9 to TS808 mod. The blue one in the low-center area of the picture, I have no clue why he changed that, and to what value. On my board, that seems to be an 8k2 resistor (if I've computed the value right by the color code). What's that resistor for, and why would it be interesting to swap it for some other value?

Thanks for reading, and thanks in advance for your replies!

[red]

I might be wrong, but from this post in another forum (http://www.ultimate-guitar.com/forum/showpost.php?p=10308153&postcount=51) it looks like the extra resistor is connected to the LED. So it's probably a "brighter LED" mod.

I'm still extremely curious about the other questions.

[Suhnton]

According to a schematic I have (which is for a TS-808 clone, but it should be very similar), the input transistor has a 10K resistor and 1uF cap off the emitter. Going into the base is a 510K and 1K resistor. The collector goes to the +V rail. Have a look for those parts and you should be right.

I used 2N3904 transistors in mine, I don't get any noise issues with it.

[red]

Thanks! I think I found the input transistor - it should be the one in the upper-left side of the pictures in my post.

Does anyone know why the old 0.02 cap has been changed? Does it have a different capacitance value, or was it just changed to metal film for better quality? Aren't the see-through yellow plastic ones also metal film?

Thanks for the transistor suggestion, I'll see if it's available around and try that one too. :beer:

[Suhnton]

I think the yellow ones are mylar caps. As for the 0.02uF cap, if you're talking about the one at the base of the input transistor, then it's about right. The value I have is 0.027uF (close enough).

[duhvoodooman]

I doubt the transistor change would have an audible effect. You'd be better off changing all the higher rated resistors (anything 100K or higher) to metal flim. Carbon films at high ratings tend to be a little hissy.

Keep in mind that you've doubled the pedal's gain with that change of the 4.7K resistor to 2.4K, and that doubles the noise, too, when the gain is dialed up. So you're fighting physics, to some extent.

I wouldn't worry about that 0.02 cap. It's at the start of the input buffer and I've never seen any mention of it being tonally important.

Can't comment on that 8.2K resistor; that's not a rating I've encountered in a TS.

BTW, your bass didn't increase with that switch to a 0.1uf cap--it just compensated for the 2.4K resistor. The two form a frequency filter, and when you cut the one in half, you have to double the other to keep the bass response the same. If you really want more low end, increase it to 0.22uf.

[red]

I think the yellow ones are mylar caps. As for the 0.02uF cap, if you're talking about the one at the base of the input transistor, then it's about right. The value I have is 0.027uF (close enough).
The 0.02 cap I'm talking about is the yellow one right above the numbers 12 and 13 (12 is where the red wire connects) in the picture labeled "TS-9 Modifications". Yes, in the schematic it looks like it's connected directly to the input transistor.

Anyway, yes, I know the value is 0.02uF, both from the schematic and from reading what it says on the capacitor in my TS9 reissue pedal, my question was: do you know if the Keeley mod uses the same value for that capacitor (if you look at the second picture I've posted in my initial post, you can see that the yellow cap has been replaced with a red metal film cap), or if the mod uses a different value for the replaced capacitor?

[red]

I doubt the transistor change would have an audible effect. You'd be better off changing all the higher rated resistors (anything 100K or higher) to metal flim. Carbon films at high ratings tend to be a little hissy.
I don't expect to completely cure the noise, just cut it down as much as possible. I found this article (http://www.geofex.com/Article_Folders/TStech/tsxfram.htm) that stated:

"Note that in going to higher gains, you will inevitably increase the noise in the output. This may be curable to a degree by changing the input transistor to a quieter part (MPSA18 or 2N5089 is good) as well as swapping to a more modern and quieter opamp like the LM833 or both."

The Keeley Baked Mod (http://www.robertkeeley.com/product.php?id=4) page also specifies:

"I also change the input transistor to a MPSA18 for lower noise operation."

Now, I like the RC4558P very much, so I won't be changing that, but I though I'd give the transistor swap a try. It might not cure all of the background noise when going to higher gain, but it if cures some of it I'll be happy.


I wouldn't worry about that 0.02 cap. It's at the start of the input buffer and I've never seen any mention of it being tonally important.
Well, this is where (http://www.geocities.com/greene_pedals/newtech.html) I've seen a mention that it might be tonally important. I've also tried it out - I put a 0.1uF cap there and I got more bass than I had bargained for. But I really like the 0.02uF cap, and I don't think mr. Keeley changes that value, so I'll just swap it for a 0.022uF metal film for better quality and leave it at that.


BTW, your bass didn't increase with that switch to a 0.1uf cap--it just compensated for the 2.4K resistor. The two form a frequency filter, and when you cut the one in half, you have to double the other to keep the bass response the same. If you really want more low end, increase it to 0.22uf.
Indeed you are right, sir. I did try a 0.22uF, I thought the pedal got too muddy though and went back to the 0.1uF.

Thank you for your reply!

[duhvoodooman]

Whenever I've seen discussions of punching up the bass on a TS circuit, it's been via that 0.047uf cap in the opamp feedback loop that you already changed. I'd agree that the 0.02uf on the input buffer is best left as is. BTW, if you still want a bit more bass, you might want to try a 0.15uf cap where that 0.047 was.

Can't hurt to change the buffer trannies, though the pinout on the MPSA18 (which is what I've used in all my TS clones) is different, as you mentioned. Just leave the legs a bit long and bend them around carefully to avoid contact between them, and you shouldn't have any problem. I recently did the same thing to change from higher gain trannies to mid-range ones in another distortion pedal, and it worked fine. I'll be interested to hear if you discern any improvement from the switch.

FWIW, my favorite opamp in that circuit (and I've tried a LOT of them) is the TI TLC2272ACP. Seems to strike just the right balance of fidelity and low noise without sounding too antiseptic. Some people swear by the Burr Brown OPA2134A, but I found it a little too hi-fi in that circuit; sounded a little lifeless to me compared to others. But I have one in my MOSFET-conversion TS clone, and it sounds great in there. So go figure....

[red]

Can't hurt to change the buffer trannies, though the pinout on the MPSA18 (which is what I've used in all my TS clones) is different, as you mentioned. Just leave the legs a bit long and bend them around carefully to avoid contact between them, and you shouldn't have any problem. I recently did the same thing to change from higher gain trannies to mid-range ones in another distortion pedal, and it worked fine. I'll be interested to hear if you discern any improvement from the switch.
Thanks, I'll let you know as soon as I make the change - I had to order the MPSA18 online and now it's still on its way, should be here sometime next week I think.


FWIW, my favorite opamp in that circuit (and I've tried a LOT of them) is the TI TLC2272ACP. Seems to strike just the right balance of fidelity and low noise without sounding too antiseptic. Some people swear by the Burr Brown OPA2134A, but I found it a little too hi-fi in that circuit; sounded a little lifeless to me compared to others.
Thanks for the suggestions, I'll hunt around for those and try them out, should be easy enough with the IC socket I've installed...

[sumitomo]

This will be cool I just got one of these ts9's today with a blues jr I bought and next week I should be getting the Rodent Vood is making so I should have a fresh opnion of these pedals because I mainly just plug in an amph.Sumi:D

[red]

Can't hurt to change the buffer trannies, though the pinout on the MPSA18 (which is what I've used in all my TS clones) is different, as you mentioned. Just leave the legs a bit long and bend them around carefully to avoid contact between them, and you shouldn't have any problem. I recently did the same thing to change from higher gain trannies to mid-range ones in another distortion pedal, and it worked fine. I'll be interested to hear if you discern any improvement from the switch.
I've done it. There is almost no difference until the DRIVE knob is turned up about 2/3, but after that the difference is audible. It's not a huge difference, the background noise is still there, but IMHO it's definitely an improvement. Which is great for me, because I'm using only single-coil guitars (mostly a Telecaster) and I can use all the breaks I can get from the hum. I'll probably shield the pickup and controls cavities as well in the near future.

The change is probably not worth it at all for people using humbuckers or noiseless single coils, but it did make a difference, albeit small, for me. YMMV.

[thegnu]

Hey all, I'm new. I did the Monteallums mod to my pedal, and I was just cleaning up the work inside since I've gotten better. So then I wanted to implement some of the Keeley mods, and I read about those MPSA18's, and I had just taken some out of a dynacomp, so I thought, "cool"

I've only got two transistors on mine, I think because it's the true-bypass Maxon OD-9. They are circled below:
http://img10.imageshack.us/img10/3905/ts9modsgraphic.jpg

Can anyone tell me the model or pinout of the transistors I had there, because I was tired, and dropped one of them in the carpet, and I can't remember what I did to the other one. :thwap:

Also, that cap isn't in the signal path, is it?

I appreciate any help.

[Suhnton]

The transistors I used in mine are 2N3904's.
The pinouts are (with the flat side facing to the right): C-B-E top to bottom.
The GenaralGuitarGadgets (GGG) schematic uses 2SC1815's. I think they are the same pinout configuration.

[duhvoodooman]

The GenaralGuitarGadgets (GGG) schematic uses 2SC1815's. I think they are the same pinout configuration.
No, they're not. 2SC1815's have the collector pin in the middle position:

http://www.datasheetcatalog.org/datasheet/toshiba/964.pdf

MPSA18's are the same pinout as the 2N3904's, but much higher gain.

[thegnu]

No, they're not. 2SC1815's have the collector pin in the middle position:

http://www.datasheetcatalog.org/datasheet/toshiba/964.pdf

MPSA18's are the same pinout as the 2N3904's, but much higher gain.
So, I remember one of them was a 2SC1815. Praise the lawd. So, what would be the practical difference between using MPSA18's and 2N3904's, because the legs on my reclaimed MPSA's are stubby, and I've got brand new 2N3904's sitting there.

[duhvoodooman]

SoSo, what would be the practical difference between using MPSA18's and 2N3904's, because the legs on my reclaimed MPSA's are stubby, and I've got brand new 2N3904's sitting there.
The higher gain transistors will keep the buffer gain closer to unity. You may lose a few percent by going with a lower gain tranny. Shouldn't be a big deal.

[Suhnton]

No, they're not. 2SC1815's have the collector pin in the middle position:

http://www.datasheetcatalog.org/datasheet/toshiba/964.pdf

MPSA18's are the same pinout as the 2N3904's, but much higher gain.

Cheers for the correction (I was too lazy to look it up:) )

[thegnu]

Hey guys, thanks for the help. You guys are pretty darn friendly, as forum inhabitants go. :)

I had some MPSA18's that weren't too short I couldn't put small pieces of old heatshrink tubing and twist its feet around. So that's done.

I've got interesting stuff to report (at the very least for future googlers). On the OD-9, the TS808 mod resistors are already in place. As for the Keeley more/less mod, the "4k7" is a 2k2 resitor, and the "51k" resistor was a 20k. So I just let them be.

The Monte Allums tri-gain mod installs a switch that gives you stock 1n914a silicon diode clipping on one side, LEDs on the other, and a blend in the middle. I had some issues with the solder joints falling apart, so I took a little bit of that radioshack board with the copper pads per-hole, so I could mount the mod on a more stable surface. Now I've got asymmetrical germanium diode clipping, and symmetrical silicon diode clipping.

Except I screwed up my switch from resoldering it too many times, and for now, only the germanium diodes are working.

Thanks for your help! :rockon:

[red]

For future Googlers, as you say :), usually looking up the transistor type gives you the datasheet, most often in PDF format, and that shows you a picture of the transistor with the pinout clearly labeled.
Example:

MPSA 18 (http://www.fairchildsemi.com/ds/MP/MPSA18.pdf)
2SC1815 (http://www.toshiba.com/taec/components2/Datasheet_Sync//50/6455.pdf)

About lifting traces or solder pads, make sure you're using a proper wattage soldering iron (a 25W should be used for PCB work), and be very careful when desoldering (don't let the desoldering braid, if you're using that, cool and stick to the solder pad and then force it off, and don't pull on the components that are not ready to come out). If a trace does come off, you can fix it with a jumper wire (most of the time a resistor leg can be used for that), or you can scrape off some of the insulating coating from the track near the old solder pad and solder the component leg to that.

As for having the option to switch between clipping devices, my (unsolicited :D) opinion is that these things are better done on things that are designed to do that and are more experimental in nature, such as a BYOC OD2 kit.

And finally, about MPSA18 vs. 2SC1815, the swap is not really worth it unless you triple the ammount of distortion your pedal can do, or you're using the pedal with very noisy single coil guitars. Even then, the difference is subtle. I'd be surprised if anyone could hear it in a live situation.

Hope this helps.

[thegnu]

As for having the option to switch between clipping devices, my (unsolicited :D) opinion is that these things are better done on things that are designed to do that and are more experimental in nature, such as a BYOC OD2 kit.
Yeah, in retrospect, I don't much care for the switch, but I already have a hole where the switch goes. :) I'm in a learning and tinkering phase, so there's no harm in playing around as long as I'm careful not to blow things up. And as for removing components, I got the cheapo Radio Shack desoldering iron, and it's made component removal way, way easier that using a regular desoldering pump and desoldering braid.

Thanks for the feedback. Much appreciated. :)

[Stormrider]

IME the 2SC1815 is already a pretty good device. I don't think there is any point in changing them out in these pedals. If they had used 2N3904's from the factory, then maybe, but I'd let the 2SC's be.

[thegnu]

IME the 2SC1815 is already a pretty good device. I don't think there is any point in changing them out in these pedals. If they had used 2N3904's from the factory, then maybe, but I'd let the 2SC's be.
Again a little further off-topic, if you had a device you were trying to lower the noise floor in, and you changed out the capacitors based on Some Internet Guy's advice, but it made no difference, and it had 2N3906's in it (the PNP version of the 2N3904's, to my knowledge), would you swap them for something else? And what would you swap them for?

This device is a Boomerang+ phrase sampler.

[red]

Again a little further off-topic, if you had a device you were trying to lower the noise floor in, and you changed out the capacitors based on Some Internet Guy's advice, but it made no difference, and it had 2N3906's in it (the PNP version of the 2N3904's, to my knowledge), would you swap them for something else? And what would you swap them for?

This device is a Boomerang+ phrase sampler.
This probably won't help you much, but since we're drifting further and further away from the original topic here's my own contribution to that.
The boutique guys swear by silver mica capacitors and WIMA metal film capacitors. The wiring's also a thing with them (teflon coated wire mostly), and almost everyone seems to prefer 1% metal film resistors. Some people prefer higher fidelity op-amps where applicable.

Not many people care about transistors though, the only mention I've seen about "better" transistors were recommendations for MPSA18s. I guess you could always find all the compatible transistors with what you have in your gadget now and look up their spec sheets for more details. That would certainly help you make a decision based on more than some Internet guy's advice.

[Stormrider]

Again a little further off-topic, if you had a device you were trying to lower the noise floor in, and you changed out the capacitors based on Some Internet Guy's advice, but it made no difference, and it had 2N3906's in it (the PNP version of the 2N3904's, to my knowledge), would you swap them for something else? And what would you swap them for?

This device is a Boomerang+ phrase sampler.

The transistors won't make too much of a difference in most guitar effects circuits. Shielding all the wires carrying signal within the unit, using carbon film or metal film resistors instead of carbon comp, ensuring good ground connections (all connections really, not just grounds) etc... will make more of a difference than just changing the transistors.

That said, if i was building a circuit from scratch i would just use the better transistors as there is little cost differance; but i wouldn't bother to mod an existing circuit just because of them.

Better opamps can help; changing 4558 chips to NJM4580, NE5532, or the OPA's will help. I would, however, shy away from the OPA and burrbrown type opamps as they are almost TOO good for some circuits. They can introduce noise and oscillations unless the circuit layout, grounds, etc... are perfect

Caps won't make much of a difference as long as they work. Using poly or mica types for power supply bypassing (battery or whatever) will actually make things worse as ceramics are more suitable. Then again, i don't go for the audiophile "this cap sounds better than that cap" nonsense. :thwap:

All IMO. :rockon: