Wednesday 17 September 2014

EHX Germanium Overdrive

Thanks to Miro for the excellent trace job he did on this.

Hope you're stocked up on 470nF caps.  They were all polarised but I have shown poly for the caps with a 5mm pitch because I expect that is what most people will use.  I have shown electrolytics for the two with only a 2.5mm pitch just so people know the polarity if anyone would like to build it with electrolytics, but again I expect most would use poly or maybe 2.5mm multilayer ceramics for them.  Any PNP germanium can obviously be used an an alternative to the NKT275, and I suspect more people will have 2N3906's in stock than 2N5087's, so try those for Q2 and Q3.

Info about the original:
Classic '60s Germanium transistor overdrive and more. In addition to a Gain control, the Electro-Harmonix Germanium OD allows the user to control Bias for attack and the circuit Volts for that sweet spot found when a battery's voltage dips. Dial up that slightly torn speaker sound OR dial in total Germanium "Satisfaction."

Germanium is an element like carbon or gold. Germanium transistors were used extensively back in the 1950s and ‘60s. Eventually they were replaced by components that were considered “better.”  The EHX Germanium OD and Germanium 4 Big Muff Pi use these old school transistors. They have a distinctive sound, some say smoother or “spongier.” You be the judge, hear the Germanium OD here.








and the schematic




Thanks to ξεναγός νεκρόπολης for verifying and providing the following working voltages.

Q1
e 5,41
b 4,17
c 3,76

Q2
e 4.37
b 3,75
c 0

Q3
e 3,02
b 2,43
c 0

33 comments:

  1. Awesome one mark. I bought one of these years ago and returned it cause I thought if sounded like crap. But the funny/sad thing is that i didn't understand how it worked and understand the idea of bias (I was like 14 when I bought it). I've always wanted another, but I figured I would live to build it instead, allowing me to really play with this circuit. It sounds amazing, especially when you understand what the controls do. Lol

    ReplyDelete
  2. build it but can't make it work properly .
    i can hear that germanium overdrive sound(btw i've used ac128 and 2n3906)
    BUT
    all pots do "something" but when i turn volts all way down for example everything stops. like volume pot. i don't know if this is how it should work but i didn't expect it.
    i always have a hzzzzzzzzzzzzzzzz sound and only in some sets output is in unity.in all other sets it's below., and i'm pretty sure i don't have any sort or bridge. it's a really small board so i've checked everything lot of times.
    now for the voltages, i don't know how i should measure them, as by changing pot settings voltages change as well.
    the strange thing is that i get same voltage in all 3 germanium legs no mater how i have set the pots...
    i have to go to work now..late at night i'll be back and recheck the hole thing.
    Ah! and i checked the schem too and everything is ok between them!
    hole i'll make it cause it has my 3 favorite phrases...electro harmonix, germanium and overdrive...!!!

    ReplyDelete
    Replies
    1. I must say that my original doesn't sound nothing like that PGS demo. So there is a chance of it working properly.
      +m

      Delete
    2. Measure everything with the pots in the centre positions, it'll still give us a clue if something looks drastically wrong

      Delete
    3. With the first set of settings on the PGS demo, my original unit is way under unity level. This may suggest that the demo may not reflect reality to a degree we'd hope.
      +m

      Delete
  3. Funky pedal but it is pretty cool... I would probably put a booster after this one for certain settings.
    Thanks! Added to my pile hehhe

    ReplyDelete
  4. In the demo it says the voltage pot drops it all the way down to just 3v DC, and the sound does get very sputtery down there, especially if you also drop the bias (which is technically a reverse pot, but you get the idea), so I think that's just the way the pedal works.

    The original Ge transistor was rare even back in 2006 according to some sites I checked, but I don't know how much difference having the original tranny makes. On Ebay, you can get 3 of them for $19 plus shipping. http://www.ebay.com/itm/like/111301851966?lpid=82

    New today, this entire pedal (without the original tranny, I suspect) only costs $74 retail.

    Very interesting pedal from a design standpoint, - having bias and voltage sag as user controls is pretty unusual. Thanks for the layout and the demo.

    ReplyDelete
    Replies
    1. Those are not original nlt275'd, they're repos that are given the same number, but they do not sound nor act the same as the real ones. To give you an idea of how much the real ones are the last time a red dot nkt275 transistor was on eBay, just one mind you it sold for over $80USD. I have 4 sets of real red dot nkt275's in fuzzface range sitting waiting to be used that I got before they disappeared and got so valuable.

      If the bias goes down to 3v it shouldn't turn the pedal off, I would expect it to get quetwr, but not turn off so something's actually wrong that has to be ironed out.

      Delete
    2. Ii meant nkt275's. It's hard to type on an iPhone with my stupid sausage fingers. Lol

      Delete
  5. What a great sounding pedal. Great pick man. This is like a soft germ fuzz. Excellent.

    ReplyDelete
  6. verified!
    yesterday i guess i overheated the ac128 with my fingers, and the next moment i tested it for 5 minutes.then i made the comment, went to work, and today it works!!!!!
    auto-debugged!
    now volume is still there when volts pot is at max.(yesterday i wanted to say "all the way up")

    anyway.
    very nice overdrive. not too flexible though..
    working voltages now!
    ac128
    e 5,41
    b 4,17
    c 3,76

    q1
    e 4.37
    b 3,75
    c 0

    q2
    3 3,02
    b 2,43
    c 0

    all voltages with the pots at center positions
    thanks Mark and Miro!

    ReplyDelete
    Replies
    1. Cheers!
      That actually explains everything. For a few germanium fuzzes i've built, i've ran into this exact issue. It'll take a few hours for the transistor to get back to normal working range after its soldered in. And sometimes the transistor doesn't come back alive, no matter how carefully one solders it to the board. (This is why i don't do much Ge builds myself anymore - those are unreliable crap)

      Great to hear it is working!
      +m

      Delete
    2. Oh. And separate thank you for the voltages! Those should help others in case of problems.
      +m

      Delete
    3. Thanks matey, I've added the working voltages to the main post

      Delete
  7. Awesome. That's the exact reason I don't solder my transistors, Ge's in perticular. I have this fear that I'm going to fry them from the hear of my iron, so I just use sockets.

    ReplyDelete
  8. Johns: "I have 4 sets of real red dot nkt275's in fuzzface range sitting waiting to be used that I got before they disappeared and got so valuable."

    When did you acquire those? - or more accurately, about what year did the originals become very hard to find?

    ReplyDelete
    Replies
    1. They have been hard to find for a lot time now, not sure the exact time. I got mine maybe a year or two ago when I first got into building, and long sorry short met someone that as able to get them, so I got 8 and was lucky enough to get 4 different fuzz face sets out of them (holy grail pair, hot pair, low pair, even pair).

      The thing is just cause it has the part number doesn't mean it's the real thing. The real ones are short and fat and have either a white or red dot on them, like in the link to a picture below:

      http://i.ebayimg.com/t/2x-NKT275-Germanium-Transistors-tested-for-Fuzz-Face-nkt-275-nkt-275-nkt-/00/s/OTU2WDE2MDA=/$(KGrHqF,!o0FCyLJ2j7OBQ2juwd+)w~~60_12.JPG

      Some company, I think in the US, began to make "new" ones, and used the same part number, but they are NOT the same, and do not sound the same. The repo's look more like AC128's and are long and skinny as seen in the link to a picture below:

      http://i.ebayimg.com/00/s/MTIwMFgxMjAw/z/0SQAAOxyoahSWZc3/$T2eC16Z,!)4FI,TR-mUnBSWZc2ZPr!~~60_35.JPG

      The repos are being sold as the real deal, and they're not. I don't know how much luck you'll have finding real ones, but the last time I saw them on eBay, some time last year they sold for $80 for 1, that was "untested." Some people in community seemed to buy up pretty much every last one, *cough* analogman *cough*

      IMO I wouldn't go searching all over creation for the real ones, or really use the repos, just use a Russian pnp and you'll get great tone and be just as happy. I thought I needed them due to the mojo magic, and saving them form some very special builds.

      Delete
  9. I'm a bit new to this, but it looks like there's a 10K resistor to the LED - isn't this usually a 2.2K? Thanks!

    ReplyDelete
    Replies
    1. Yes it is higher than I would use, but I kept it true to the circuit. Personally I would use a 2K2 resistor for most standard (not superbright) LEDs.

      Delete
  10. No need for a voltage inverter here?

    ReplyDelete
  11. I didn't solder the transistor off the board on my original, so i don't know deeper specs about it. AC125 should work fine.
    +m

    ReplyDelete
  12. Thanks... So out of curiosity, why no voltage inverter despite the use of PNP transistors? Feel free to ignore me if it's an annoying question;)

    ReplyDelete
    Replies
    1. PNPs can be wired in reverse, so the current flows like it should. This just means that one takes the collector to ground instead of emitter. Fuzz Factory is another example of this method.
      +m

      Delete
    2. Thanks mirosol. I suppose I really should read up a bit on transistors.

      Delete
  13. Hi. I tried this one but doesnt work. I used bc108 because I didn't find germanium transistor. Also I don't understand the led on the schematics. I link the positive on the board and another part? Sorry my poor english. Thanks.

    ReplyDelete
    Replies
    1. you need a Germanium transistor for Q1, the overdrive really works on altering the bias of the transistor and give it it's distortion characteristic. without it, the pedal will be a no go. the +LED on the board is for the ON/OFF indicator, so the negative lead goes to the switch like you would put when wiring the pedal without the LED being connected to the board.

      Delete
  14. The BC108 is NPN and the Ge transistor is PNP. It won't work if you use BC108 in place of the Ge transistor.

    At the top of the page there is a tab called "offboard" which shows the offboard wiring

    ReplyDelete
  15. I used soviet gt402b from 1978)) its kind of higain overdrive now!

    ReplyDelete
  16. I used soviet GT403B for Q1 and C1815 for Q2, Q3. Added an extra capacitor from Q1 E to ground, nominal 0,22uF, did not like the sound with Gain pot in total left side position.

    ReplyDelete
  17. Built this today with 2N5087s (as specified in the layout) and an AC128. There's no mention of hFE or leakage in this thread, so I auditioned several transistors before finding the ones that sounded and reacted to the knob settings most like the PGS demo.

    To my ears, 50-70 hFE with very low leakage (<0.04mA) sounded best. I tried some OC42 with 0.0mA leakage, but those were a bit dark and didn't sputter or spatter when starved. So, it seems perhaps a wee bit of leakage is required. hFE >130 or so starts to sound creamy indeed, if that's what you're after, but not a lot like the demo on certain settings, IMO. Since ODs are often meant to push an on-the-edge amp into "overdrive", I opted for the lower hFE and less "cream".

    Lastly, to get the Volts control to work like the PGS demo, I swapped lugs 1 and 3.

    Great sounding overdrive to push the pre OR power section of your all-tube amp (if you have an FX loop). Just my $0.02. YMMV.

    ReplyDelete
    Replies
    1. Y'know ... I shouldn't have dismissed those OC42s so quickly. They sound great in here too -- just a bit different than the demo. hFE ~70. Try 'em if you got 'em!

      Delete