Announcement

Collapse
No announcement yet.

How to switch an entire gain stage??

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

  • How to switch an entire gain stage??

    In order to make an entire gain stage in/out of the chain, what should be done: just connect the plate of the previous gain stage to the grid of the next stage, shunting the grid of the bypassed one to ground, or anything else?

    The idea here is to make this bypassed stage a switchable tube booster for the gain channel...

  • #2
    You should research how the many "channel switching" amps do it. Most are not channel switching at all but rather an insertion of a gain stage (or two or even three) just as you propose. The basic idea is the same with most. Certain senarios do require special attention and I don't know your schem. Your basic idea is correct , that is, you would shunt the unused gain stage grid to ground. Sometimes it's also possible to shunt the unused stage output to ground. This may work out better for some designs strictly as a matter of convinience with the switching system. Again, I don't know your schem so I can't comment on specifics. I will add that you need to pay attention to a couple of things. One is that when switching you need to have a 0V reference on both switch ends to avoid noise due to a voltage differential. Another is that you must make sure that if you are using a switch on a grid that it never "floats" the grid eliminating the tubes bias condition. Like I said, look up some schems for channel switching amps and you'll see these things in practice. You may find a design that will swap easily into your plan. The switching devices used today are typically LDR's and they are very effective for this sort of thing but if you plan to use relays there are plenty of schems to represent that too.

    Chuck
    Last edited by Chuck H; 10-27-2010, 02:43 AM.
    "Take two placebos, works twice as well." Enzo

    "Now get off my lawn with your silicooties and boom-chucka speakers and computers masquerading as amplifiers" Justin Thomas

    "Being born on third base and thinking that you must have hit a triple is pure delusion!" Steve A

    Comment


    • #3
      @Chuck H
      Thanks mate!

      In my design, the main idea is to have a clean channel (still don't know if it'll be an EF86 based or a Fender Twin/Bassman like) and a british gain channel, with an additional switchable gain stage... I didn't decide yet about the way that these switches will be done (optocoupler, relay, ...), but my intent is trying to do this with relays (mechanical switch)... however, as I never done a "switching system", I'm kind of lost thinking about how to do the work in here, specially with this additional gain stage... for convenience, the topic below is the one that I opened looking for some advice in my preamp concept:
      Ogrus British Design (from the scratch) - need backup!

      Your advices will be helpful, for sure... althought, I still don't know exactly how to procede:
      - to avoid the pops, if using relays (is your idea of keeping 0V reference in both ends enough?);
      - to keeps the tube safe, by ensuring that its grid don't floats.

      Looking the web, I found this schem that shows how to switch an entire gain stage from the chain:
      http://www.lynx.bc.ca/~jc/JCM2chanPr...9%29300dpi.gif
      Can be a good starting point!

      Diogo

      Comment


      • #4
        That channel switching design looks like a suggestion rather than an actual circuit in use in an amp. Perhaps some kind of mod (modification) to a stock circuit. The problem is that this circuit DOES float the grids during switching. This will almost certainly POP when switched and at it's worst, if the switch were to fail it could leave a tube with no bias. The solution is to hang a 1M resistor to ground from the pole of each switch. This provides the switched grids with an uninterrupted 0V reference. The additional load by adding these resistors can be compensated with other circuit changes if you wanted to. But it usually isn't an issue.

        So... You want two seperate preamps and one will have a gain switch, right? How do you want to switch between preamps? You may be planning to have an input for each preamp and choose one, both or the other with an ABY pedal. But keep in mind that if you do this the hiss and noise from the "other" channel will still be present at the amps output. This can be really annoying if your using the "Fender" channel and the "British" channel has the gain boost on. But I have done amps this way and it "works" fine and performs well on stage since you can use different pedals in front of each preamp.

        The amp you propose in the other post (and I suppose this one) is really quite elaborate. There are many simpler designs that will get most of what you want with fewer tubes and layout problems. You would have to sacrifice the Fender/British option though and just have one or the other. So maybe that's no good for you.

        A "British" channel with an extra gain stage can be done with two dual triodes. If you went with an AB763 type "Fender" channel you could do that with one dual triode. Now add a PI tube, a triode for the FXL and probably another triode for mixing and your total preamp tube count could be as low as five. If you omit the FXL and mixing stage you can do it with four preamp tubes. If you keep the FXL and mixing stage and use the Bassman/Twin preamp you'll need six preamp tubes and you'll have an unused triode in the circuit. I don't know why but that always bugs me for some reason.

        A little more discussion... The Bassman/Twin preamp IS a Brittish preamp. Or more correctly, the Brittish preamp is basically a Bassman/Twin preamp. That is, gain stage, gain stage, cathode follower, tonestack and then to the phase inverter. It's worth considering this because in the same amp with the same speakers and the same output tubes and transformer the two preamps might sound very similar. EQ would be the only real difference between them. That's why I mentioned the possibility of an AB763 preamp.

        Also, if you only plan on using one preamp at a time you may not need a mixing stage but just use mixing resistors behind the phase inverter instead. This would allow you to use another triode as another gain stage for the pedal boost. I think this is a good idea because with a dedicated master volume on the "British" channel that master vol will need to be behind the phase inverter. I can tell you from experience that a British type preamp with one extra gain stage and a pre PI master volume won't get you into that uber gain zone with the preamp alone. So you could series two gain stages as the inserted boost and really have some modern gain levels from the preamp.

        Things to consider.
        "Take two placebos, works twice as well." Enzo

        "Now get off my lawn with your silicooties and boom-chucka speakers and computers masquerading as amplifiers" Justin Thomas

        "Being born on third base and thinking that you must have hit a triple is pure delusion!" Steve A

        Comment


        • #5
          That channel switching design is a suggestion to have a Plexi-like channel (when the gain stage is bypassed), allowing to a lesser gain circuit, and a hot rodded JCM800-like channel (when the gain stage isn't bypassed), allowing to a higher gain circuit (as it's a bit different than the stock 800, w/o some of the interstages components used to keep it "calm")... althought, as you pointed, this design does float the grids during switching. But, with the additional load that those 1M resistors to ground provide, maybe (and only maybe) some tweeks in the stage should be done...

          About my project, I might not have expressed myself very clearly... the intent is to have only one input, feeding both channels... then, I can choose which channel I want by switch between them... the boost can be actived only in the drive (crunch, dirty) channel, by another switch...

          In one hand, the trouble that I see with using a mixing stage is that the output of the channels will be out of phase, in the situation where the chosen clean channel is the "Fendery" one, as we have two gain stages (inverting) in it, and three in the British one... think it could be seen better in the attached block diagram... and, as you pointed, both the Fender and the British one are quite similar, as the second was closely based in the first one... but, in this case, different components in every stage, and even different plates voltages, should be applied, making each channel with its own feel...

          In the orther hand, I can forget about the "Fendery" clean channel (one and a half 12AX7), and use only a EF86 to this job (whose block diagram is attached too)... but, in this situation, if I decide to keep the boost into the drive channel, an entire gain stage (half 12AX7) will be lost in the amp (useless), cause four of them (three gain stages, plus the cathode follower) will be used in the British channel and only a half will be used to the boost (not counting the entire 12AX7 for the FX Loop and the entire 12AX7 for the Phase Inverter)...

          So, at this moment, I have the "skeleton" of the amp in my head, but am still trying to put it on paper, to (finally) finish the design of it and begin the building... but, I'm still puzzled about somethings, like:
          - how to switch the channels
          - how to switch the boost in the drive channel
          - how to add a volume level for the boost gain, to avoid the boost/lead to be louder than the drive/crunch
          - how to design both channels to be pedal-friendly (especially for use with fx that I like to use before the drive, like chorus, phaser, envelope filter...)

          Imageshack - britishdesignblockdiagr.png

          Imageshack - britishdesignblockdiagr.png
          Last edited by Ogrus; 10-27-2010, 05:43 PM.

          Comment


          • #6
            Your block diagram already has the switching figured out for you. Use a relay for the contacts shown in the diagram. Same thing for switching in the boost. You can find alot on this topic by search for channel switching and relays on this forum and others. A schematic is your next step, the block diagram is pretty much good to do.

            As for keeping the boost the same volume level, switch in a fixed voltage divider or volume control when as the same time the boost is engaged. You may find you don't need this feature. Prototyping the build will answer alot of questions that all the planning in the world can't answer.
            -Mike

            Comment


            • #7
              @defaced
              Thanks mate!

              Well... just one thing... in both of those block diagrams, I tried to make clearer, for everyone who could be puzzled with my first explanation, what I expect to do into the preamp, and especially all things which are to be changed by the switches, althought I'm not so sure if I picked the right places for them, cause:
              - maybe, bypass the first (and not the second) gain stage could be a better choice for the boost
              - maybe, to chose which channel will be used, I need to bypass the entire chain, and not only select the one that I want simply using the relay at the final of the chain (before the master), as I put in the block diagram

              I'm still studying some designs to "finish" the prototype schematic, but already have some ideas... however, as you said, many questions will be answered only when I build the amp... until there, it'll be only speculations...

              For now, I will need to:
              - keep looking for higher gain amplifier schematics, to close the drive channel
              - decide which clean channel to do: the 12AX7 one or the EF86 one
              - ugrade (after the two steps above) the psu
              - decide how to do the switching system

              Cheers!

              Comment


              • #8
                The location you're switching stuff in and out looks good. Its now time to just build something to fine tune what you're going to do.

                maybe, to chose which channel will be used, I need to bypass the entire chain, and not only select the one that I want simply using the relay at the final of the chain (before the master), as I put in the block diagram
                I've seen it done both ways. You may need to ground the signal coming from the unused channel to keep bleed through to a minimum. This will be a mature prototype question, and maybe something you patch together on the final build. More commentary on it below.

                - keep looking for higher gain amplifier schematics, to close the drive channel
                The Soldano SLO 100 is a bare bones starting block four gain stage preamp, that sounds REALLY good. The Soldano Avenger, which is just the drive channel from the SLO, bypasses the first stage to get a three gain stage preamp (labeled high and low inputs). In contrast, the VHT (now Fryette) Deliverance bypasses the third stage to go from four stages to three. Both sound great, both have the same basic architecture, both sound rather different, both are very different when you look at the details (the Deliverance is pretty complicated for a "single" channel amp). There is a whole forum dedicated to cloning the SLO and TONS of discussion on high gain tube amp designs. Google it, register and read
                -Mike

                Comment


                • #9
                  Hum... if the sites where the switches should be placed seems to be good ones, it's already a beginning (rs)...

                  About the prototyping, is there some way to do this not in the main turret board, but into an external board, made just for tests? Or making the tweaks into the final turret board is the better thing to do? The trouble to begin to prototype is that I'm having difficulties to find some components in here (good caps, turrets, high wattage resistors), so I'm trying to define the basic project, to import the components, build, and so, make all the necessary tweaks in the board... very encouraging!

                  Ah! I'm already a member of the forum that you are talking about... tried to start this thread in there too, but no one answered, until now...

                  Comment


                  • #10
                    Here is a recent thread discussing some ideas on prototyping; how I prototype is included in this thread: http://music-electronics-forum.com/t22155/

                    I wish I could help you on sourcing components, but I only buy parts from companies like Mouser and they usually charge a good bit for international shipping.
                    -Mike

                    Comment


                    • #11
                      Awesome! This thread will be helpful!

                      About the components, I'm trying to find something with the brazilian handmakers, before give the last step: import (which probably won't be so bad, cause everything in here is expensive, because of the taxes)! I'll look into the Mouser's website too (as exists an brazilian site, but with the prices in dollars)... let's see what's happen... the building already taked more time that I expected, so, a little more time wasted won't be too bad...

                      Ah! I forgot, in the last post, to coment about the SLO... I like how it sounds, but I'm seeking for something a bit more british (with crunchy mids)... the SLO is too "clear" in the mids... the idea, as I posted in the other thread (linked above), is to try to achieve a drive like the Iced Earth's one, or like the channel two on the PT-100...

                      Comment


                      • #12
                        Mixing the stages rather than switching out once or the other

                        Originally posted by defaced View Post
                        The location you're switching stuff in and out looks good. Its now time to just build something to fine tune what you're going to do.

                        I've seen it done both ways. You may need to ground the signal coming from the unused channel to keep bleed through to a minimum. This will be a mature prototype question, and maybe something you patch together on the final build. More commentary on it below.

                        The Soldano SLO 100 is a bare bones starting block four gain stage preamp, that sounds REALLY good. The Soldano Avenger, which is just the drive channel from the SLO, bypasses the first stage to get a three gain stage preamp (labeled high and low inputs). In contrast, the VHT (now Fryette) Deliverance bypasses the third stage to go from four stages to three. Both sound great, both have the same basic architecture, both sound rather different, both are very different when you look at the details (the Deliverance is pretty complicated for a "single" channel amp). There is a whole forum dedicated to cloning the SLO and TONS of discussion on high gain tube amp designs. Google it, register and read
                        Can anyone direct me to resource material for " mixing" the control grids of a triode gain stage and a pentode (rather than switching one or the other) and tying the plate outputs prior to the PI? My idea would be to use a dual gang pot, just after the tone stack, one extreme going to the triode, the other to the pentode. Is this even possible?

                        Comment


                        • #13
                          Hi Mudd, and welcome. What you propose shouldn't be difficult at all. I'm at lunch right now so I don't have time to look up and link anything similar until after work. I keyed in to say that your idea for the dual ganged pot has one drawback. You wouldn't be able to get full output from both tube stages simultaneously. It would be all of one, all of the other or both at half gain with variation in between. Since gain stages amplify voltage rather than current you don't get much more gain from two tubes in parallel. So you would actually have the lowest gain when both the triode and the pentode are sharing equal duty. That's if you use a normal dual ganged pot. If you use an M/N pot it would allow you to turn the triode down in one direction, the pentode down in the other direction and both stages at full volume in the center position. M/N pots weren't really available on the retail market for a long time, but Bourns has recently offered their "blend" pot which guitarists are using for pickup blending. I believe it's 500k and I don't think a 1M pot is offered, but 500k should be plenty high enough.

                          Something to consider. I'll check in later and write something about using stages in parallel if whatever I have to offer hasn't already been covered by the time I get back.

                          For future reference, it's usually better to start a new thread for independent subjects. You'll get more responses that way and it makes it easier for others with the same problem/question to locate the information with a search.
                          "Take two placebos, works twice as well." Enzo

                          "Now get off my lawn with your silicooties and boom-chucka speakers and computers masquerading as amplifiers" Justin Thomas

                          "Being born on third base and thinking that you must have hit a triple is pure delusion!" Steve A

                          Comment


                          • #14
                            I suggest not to crossfade/blend between the inputs but rather between the outputs.
                            As long as no grid current is involved I would expect less mutual loading effects when tying the grids (via coupling caps) as opposed to tying the plates.
                            Last edited by Helmholtz; 06-14-2019, 10:24 PM.
                            - Own Opinions Only -

                            Comment


                            • #15
                              Yes, search how Hi Fi amplifier "balance control" is made, or active bass pickup blending.

                              Basically:
                              1) you join grids
                              2) plates are separate, each one goes to its own output volume control.
                              You have a dual one.
                              One half (say, the front one) is wired normal, so, say, pentode goes from 0 to max volume when pot swept 0>10
                              3) triode goes on back half, but wired upside down: 10 is zero volume and 0 is maximum
                              4) you mix pot wipers with a 220k resistor coming from each one, resistor joint goes to next stage grid.
                              5) so now 0 means full triode, 10 means full pentode, 5 means "half and half" and so on.
                              remember to use coupling caps as needed.

                              There are other ways to do it.

                              EDIT: best is a custom made pot, half Log ; half anti Log , but those are hard to find and expensive, so next option is using a dual linear pot, because by definition both halves are symmetrical.
                              Juan Manuel Fahey

                              Comment

                              Working...
                              X