Understanding the schematic

[mcrushing]

Hello, all!

This'll be my first post, so my apologies if this question has been asked/answered elsewhere already.

I'm an enthusiastic newb to DIY audio. I can solder and have upgraded household circuits without incident, but since I have only a beginner's knowledge of audio electronics, I've been spending time with the manual before Crack kit ships... as well as trying to better understand what those electrons are getting up to while they move through the circuit.

I think I correctly understand how a triode works: The heater causes the cathode to emit electrons; the B+ voltage attracts those electrons to the plate; current flows thru the tube and gets modulated by AC signal applied to the grid. But something I noticed in the Crack's diagram has me scratching my head:

On most triode schematics I've seen, the lead from the power tube to the output jack is connected between the B+ rail and the tube's anode. On the crack, that output lead is connected between the cathode and ground rail. Why is this? Does it not matter where that lead connects, as long as current is flowing in the tube and a coupling capacitor blocks DC at the output?

Or is there some other reason? Thanks for any insights!!

[Paul Joppa]

The output tube circuit is a "cathode follower" circuit, which has a voltage gain less than 1 but much higher current and lower output impedance, enabling it to drive headphones without an output transformer.

[mcrushing]

Wow, thanks for the quick response! As expected, the answer leads to more questions.

Some light Googling reveals a number of articles on cathode follower circuits from guitar-oriented sites. From what little I know, I focused on descriptions of AC-coupled versions (figuring any hifi signal is AC, correct?) A lot of what I initially found quickly went over my head.

This may sound naive, but is a cathode follower essentially an upside down voltage gain circuit? That would track with what you said about the voltage gain being < 1 … (assuming “1” means V in = V out). If so, would it be accurate to say all Crack’s voltage gain comes from the 12AU7 stage while the 6080 provides current gain to overcome the headphone’s impedance?

It’s becoming clear that I need to study a lot more in order to even ask questions…but I appreciate any time you’re willing to spend schooling me.

[Paul Birkeland]

This may sound naive, but is a cathode follower essentially an upside down voltage gain circuit? That would track with what you said about the voltage gain being < 1 … (assuming “1” means V in = V out). If so, would it be accurate to say all Crack’s voltage gain comes from the 12AU7 stage while the 6080 provides current gain to overcome the headphone’s impedance?

This is not as over your head as you think it is apparently!

[mcrushing]

That’s very kind of you to say.

I’m always wary of things I learn from Google university, so I appreciate your willingness to help with my gaps and misconceptions - this forum really is as good as advertised. My goal in going this route is to learn as much as possible along the way - and because I really am a newb and I have many questions for if/whenever you feel like answering them…

What’s the purpose of the resistor that’s wired in parallel between the output lead (after the coupling cap) and the ground rail?

Does this amp (or any otl amp) use feedback? Also, what exactly is feedback in an amp, what makes it global vs local and why does it have a polarity?

Where does harmonic distortion come from? Why do SET heads wax poetic about the 2nd harmonic while the dudes on Audio Science Review say we’re nuts?

What does it mean to say an amp is oscillating? What causes it?

Again, thanks so much and don’t feel obligated to explain all this. Oh, and happy holidays!

[Paul Birkeland]

What’s the purpose of the resistor that’s wired in parallel between the output lead (after the coupling cap) and the ground rail?

Without those resistors the output side of the coupling caps will float up and present a high DC voltage at the headphone jack, which could damage your headphones or shock you when you plug them in.  The jack is a shorting jack so we don't absolutely rely on those in the 1.1 version like we did in the 1.0 version, but they are still necessary for safe operation in my opinion. 

Does this amp (or any otl amp) use feedback? Also, what exactly is feedback in an amp, what makes it global vs local and why does it have a polarity?

There are many different kinds of feedback.  Cathode followers themselves function from local degenerative feedback.  A global feedback loop will be a connection from the output of the circuit all the way back to the first stage, which is not a type of feedback we use in our products (except the Tode guitar amp we offered some years back).  I suppose you could say there's a polarity, as both positive and negative feedback are used at times, though positive feedback isn't something I run into in very many designs.

Where does harmonic distortion come from? Why do SET heads wax poetic about the 2nd harmonic while the dudes on Audio Science Review say we’re nuts?

Distortion comes from non linearities in the devices we use.  You can think about linearity as sending 1V of signal into a tube and seeing 10V come out at the plate, then sending 10V of signal in should get you 100V coming out at the plate.  This would be a really linear devices, but tubes certainly are not perfect devices and there are deviations from this ideal, and these variations are usually asymmetric, meaning that your device may swing down 50V and up 45V.  The directly heated single ended triodes we use in our power amplifiers do a really admirable job of amplifying linearly, and we are careful about how we design our driver circuits so they contribute as little deviation as possible.  For the bulk of amplifier designs, whatever appears at the speaker jacks is routed back to the input stage through a voltage or current divider in order to correct for issues like these, and thusly you can drive THD way, way down.  Some designs will use both local and global feedback (the Mac tube power amps are a good example of this, as is the HK Citation II).

In a circuit like the Crack, the 12AU7 just has to drive the grid of the 6080, which is almost like driving nothing at all.  With the Speedball upgrade, the 12AU7 is loaded about as well as it can be and we really get all the linearity we can out of that tube. 

As far as ASR goes, there will definitely be people who think the zero feedback SET products are a little crazy, but we go out of our way to provide honest measurements and there are plenty of 0.001% THD power amps out there if that's what someone is into.  Half the fun of something like the Crack is building a super old school point-to-point kit, and that's just not an experience you can get from soldering an op-amp, a few resistors, and a few capacitors to a PC board.

What does it mean to say an amp is oscillating? What causes it?

An oscillating amp is one where with no signal present at the input, there is some kind of periodic signal appearing at the output.  There are lots of reasons this can happen and for a circuit like our Mainline, there are lots of parts that appear to be superfluous that are there to prevent this from happening.  Since you have asked about feedback and polarity, there are mechanisms which can cause negative feedback to become positive feedback, and this will cause an amplifier to oscillate.  If you take a push-pull amp with global negative feedback and you hook the primary of the output transformer up backwards, you will convert the negative feedback to positive feedback and get a nasty howl!

[mcrushing]

Wow, thanks for all of this info. I’ll dig in/read up on a few of these answers over the long weekend. I don’t think I clearly understood the concept of linearity in gain stages and you explained it really simply, which helps a lot. Looking forward to my build and surely more questions will arise. But thanks again and have a merry Christmas Paul!

[mcrushing]

Hi, Paul.

After some further study, I have some follow-up questions for whenever you have time.
 
Input stage question:
I found the Youtube channel of "Uncle Doug," a retired teacher who repairs guitar amps, does great explainers (and has entertaining pets). He's got a series on grid leak, stopper, cathode bias, and plate resistors that's really informative, but I don't see leak or stopper components in the Crack's input. It occurred to me that the volume pot might actually serve those purposes, with the part of the resistor that's on the input side of the pot's contact as the stopper, and part that's on the ground-referenced side as the leak.

1. Am I right?
2. If so, don't the values change when the volume does? 
3. If not, why no grid leak or stopper in the circuit?

Output stage questions:
Online examples of cathode followers always explain how they function as preamp outputs or drivers for tone/effects loops in guitar amps. Applied as the output of an OTL headphone amp, there are still some things I just can't wrap my head around...

First, that output resistor. You said:

Without those resistors the output side of the coupling caps will float up and present a high DC voltage at the headphone jack, which could damage your headphones or shock you when you plug them in. 

I racked my brain trying to understand how DC gets onto that jack if I remove that resistor (don't worry, I won't leave it out of my build). It might be that I'm misunderstanding how the coupling cap functions. I *think* DC volts see a capacitor as an open switch and AC volts see it as a closed switch. And if that's the case, there's no DC continuity between ground, the jack, and anything with +VDC on it. What am I missing?

Second, (since I've now learned enough to become dangerous), I have more detailed questions about how a cathode follower works.

You said I guessed right about the 12AU7 stage being a signal voltage amplifier, with the 6080 amplifying current in order to lower the overall output impedance so those signal volts can drive the cans. I also noted the current gain stage looks like an inverted version of the voltage gain stage. But more I learn, the more I notice dissimilarities... First, there's nothing that looks like a grid leak or stopper resistor for the 6080. Is that because the 12AU7 is serving those functions? If so, does the 6080's leaking grid current result in a slight decrease in DC volts on the 12AU7's plate? Or does the whole 12AU7 circuit act as a signal ground reference for the 6080? 

The reason I ask is because my understanding of triode operation - at idle - is that bias current flows through the tube because the plate has a very high + DC voltage with reference to the cathode, while the cathode has a slight + DC voltage with reference to the grid (which has no volts at idle because it's referenced to ground via the leak resistor). Then...

the 12AU7 just has to drive the grid of the 6080, which is almost like driving nothing at all.

WHEN THAT HAPPENS.... why is amplified current available on the cathode side of the 6080?
AND...is the 6080's cathode resistor ALSO essentially acting as plate resistor as far as the VAC signal is concerned?
AND... wouldn't that also essentially make the coupling capacitor for the output double as a bypass capacitor for the 6080 cathode resistor?!

Recognizing that I don't actually need to know any of this to build the kit, I TRULY want to thank you for your efforts to satisfy my curiosity.

Cheers, Mike

[Paul Birkeland]

Yes, the volume pot serves as the grid leak.  The grid leak value does change, but it's always low enough not to cause problems.

You could certainly add stoppers, and we use them in our higher gain circuits with high transconductance tubes (Eros, Reduction, Mainline, Kaiju, Stereomour). 

For the coupling cap, if you start with the + end of the cap at 0V and then apply 100V to the + end of the cap you will need to complete the circuit with the - end of the cap in order for the cap to charge.  If there's no charge on the cap, there's no voltage across the cap, and the - end will be at the same voltage as the + end.

For the 6080 grid leak question, the circuit is directly coupled, so the 12AU7 provides bias voltage to the grid of the 6080 and a grid leak resistor would tend to bleed this off.  The small amount of grid leak current is well swamped out by the 12AU7s operation.

I think amplified current might not be the most intuitive way to think about what's going on with the 6080.  If we consider a 300 ohm pair of headphones that want 5V of signal, that would be 17mA of current (all in RMS), so our output stage must be able to swing this much current to drive the headphones.  The ~3mA of current in the 12AU7 isn't enough, so the first order of business in the output stage is running enough current. 

The 3K cathode resistor on the output stage is indeed similar to a plate loading resistor in functionality and it is in parallel with the headphone load (which is part of why you want the Speedball upgrade). If you made the 6080 output stage with the 3K resistor hooked up from B+ to plate and added a coupling cap, you could make a different amp that would still be functional.  Without the local degenerative feedback of the cathode follower, this alternate circuit would have much higher output impedance that would need to be addressed.

For the coupling cap to act as a cathode bypass cap, the - end would need to be at AC ground. 

[Doc B.]

Not sure if this was mentioned further back in the thread, but the original idea behind the cathode follower was as a less expensive substitute for an output transformer - which also converts high impedance to low impedance in order to drive current into a load.

Regarding the lack of a grid resistor between the plate of the 12AU7 and the grid of the 6080 you might want to do some study of direct coupled circuits.

http://www.valvewizard.co.uk/dccf.html

[mcrushing]

Thank you all so much. You guys really helped me get to a point where I'm (at least somewhat) confident I understand what every component of the Crack is for, and can understand - or at least ask the right questions - about the schematics of other amps.

(That said, I'm sure you'll see me back on this thread once I decide to add the speedball )

Also, thanks for turning me on to the Valve Wizard, Doc. I'll definitely spend some time on that site, maybe even buy the hifi book. I'll be lurking on this forum as well... you've built a great resource and community here. Til then, looking forward to building my amp!