Rangemaster - Modern germanium transitor equivalents?

[tubeswell]

I love the sound of Clapton's 'Hideaway' guitar, and seeing as how I have a 'JTM45 equivalent' in the cupboard, I decided I'm GASin' for the promise of sounds I could get with the Rangemaster.

I see from RG's article, that the OC44 equivalents could be NKT275, 2N527, 2N508, 2N404, 2SB75, and that ECG158 works well 'if you can get the right one'. So in today's market (for one like myself who doesn't understand much about SS), what is the 'right one', and/or what off-the shelf equivalents should I keep an eye out for. TIA

[kleuck]

Hello.
First, i must say that i do not thing that EC ever used a TB, i cannot hear one in any of his tracks (it's not always obvious though)
I am building modern treble boosters -with NO germanium transistor- and recently i tried to build a modernized Rangemaster, still a work in progress but i already sold a few ones.
Just to expose who i am.
I don't expect to sell a lot of my pedals to US considering the euro/dollar ratio and the shipping costs, so I'm not in a marketing, hum "path" ? (sorry i'm french) but actually i had to compare my work on the "modernized germanium booster" to a good RM son or grand son, and used an Analogman Beano as a reference (because one of my first customer had one and was willing to be my first tester, not too far away from me so we can meet etc, casual stuff....) so i can say : if you want a good (indeed great) germanium treble booster, buy a Beano.
Sound is awesome, construction is realy good, the man added a lot of dc filtering, a really really good pedal.

[kleuck]

I'm not sure to have understood the question so, if the point is "how to build a good one by myself ?" i will say that the exact transistor reference is not so important, as long as you can get a good one (not too leaky and at least 40 of hfe) and make your circuit tunable ie : you can adjust the idle current to your taste.

[dai h.]

if you happen to find some and they are old Japanese ones, I would be careful not to overlook them. They can be quite good IMO. There was a guy on another forum (who was in NZ) who found a huge stash of 2SB172(?) and I advised him to just buy the whole lot (if cheap enough or cheap) and he ended up selling selected pairs (for Fuzz Face use) to DIYers (for a reasonable fee) and kept some for himself.

re: the supposed Clapton connection, I also am not quite convinced and think this could be more of an internet myth. Having said that though, the thing is just so simple I think you should build one for the heck of it and have fun with it regardless of whether he used it or not. I'd also suggest trying cheap ceramics for one or both of the small caps (that's what some units had--to be specific, old style red "dogbone" shape ones). Might give a bit more "colour" and make it a bit more interesting (or maybe go the other direction using something like a polystyrene or polypropylene for a "purer" less coloured tone).

[tubeswell]

Okay so if I can't find germanium transistors here (in NZ), what sort of PNP transistor would I put in place of the OC44 in that booster circuit? (I can't find any of the other transistors that R.G. had in his article either). I understand very little about transistors and their classisfications. (R.G.'s article is attached BTW)

[kleuck]

A germanium is mandatory in this circuit, and it must have at least a gain (hfe) of 40.
I use OC75, AC128, SFT323 and less known japanese transistors (2SAxxx).
You can also build a modern Treble Booster like the Omega : Omega

[dai h.]

I would try whatever you can get ahold of cheaply or reasonably. I would guess the maker just used cheap gen. purpose ones (though way back then, transistors might not have been super cheap as you can find today). IIRC some came with OC71, so (I'm guessing here, but) maybe OC44 was due to availability or perhaps they were lower noise. re: the gains, FWIW in my limited experience (measuring a couple hundred I had gathered over a few years here and there) finding a Ge with a gain of around 40 should be far easier than getting one for the (2nd?) one in a Fuzz Face (120 or so with low enough leakage) as far as (the apparent) relative ease of sourcing an appropriate transistor (meaning that I personally found many more in the lower gain range than the higher).

[Steve Conner]

Why is a germanium mandatory? Treble boosting is treble boosting, you can do it with tubes, op-amps, whatever. A silicon transistor may not sound right in the Rangemaster circuit, but you don't have to use the Rangemaster circuit.

I'd be tempted to try the MPSA92, it's a high voltage silicon transistor, and being high voltage it has low gain. Or MPSA42 if you want a NPN booster that runs off the same supply as your other pedals.

I got a batch of germanium transistors from a friend, intending to make a Fuzz Face, but they were mostly too leaky to be usable. There was an OC44 in there, but it had too much gain and was noisy. Nevertheless the best FF tone I got was with that in the first position and the least leaky AC128 in the second. I probably should try the OC44 in a Rangemaster circuit.

[dai h.]

(Electro-Harmonix) Screaming Tree is a silicon transistor one (treble booster). Then there's that one Brian May supposedly used (also Si transistor) the schematic of which should be floating around on the interwebs somewhere. I suppose you could use an EQ as well. Oh, and the way Ed Van Halen used the Phase 90 is treble booster-ish as well.

[kleuck]

Why is a germanium mandatory? Treble boosting is treble boosting, you can do it with tubes, op-amps, whatever. A silicon transistor may not sound right in the Rangemaster circuit, but you don't have to use the Rangemaster circuit.

Because, no, a treble booster with a germanium transistor is NOT a treble boosting device, but a mids & hi-mids booster.
If you try a silicon one in the RM circuit, apart from the biasing wich would not be right, you'll end with an ear-piercing device.
That's why the Brian May circuit for example has a low pass filtering, cutting the highests frequencies, where a RM doesn't need cause it has a poor response in highs actually.
Apart from that, a RM is producing hi levels of even order harmonics and has great sounding and chewy distortion on his own, where a Si one....

[Steve Conner]

So you're saying that the germanium transistor itself has such a bad frequency response that it limits the treble boost? I've heard that they have a very low ft, but didn't realise it was as low as that.

Any cheap silicon transistor will amplify to several MHz, so that would be a problem. Maybe you could whack a large capacitor from collector to base, to slug it.

If a Rangemaster doesn't actually boost treble, maybe that explains why it isn't so obvious on the Beano album. I heard that Clapton turned the tone knob on his guitar down and used the neck pickup, then used the treble boost to put some of the highs back. That's why the Beano tone sounds so fat and creamy, but still has some detail in it.

[kleuck]

Not only the transistor by itself actually, but also the low impedance of a germanium treble booster.
The low gain of the transistor implies a low base impedance, wich comes in // with the bias bridge, so the real input impedance of a RM is ver very low (around 12 Kohms) and cuts out all the higs just like a 50 ft long crappy cable.
In my modern TB i had to put an lowpass filtering too, and it's the same ith the Omega.

If a Rangemaster doesn't actually boost treble, maybe that explains why it isn't so obvious on the Beano album. I heard that Clapton turned the tone knob on his guitar down and used the neck pickup, then used the treble boost to put some of the highs back. That's why the Beano tone sounds so fat and creamy, but still has some detail in it.

Yes but usually you can her the "grit" of the transistor, wich is very very unique, remaining the dirt you can get from a pentode.
Anyway, a TB wathever technology it's build from, is a really usefull cirucit with tubes amps, and indeed, the Beano (wich is mostly a RM clone) is fat fat fat (i try to nail this fatness with my own modernized Ge TB, but it's really not easy.
Here a sample trough a Badger 30, the guitar is an ash Telecaster, first the amp only, then the Beano, then my own pedal : btb_1.mp3 - File Shared from Box.net - Free Online File Storage

[J M Fahey]

Why is a germanium mandatory? Treble boosting is treble boosting, you can do it with tubes, op-amps, whatever. A silicon transistor may not sound right in the Rangemaster circuit, but you don't have to use the Rangemaster circuit.

I'd be tempted to try the MPSA92, it's a high voltage silicon transistor, and being high voltage it has low gain. Or MPSA42 if you want a NPN booster that runs off the same supply as your other pedals.

At last the (or "a") voice of Reason.
I didn't want to get into Faith based discussions, but now that I see I'm not alone, let me state that we do not only not need Germanium, we do not need it to be PNP; even more, we do not even need it to be a Transistor.
For me, *parts* are not important; *what it does* is.
It can be reasonably cloned using that wonderful building block: the Op Amp.
What does it *do*, in essence?:
1) Amplify? ... check
2) have a low and reactive input impedance, such as .005uF in series with 12K ? ... check
3) Have unity gain at 80 Hz? ... check
4) Have a 6dB/octave rising response? ... check.
5) Have a 6dB octave rolloff from 5KHz up, so the gain is not 60x but 30x (because of 6dB loss at that frequency?) ... check.
6) Have the same (constant) gain (around 30X) from that frequency up, because input rising 6dB/oct gets compensated by 6dB/oct rolloff, leaving net change around 0dB? ... check.
7) There are claims of second harmonics and nonlinearities .
Can't calculate them out of the top of my head as in points 1 to 6, but the practical approach would be to build an original clone, germanium transistor and all, hit it with various frequencies at different levels, and duplicate that adding an approppriate non-linear network in the feedback loop, one specialty of Op Amps.
8) Noise/unavailability/lack of repeatability/unbearable crackle/major Pop ? ... we can leave them out of the equation, can't we?.
9) Let's not forget that the Rangemaster does not sound by itself (nobody sends it right to the recording console) but is an equalizer and pre-processor for a classic overdriven tube amp, which does 95% of the work.
10) Where's the Mojo? ... good question.

[dai h.]

but you won't be able to charge as much for it since, "it's not exactly what Clapton used"(supposedly). Getting the same result using a different way is outside the realm of mojo thinking. (and beyond my skill level also, lol)

[kleuck]

At last the (or "a") voice of Reason.
I didn't want to get into Faith based discussions, but now that I see I'm not alone, let me state that we do not only not need Germanium, we do not need it to be PNP; even more, we do not even need it to be a Transistor.
For me, *parts* are not important; *what it does* is.
It can be reasonably cloned using that wonderful building block: the Op Amp.
What does it *do*, in essence?:
1) Amplify? ... check
2) have a low and reactive input impedance, such as .005uF in series with 12K ? ... check
3) Have unity gain at 80 Hz? ... check
4) Have a 6dB/octave rising response? ... check.
5) Have a 6dB octave rolloff from 5KHz up, so the gain is not 60x but 30x (because of 6dB loss at that frequency?) ... check.
6) Have the same (constant) gain (around 30X) from that frequency up, because input rising 6dB/oct gets compensated by 6dB/oct rolloff, leaving net change around 0dB? ... check.
7) There are claims of second harmonics and nonlinearities .
Can't calculate them out of the top of my head as in points 1 to 6, but the practical approach would be to build an original clone, germanium transistor and all, hit it with various frequencies at different levels, and duplicate that adding an approppriate non-linear network in the feedback loop, one specialty of Op Amps.
8) Noise/unavailability/lack of repeatability/unbearable crackle/major Pop ? ... we can leave them out of the equation, can't we?.
9) Let's not forget that the Rangemaster does not sound by itself (nobody sends it right to the recording console) but is an equalizer and pre-processor for a classic overdriven tube amp, which does 95% of the work.
10) Where's the Mojo? ... good question.

Many people tried, as far as i know, nobody suceeded.
With my Ballast (non ge modern version), i nail the RM sound, but it's not exactly the same grit.
YouTube - Ballast "Trouble Booster" demo 2 - Treble Booster pedal by Le Gecko Electrique
Even with tubes it's not easy to emulate the EF86 sound.
Yes, of course a TB is meant to be used in front of a Tube amp, just as a guitar is meant to be plugged in a guitar amp, not an HiFi one, what's the point ?
A good clone of RM does not crackle or pop...but it can be a good radio tuner, and can hiss
Anyway i was not saying that you must use a Ge transistor to make a TB, just that the RM circuit can't give a good sound with a Si one.

[J M Fahey]

Fine with me.

[Steve Conner]

Yes, I agree too...

I have one question though. What do you mean by "the EF86 sound"? Surely it is very easy to emulate it with tubes, by using an EF86

It would be interesting to see whether the grit comes from non-linear Miller capacitance. That could be simulated by putting a large diode (1N4001) from base to collector of a silicon transistor.

[kleuck]

Because a Germanium transistor used as in the RM makes the same kind of harmonics ratio than a hard-driven preamp pentode like the FE86.
And i wanted to say "if you want to get rid of the EF86 microphony by using some other tubes, it's not an easy task to emulate its sound"
If you heard at my sound sample, you know that even with a Ge transistor, it's not easy to emulate the true RM/Beano sound.
If you design a modern circuit, with less issues (impedance, stability), it doesn't sound the same.
Perhaps is it in close relation with the exact transistor though, i do not have OC44 yet, so i tried only OC75, AC128, SFT323 yet.
Pentodes have a low Miller effect Miller Capacitance

[Steve Conner]

OK, well the way to fix that is to use only NOS Mullard EF86s on rubber shock mounts. Works for me.

I know EF86s have low Miller capacitance, and that's what makes them sound bright. I'm suggesting that germanium transistors have a bad frequency response because they have extremely high Miller capacitance. If that capacitance were non-linear with voltage, that would be a source of distortion. I'm saying that is maybe the "grit" you're looking for, and it could be simulated by taking a device with a large, non-linear capacitance and putting it from base to collector of a modern transistor.

Since large diodes leak a lot, it would also give authentic vintage bias drift with temperature.

The low input impedance is an important part of the sound, as with the Fuzz Face. But any low gain transistor will have that low input impedance.

[J M Fahey]

If you want a *large* and *voltage modulated" collector to base capacitance, look no further than a BB104 or some from that family.
It might provide true grit even on an Op Amp.
Ooops !! I did it again. Sorry.
You can me if that pleases the Gods.

[Steve Conner]

Wow, it goes up to something like 30pF, I think that would be overkill.

The OC44 has a Ft of 6-15MHz, so that kind of kills my hypothesis.

[tubeswell]

Yay! I scored a NOS Mullard OC44 for NZ$25 - more fool me I s'pose.

http://images.trademe.co.nz/photoser.../163002806.jpg

[Alex R]

Phillips Miniwatt NOS EF86 are low-microphonics too ime, and cheaper than Mullards. (@ Steve)

[tubeswell]

Double yay! My (very expensive) Mullard OC44 arrived in the post today (and I feel like a tin-arse). How do I test it in my multi meter? - I have a Digitech N287 meter with a socket with 'E' 'B' 'C' and 'E' holes for NPN and for PNP, but which lead is which on the transistor? There is a red dot on one side (and a yellow dot on the top)

[J M Fahey]

Well, you can hook it to a 300V, 15A Lab PSU in certain way, and if it doesn't produce a mushroom shaped cloud then ......
ok ok, just kidding .....
seen from below leads form a triangle: E B C with the red dot being the collector.
Plug it into your multimeter and you should hav a rough idea of its gain (Beta) which should be around 30-40.
*This* test is non-destructive.
The yellow dot on top should give an idea of the gain, but don't remember the original coding.
Good luck.
Remember reading the approppriate issue of Beano magazine helps a lot.

[tubeswell]

seen from below leads form a triangle: E B C with the red dot being the collector.
Plug it into your multimeter and you should hav a rough idea of its gain (Beta) which should be around 30-40.

The yellow dot on top should give an idea of the gain, but don't remember the original coding.

Thanks. The leads are in a line on the bottom. So the one closest to the red dot is the collector, and the next one must be the base?, and the furthest one must be the emitter? But do I read it in the PNP side or the NPN side? (Never mind I tested it - it reads at '41' on the hFE)

Remember reading the approppriate issue of Beano magazine helps a lot.

Is that a mag for South American electronics enthusiasts? (Haven't seen it over here) ;-)

[tubeswell]

So to re-cap, I have my OC44 and it measures 41 Hfe on my meter.

R.J. has some Rangemaster circuit tuning and improvement suggestions, (including 2M2 switching noise snubbers, and a couple of tuning pots for the bias, and a 25kA boost level pot where you have a transistor with a lower Hfe) which I think I have captured in the attached schematic. (Are those 2M2 resistors in the right place R.J.? TIA)

[R.G.]

Nota Bene
The germanium vs silicon issues in the Rangemaster seem to stem from the near-cutoff behavior of early germaniums. Back when I messed with this, getting the collector into near-cutoff was important to get a good sound to my ear at least. That's what all that stuff about getting -7V on the collector was.

I think that like many early effect circuits, this one happens to use an undocumented side effect of the crude semiconductors. In this case, it's the near-cutoff behavior of hfe change. It's a much softer distortion than more modern devices. I've built this with germanium devices of more modern manufacture, and that doesn't sound the same to me, either. In my mind, early germanium and crude processing is what gives the exploited side effect. I'm pretty sure I could whip up something with opamps to do a similar transfer function with a little time. It is likely not that it's germanium, but that the transistors get squirrel-y near cutoff.

It's a theory anyway. I did the Rangemaster stuff almost 11 years ago. Even the revival is becoming archaic. 8-)

[kleuck]

You must adapt the bias to the transistor you are actually using, it's an issue i encountered with my modernized RM, OC75 sounds best at 6 volts where OC44 are fine with 7, etc etc
I obviously ended up with trim.

[tubeswell]

Thanks gentlemen. So are those 2M2 in the schematic I posted in the right place or...? (TIA)

[kleuck]

The input resistor is fine, but the output on should be tied to the circuit output, right after the 100 nf cap.

[tubeswell]

Thanks kleuck - I did wonder about that .01uF (in terms of how to treat the audio path going to the output) - corrected version attached (with pretty colours)

[kleuck]

And 2M2 is pretty high for this circuit -low values=less pop, usually you can't go too low, cause you lower the input impedance too.
The RM input impedance is very low, so 500 K or 1M would be enough on the input, on the output 100 k is enough.
And you don't want to power down the circuit when switching the effect off, no circuit can be biased and start instantly, so you ill have a lot of noises when using the switch.
You'd better wire a true bypass Many Bypass Switching Options

[tubeswell]

Many thanks for that handy link kleuck. I'm eternally grateful. This is how I interpret your suggestion. (The battery is supposed to be wired through a switch on the output jack with this DPDT switch, but I acknowledge I could use a 3PDT switch in that regard).

[kleuck]

The input resistor should be soldered right before the input cap of the circuit.
Here's a modernized clone : http://diy.musikding.de/images/stori...angeschalt.pdf
Beware, it's a negative ground one (but you can build the same way, the Beano is a negative ground too)
Usually there's no power switch in pedals, you just need a stereo jack at the input : http://www.generalguitargadgets.com/...8044a841b3e85a
When there's no jack plugged, the battery is off circuit (no connection to ground)
Beware, this is a generic negative ground schematic, if you go positive ground, the "+" from the battery has to be tied to the ring of the jack.

Here a positive ground clone : Rangemaster

I realized i did a mistake : i use 1 µf output cap, but if you use a 10nf one, the output resistor should be high, to avoid too much low loss from the output, so stand with 1M or 2M2.