S.E.X. or Crack2A - which is better for learning and experimenting?

[Deke609]

Thanks PB. That's a good article.

[Deke609]

Sorry - I have another question/problem.

As I "understand" it (I am taking some liberties in using that term), the differential use of the 6J6s results in a complete signal, but with the first half of every waveform being inverted - so you end up with 2 camel humps, instead of a sine wave.  So something needs to be done to fix this.

Are the OTs being used "backwards"? I.e., the secondary is used as the primary, and vice versa?  And the B+ wire acts as a center tap that inverts 1 half of the AC signal putting the the whole signal back into phase?  If so, is that the only purpose of the OTs?  Or have I completely missed what's going on?

Many thanks,

Derek

[Paul Birkeland]

As I "understand" it (I am taking some liberties in using that term), the differential use of the 6J6s results in a complete signal, but with the first half of every waveform being inverted - so you end up with 2 camel humps, instead of a sine wave.  So something needs to be done to fix this.

You get one sine wave on one plate and one sine wave on the other plate, but out of phase.  There's nothing to fix.

Are the OTs being used "backwards"? I.e., the secondary is used as the primary, and vice versa?

The nomenclature of "primary" and "secondary" windings gets into a gray area with a part like the 124B. 

And the B+ wire acts as a center tap that inverts 1 half of the AC signal putting the the whole signal back into phase? 

No phase inversion is done in the output transformer.  It is feeding B+ to each plate (DC) and acting as a load for each plate (AC). 

If so, is that the only purpose of the OTs?  Or have I completely missed what's going on?

The 6J6 has too much gain to be useful as a linestage (38x or 31dB approximately).  The combination of the differential circuit and the output transformer dramatically cuts down the gain even more.  The output impedance of a 6J6 on its own is also a bit higher than you'd want for a linestage, but the step-down ratio of the output transformer takes care of that for you as well. 
Many thanks,

[Deke609]

Thanks PB.

You get one sine wave on one plate and one sine wave on the other plate, but out of phase.  There's nothing to fix.

Hmm. I got that completely wrong. So you get the original signal and its inverse superimposed on it? Doesn't that muck with sound?

[Paul Birkeland]

If you had the original signal and its inverse superimposed, then you'd have no signal at all. 

[Deke609]

Yeah, true! So are you saying that the signal from both cathodes are in phase with one another, but out of phase (inverted) as compared to the original?

Edit: I think I see what you mean about using the differential circuit to keep the gain down. I had been thinking (without actually checking) that the 6J6 was grid-biased with -ve DC voltage, but now see that it is cathode-biased in an intentionally weird way to lower the voltage gain.  Whereas a capacitor would normally be added to bypass the cathode resistor to send AC signal on the cathode to ground, in this circuit it is intentionally left out with the result that the cathode acts as a signal source and lowers the output the tube in doing so -- I don't yet understand the "mechanics" of the power lowering effects -- but only that this is something that one usually seeks to prevent by adding a capacitor bypass. This seems a very strange amp.  Is there a sonic purpose to its design? Or was it born as a result of someone smart having a two 124bs and two 6J6s kicking around and deciding to see if they could make something work? Either way, it seems pretty unique and clever.

[Paul Birkeland]

Yeah, true! So are you saying that the signal from both cathodes are in phase with one another, but out of phase (inverted) as compared to the original?

There's only one cathode in a 6J6.

I had been thinking (without actually checking) that the 6J6 was grid-biased with -ve DC voltage, but now see that it is cathode-biased in an intentionally weird way to lower the voltage gain.

It is biased with negative grid voltage.  The grid is at DC ground and the DC current of both plates flows through the bias resistor, so the cathode is at a positive DC voltage with respect to grid.

Whereas a capacitor would normally be added to bypass the cathode resistor to send AC signal on the cathode to ground, in this circuit it is intentionally left out with the result that the cathode acts as a signal source and lowers the output the tube in doing so

This is the case for the common cathode stage, but a little more complicated for the differential circuit.  If you bypass the cathode bias resistor in a differential circuit, it's no longer differential (the side with the grounded grid is no longer acting on the signal, but rather just sitting there sipping down power supply current).

-- I don't yet understand the "mechanics" of the power lowering effects -- but only that this is something that one usually seeks to prevent by adding a capacitor bypass. This seems a very strange amp.  Is there a sonic purpose to its design? Or was it born as a result of someone smart having a two 124bs and two 6J6s kicking around and deciding to see if they could make something work? Either way, it seems pretty unique and clever.

I don't recommend using "power" in the discussion of preamp design, it can get to be a problem.  I'm not so sure how the design originated, but it's a really easy circuit to lash together and listen to (other than the lack of phase indicators on the 124B, but you will clearly hear if the channels aren't in phase and you can flip one pair of output wires from one transformer around).

[Deke609]

Thanks PB - Sorry, my message was sloppily worded - this is all new terminology for me. It's going to take me while to get the lingo down, and eons to become proficient with the math! I'm not going to tackle the math until after I get the amp built - for now, I'll just trust that the values are correct and focus on trying to understand how the different components within blocks of the schematic function together.

Edit: And just to be clear: I think you meant to write that the cathode becomes positive relative to the grid.

[Paul Birkeland]

Oops, yeah, fixed that typo.

The problem with talking about power and preamp is that something like the BeePre will push a lot of voltage into loads from 1.2K to 100,000K.  This makes nailing down any kind of "power output" pretty difficult.

[Deke609]

Recommended prototyping chassis material?

@PB - the last of the parts for the Nickel Wonder just arrived, and I need to decide on a cheap chassis. I don't want anything expensive or difficult to work with since I want to be able to move things around and add/remove components.  I am leaning towards perforated pressboard - the stuff with holes every inch or so that you sometimes see in stores for displaying/hanging wares.  I figure I can fold up some copper sheet and use that as chassis ground since it looks like the schematic calls for a ground plate to complete some circuits.  Would this potentially work?

If that's a no go, my fall back is to use aluminum baking sheets - but cutting holes in this will be a little more involved. So I'd like to avoid metal if possible.

Your thoughts?

Many thanks,

Derek

[Doc B.]

Screw t strips down to a piece of plywood.

[Deke609]

Thanks Doc! I have a bunch of plywood scrap kicking around that would work nicely.

Full metal terminal blocks or the kind with metal tabs and mounts that BH kits use?

Edit: I am assuming that "t strip" stands for terminal strip and not t-profile aluminum used for carpeting, tile, etc.

[Deke609]

Assuming that Doc meant that I should screw terminal strips to the plywood, should I add a copper strip to act as chassis ground, or just solder all ground terminations to a common lug that is then wired to the ground of the IEC inlet?

I ask b/c I don't know whether surface area or mass of the chassis ground is important.

many thanks,

Derek

[Paul Birkeland]

Try a cake pan.

[Deke609]

Ha! Thanks PB. That would work nicely for a permanent build, but I think it would make it difficult to experiment with different components and OT positioning relative to PT.

So I like Doc's idea of using plywood to prototype and experiment.  What I don't know is whether it's better to add a copper strip as ground plate or simply wire all connections to ground to a common grounded lug.