Announcement

Collapse
No announcement yet.

Are all transistors created equal?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Are all transistors created equal?

    Of course the answer is 'no', but how unequal are they? The reason I ask is that I have built a Keeler Kick from t he guitarFx forum. The specified transistors are BC182L (NPN) which have the Collector as pin 2. While waiting for the delivery of those specific transistors I thought I'd try it with some s9013 (also Npn) that I have, and twisted the legs around to get the right orientation. After a couple of attempts (a bit of reflow was necessary), I got some sound coming out but it's very quiet and not at all fuzzy.

    Both transistor types are described as 'general purpose npn', but are the characteristics different enough to give the results that I have experienced? There doesn't seem to be the joyful knowledge of this vacuum oriented forum in the silicon world and what I know about transistors could be written on a, well, transistor.

    Many thanks

    Jeff
    It's not microphonic - it's undocumented reverb.

  • #2
    Short answer, no.

    "Small signal" transistors are quite interchangeable without nuch pain.

    Please post the schematic here so we can suggest some voltage readings.
    Juan Manuel Fahey

    Comment


    • #3
      Keeler Kick Schematic & layout

      Click image for larger version

Name:	Keeler Kick.png
Views:	1
Size:	61.6 KB
ID:	830119
      Originally posted by J M Fahey View Post
      Short answer, no.

      "Small signal" transistors are quite interchangeable without nuch pain.

      Please post the schematic here so we can suggest some voltage readings.
      Hi JM, thanks for the reply and offer of assistance. I have spent the last 3 hours drawing up the schematic based on the vero layout. it is draft, hasn't been checked by anyone else and i do not know the accuracy of the source - so there may be a few 'variables' to contend with. Both are attached here. as a reminder, the symptoms are that I am getting a very low volume, not very fuzzy output. Bypass works fine!Click image for larger version

Name:	Keeler Kick 0.1.png
Views:	1
Size:	53.6 KB
ID:	830118
      It's not microphonic - it's undocumented reverb.

      Comment


      • #4
        This schematic just looks plain wrong. :-O

        As a sanity check can you please measure the voltages at the collectors of all the transistors and get back to us with the results?
        "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

        Comment


        • #5
          Agree.
          The first stage (Q1) is fine, up to and including the clipping diodes and the following 1nF cap .
          Please label *all* parts, including caps, with some reference (C1 , R3, etc.) for clarity.

          The second stage , *as drawn*, is wrong.

          Second guessing the designer, this is similar to what's spread all over the Net, not "real" designs (as in setting a goal, calculating parts values, biasing, etc.) but what I call "Photoshop designing", meaning cut and paste fragments of *other* designs.

          This particular one roughly appears to be a first single transistor booster, similar to Electro Harmonix LPB1 (check it), followed by a Big Muff type tone control, them a Germanium diodes clipping stage (won't clip much because stage 1 does not have that much gain to begin with and is followed by a lossy tine stage), followed from what probably is the Silicon NPN version of a classic Fuzz (think Fuzz Face and its family).
          The Fuzz pot and associated capacitor (labels please) top ends should go straight to ground.
          Dig the various Fuzz Face schematics (probably the Silicon Face or something) and compare.

          As a side note: the Keeler Kick designers *should* supply a schematic, besides just a layout.
          Juan Manuel Fahey

          Comment


          • #6
            (1) All transistors are not equal.
            (2) All transistors of the SAME TYPE NUMBER are not equal.
            (3) Even the same transistor is not equal to itself if temperature or current changes.
            (4) Given that, competent design work makes the circuits using the transistors compensate for the differences.
            (5) In a -> well designed circuit <- things have been chosen so that if the transistors used are (a) the same polarity and (b) have enough gain the circuit will work reasonably well over a very wide range of transistor variation.

            In general, a 2N5088 will work in most (a) well designed and (b) small signal circuits with less than about 30V of power supply. This leads to Keen's Second Law: when in doubt, use a 2N5088.
            Amazing!! Who would ever have guessed that someone who villified the evil rich people would begin happily accepting their millions in speaking fees!

            Oh, wait! That sounds familiar, somehow.

            Comment


            • #7
              Originally posted by J M Fahey View Post


              As a side note: the Keeler Kick designers *should* supply a schematic, besides just a layout.
              The Keeler Kick is a commercial product that has been cloned. I chose that one just because - I don't know anything about it. It's good learning.

              Speaking of learning, I just found a bit of a school boy mistake in my build <embarrassed> whilst checking voltages as per Steve O'Connor's request. Which now gives me volume and a bit of fizz. Not fuzz. Only the volume control seems to have any effect. So before drawing on any more of your time, I'll go and do the Father's Day thing and do some more trouble shooting and revise the schematic. Then I'll report back. I'll have little bench time this week so depending on progress, updates may be scant.

              Thank you everybody.

              Jeff
              It's not microphonic - it's undocumented reverb.

              Comment


              • #8
                Updates - diagram and voltages

                Originally posted by R.G. View Post
                (1) All transistors are not equal.
                (2) All transistors of the SAME TYPE NUMBER are not equal.
                (3) Even the same transistor is not equal to itself if temperature or current changes.
                (4) Given that, competent design work makes the circuits using the transistors compensate for the differences.
                (5) In a -> well designed circuit <- things have been chosen so that if the transistors used are (a) the same polarity and (b) have enough gain the circuit will work reasonably well over a very wide range of transistor variation.

                In general, a 2N5088 will work in most (a) well designed and (b) small signal circuits with less than about 30V of power supply. This leads to Keen's Second Law: when in doubt, use a 2N5088.
                Thanks RG and JM and indeed SO'C. I have taken JM's advice and checked out some of the circuits he mentions - on RG's website. That has certainly helped me understand what's going on. So a transistor is kinda like a triode - base is grid, collector is anode and emitter is cathode (in an analogous way), right? i have also taken the voltages as per Steve's request - here they are in table form. I did see somewhere a similar table and a narrative regarding the minimum voltage difference between a couple of pins having to be at least the voltage drop of a diode. But I can't find that now. here are the voltages of all the transistor pins (is the collector of Q2 lower than it should be?):

                B C E
                Q1 0.88 6.8 0.29
                Q2 0.57 1.12 0.0
                Q3 1.12 8.78 0.68

                And here is my diagram, updated and laid out in a way that approximates to the referenced circuits and with capacitor labels.

                Click image for larger version

Name:	Keeler Kick 0.3.jpg
Views:	1
Size:	55.9 KB
ID:	830871

                I hope this is easier for you guys to follow. It does indeed look like an LPB1 with a tone control, a silicon Fuzz Face and a texture mod to adjust the voltage at the collector of Q3. C9 (between collector and base of Q3) is a noticeable difference compared to the Fuzz Face circuit. The sound from the pedal at the moment is almost clean with a fizzy background (almost like white) noise. Only the volume pot seems to have any significant affect on sound. it changes the volume.

                many thanks for your help - I enjoy building and I enjoy the learning that I get from these pages.

                Jeff
                It's not microphonic - it's undocumented reverb.

                Comment


                • #9
                  As each stage is basically a Class A circuit, my first thought is 'what's up with Q2's collector voltage.'
                  A tad low is it not?

                  Comment


                  • #10
                    R14 applies DC feedback to the base of Q2. As currently set, the resistor values of the Q2 and Q3 stages cause the Q2 collector voltage to be low. If my memory serves me correctly, the Fuzz Face circuit did this deliberately to provide asymmetrical clipping. The resistor values could be changed so as to raise the Q2 collector voltage if that were desired. If you use a signal generator and scope to trace the signal through this simple circuit, you'll find out what's causing your problems

                    Comment


                    • #11
                      I thought that as well, but with my knowledge of transistor operation being very scant, I didn't want to jump to that conclusion. The 33k collector resistor is in the FF design so I assume it's well proven. Because the markings are so hard to read on the tiny 4 band metal film resistors I measure each resistor before inserting, but I can double check by using a lens. Neverthe less, the collector of Q2 should be the same as the base of Q3 as they are are directly connected in this and referred to FF schematics. Should the base of Q3 have a higher voltage for correct operation?

                      Thanks for looking and replying.
                      It's not microphonic - it's undocumented reverb.

                      Comment


                      • #12
                        Originally posted by ReadyTeddy View Post
                        R14 applies DC feedback to the base of Q2. As currently set, the resistor values of the Q2 and Q3 stages cause the Q2 collector voltage to be low. If my memory serves me correctly, the Fuzz Face circuit did this deliberately to provide asymmetrical clipping. The resistor values could be changed so as to raise the Q2 collector voltage if that were desired. If you use a signal generator and scope to trace the signal through this simple circuit, you'll find out what's causing your problems
                        Ah yes, now I see it.
                        thanks RT.

                        Comment


                        • #13
                          Originally posted by paggerman View Post
                          I thought that as well, but with my knowledge of transistor operation being very scant, I didn't want to jump to that conclusion. The 33k collector resistor is in the FF design so I assume it's well proven. Because the markings are so hard to read on the tiny 4 band metal film resistors I measure each resistor before inserting, but I can double check by using a lens. Neverthe less, the collector of Q2 should be the same as the base of Q3 as they are are directly connected in this and referred to FF schematics. Should the base of Q3 have a higher voltage for correct operation?

                          Thanks for looking and replying.
                          The collector of Q2 is almost surely above its Vce Sat voltage, so the transistor should work normally. Consider that it is the current through R14 that supplies current to Q2's base. If R14 were a larger value, then the emitter of Q3 would have to be at a higher voltage to supply an equal current to Q2's base. If the emitter of Q3 is higher, so is it's base voltage, and therefore, so is Q2's collector voltage. Of course, if the emitter or Q3 is at a higher voltage, the current through it will be higher unless its emitter resistor is changed. If the current is higher, then its collector voltage will be lower. If too low, Q3 will go into cutoff. As you can see, the values of the resistors in the Q2/Q3 stage have to be adjusted together for the stage to work properly.

                          Comment


                          • #14
                            Well, thank you everybody. My BC182L stash arrived and swapped them for the twisted leg S9013s. No significant difference. After JM's advice to refer to schematics of the LPB1 and silicon Fuzz Face and ReadyTeddy's advice, I changed out some of the resistors. First I changed the input resistor (R1) from 2M2 to 1M. That made a small improvement to the sound, but still a lot of room for improvement. Then I changed the base resistor (R8) of Q1, also from 2M2 to 1M. Both of these changes follow the LPB1 circuit. I did not change R3 (470k) to the 100k value of the LPB1. Small improvements, but no great shakes.

                            Then I changed R12. In the build diagram it was 4k7. In the silicon fuzz face 330R. The change was huge. Now it sounds like a fuzz effect. I haven't had a chance to play with it properly - with four knobs to twiddle it'll take me a while to dial in the sounds I like most, but I'm pretty happy with it for now.

                            Until next time - thanks everybody!
                            Jeff
                            It's not microphonic - it's undocumented reverb.

                            Comment

                            Working...
                            X