Soldano Circuit Question

C

Cap217

Active member
In this soldano circuit I see a couple things that I understand in practice but not really the science behind it. Im hoping people can help at the expense of me sounding stupid. This isnt necessarily soldano specific but its what they do and the schematic I am linking to.


Cathode Bypass
It seems that soldano started the .1/1.8k that a lot of the high gain builders adopted. I know in my amp when I switch to a .1/1.8k I get a lot more compression and maybe gain than a .68/2.7k. What is actually happening with these caps and resistors? I want to understand why combinations sound different vs just testing and hearing it.

I know there is a calculator but I want to understand what is happening vs just getting results.




Plate Coupling Resistors
Soldano uses the standard 470k with a treble peaker. I started using treble peakers to tame harshness in my amp and I assume they just bleed off frequencies but what size for what frequencies? I dont have a scope but would that really allow me to dial in with data vs my ears here?

It says in this schematic "attenuator dumps 48%" here. What is that saying? 48% of the highs are attenuated? If so, what frequency?

And why is 470k 470k the standard? Is there something about that setup that gives us guitar frequencies we are used to?

In OD Preamp 2 there is a 1m resistor to ground that says it dumps 32%. Again, 32% of what?




Negative Feedback
There is a .1 capacitor after the depth pot. This is between the .0047 on the pot and the resistor. What is this doing and how?
 
I'll address a couple. Keep in mind that many/most times a capacitor works together with a resistor to set a roll-off frequency. But for simplicity I'll most just refer to the cap itself.

Cathode resistors and bypass caps:
Look at the Duncan calculator. They affect gain and frequency response. The resistor sets the bias and affects gain across all the available frequencies of that stage. The bypass cap affects gain in the frequencies determined by the cap value. If it's big enough then it affects gain across all available frequencies. A 0.68uF cap on a 2k7 resistor increases gain across most of the guitar frequencies but not the lowest. If you just increase that resistor to 3k, it will get a little tighter. Because the gain of all those frequencies below that 0.68uF cap just got slightly reduced.

Treble peakers just allow some higher frequencies to bypass the 470k. Increases upper mids/treble and the break frequency is determined by the cap.

That .1uF cap between Depth and Presence could be blocking any DC from the Presence control. That whole NFB circuit can and often does have a big blocking cap in series. The AC still gets thru and that's all that matters there. The lack of it in old Marshalls is why the Presence pot is noisy (Marshall fixed this another way later).
 
Cap217 said:
Cathode Bypass
It seems that soldano started the .1/1.8k that a lot of the high gain builders adopted. I know in my amp when I switch to a .1/1.8k I get a lot more compression and maybe gain than a .68/2.7k. What is actually happening with these caps and resistors? I want to understand why combinations sound different vs just testing and hearing it.

I know there is a calculator but I want to understand what is happening vs just getting results.
Click to expand...

When unbypassed, the unloaded gain of the triode goes up and down with the ratio of the plate to the cathode resistor. (This is a very rough estimate, but gives on idea of what's going on.) So, that lower resistor value naturally gets just a bit more gain with the same plate resistor for the stage.

The bypass capacitor in parallel boosts gain about the cut-off frequency of the combination of resistor and capacitor. It acts as a high-pass filter. Below the cut-off frequency of the filter gain is the same as without the capacitor. Above it, the gain is much higher.

The calculation is f_c = 1/(2 x pi x R x C), so 884 Hz for the first values and 86.7 Hz for the second. That means you're getting more gain on the first numbers, but also shifting the frequency which is seeing the most gain. With the higher value capacitor you'd see more gain at lower frequencies, but it'd probably loose focus. If you want to play with things, you can shift the resistor without the capacitor change, and then step capacitors gradually down and listen to what they do.
 
SpiderWars said:
I'll address a couple. Keep in mind that many/most times a capacitor works together with a resistor to set a roll-off frequency. But for simplicity I'll most just refer to the cap itself.

Cathode resistors and bypass caps:
Look at the Duncan calculator. They affect gain and frequency response. The resistor sets the bias and affects gain across all the available frequencies of that stage. The bypass cap affects gain in the frequencies determined by the cap value. If it's big enough then it affects gain across all available frequencies. A 0.68uF cap on a 2k7 resistor increases gain across most of the guitar frequencies but not the lowest. If you just increase that resistor to 3k, it will get a little tighter. Because the gain of all those frequencies below that 0.68uF cap just got slightly reduced.

Treble peakers just allow some higher frequencies to bypass the 470k. Increases upper mids/treble and the break frequency is determined by the cap.

That .1uF cap between Depth and Presence could be blocking any DC from the Presence control. That whole NFB circuit can and often does have a big blocking cap in series. The AC still gets thru and that's all that matters there. The lack of it in old Marshalls is why the Presence pot is noisy (Marshall fixed this another way later).
Click to expand...


You say .68/2.7k to a 3k and it gets tighter. Im confused because I was looking at a chart that says this:

.68/2.7k = 232hz
.68/3k = 223hz
.68/6.8k = 180hz

By looking at this I would think it would get looser and bigger with more low end coming though. But is this a high pass filter and working the opposite way Ive been thinking?

The reason I was thinking the way I was is because if you increase cap value you get the effect I am talking about. I haven’t really experimented with the same cap and change resistor values much. I’ve done it with caps but usually over the same resistor. So I will have to work on this combo.


Is there a way to determine on treble peakers what cap value over what resistor value = what frequency?



And should I be putting a .1 cap after my depth pots (and before the NFB resistor) as good practice?
 
Last edited:
Cap217 said:
You say .68/2.7k to a 3k and it gets tighter. Im confused because I was looking at a chart that says this:

.68/2.7k = 232hz
.68/3k = 223hz
.68/6.8k = 180hz

By looking at this I would think it would get looser and bigger with more low end coming though. But is this a high pass filter and working the opposite way Ive been thinking?
Click to expand...
Did you check this like I mentioned?
https://www.ampbooks.com/mobile/amplifier-calculators/cathode-capacitor/
There's all sorts of good stuff there.

Regarding the 0.1uF cap in the NFB, I wouldn't worry too much about it. Friedman still uses the noisy 5k Presence because he feels it sounds better.
 
I don’t think many people realize this but there is a negative feedback loop created in a preamp tube circuit, or “cathode degeneration.”

Adding a bypass capacitor to the cathode resistor bypasses the specific frequencies of the cap and shunts them to ground, removing them from the negative feedback and thereby boosting them in the signal.

Putting a .1uF in the depth circuit is unnecessary. It’ll work fine without it.

Attenuators are resistors sending signal to ground, or placed inline to restrict the signal. On a high gain circuit, you must attenuate the signal more than just using a gain control. A gain control is essentially a variable attenuating resistor (voltage divider in this application), sending more or less signal to ground depending on the position of the control. The challenge is to strategically place resistors in the best position. The values depend on how much signal you need to attenuate. A 1M to ground will attenuate a noticeable amount, but say, 68k to ground, will dump a lot more. Opposite for resistors placed in-line. If you put a cap on the resistors to ground, it sends that specific frequency to ground as well, removing it from the signal. If you place them on resistors in-line, you boost those frequencies. You can really dial in an amp this way.
 
Last edited:
RedPlated said:
I don’t think many people realize this but there is a negative feedback loop created in a preamp tube circuit, or “cathode degeneration.”

Adding a bypass capacitor to the cathode resistor bypasses the specific frequencies of the cap and shunts them to ground, removing them from the negative feedback and thereby boosting them in the signal.

Putting a .1uF in the depth circuit is unnecessary. It’ll work fine without it.

Attenuators are resistors sending signal to ground, or placed inline to restrict the signal. On a high gain circuit, you must attenuate the signal more than just using a gain control. A gain control is essentially a variable attenuating resistor, sending more or less signal to ground depending on the position of the control. The challenge is to strategically place resistors in the best position. The values depend on how much signal you need to attenuate. A 1M to ground will attenuate a noticeable amount, but say, 68k to ground, will dump a lot more. Opposite for resistors placed in-line. If you put a cap on the resistors to ground, it sends that specific frequency to ground as well, removing it from the signal. If you place them on resistors in-line, you boost those frequencies. You can really dial in an amp this way.
Click to expand...
Thanks. I got snubber and peaker mixed up here.

So a 1m resistor over a volume pot on a super lead with a capacitor can really dial in a specific frequency removal? Lots of experimenting coming up if so.
 
SpiderWars said:
Did you check this like I mentioned?
https://www.ampbooks.com/mobile/amplifier-calculators/cathode-capacitor/
There's all sorts of good stuff there.

Regarding the 0.1uF cap in the NFB, I wouldn't worry too much about it. Friedman still uses the noisy 5k Presence because he feels it sounds better.
Click to expand...

I use both. I have a 0.02uF in the negative feedback path blocking DC from the long tailed pair phase inverter and also kept the noisy 5k because of the sound importance.
 
RedPlated said:
I don’t think many people realize this but there is a negative feedback loop created in a preamp tube circuit, or “cathode degeneration.”

Adding a bypass capacitor to the cathode resistor bypasses the specific frequencies of the cap and shunts them to ground, removing them from the negative feedback and thereby boosting them in the signal.

Putting a .1uF in the depth circuit is unnecessary. It’ll work fine without it.

Attenuators are resistors sending signal to ground, or placed inline to restrict the signal. On a high gain circuit, you must attenuate the signal more than just using a gain control. A gain control is essentially a variable attenuating resistor (voltage divider in this application), sending more or less signal to ground depending on the position of the control. The challenge is to strategically place resistors in the best position. The values depend on how much signal you need to attenuate. A 1M to ground will attenuate a noticeable amount, but say, 68k to ground, will dump a lot more. Opposite for resistors placed in-line. If you put a cap on the resistors to ground, it sends that specific frequency to ground as well, removing it from the signal. If you place them on resistors in-line, you boost those frequencies. You can really dial in an amp this way.
Click to expand...

I disagree completely. There’s no NFB in a simple text book class A triode stage. It’s the predecessor to enhancement mode NFETs in terms of bias and performance and is no different than how we bias MOSFET stages for gain today. It can be seen as a simple voltage controlled voltage source since the grid is high impedance to the plate and cathode and no DC current flows from grid to cathode.

The cathode capacitor provides maximum gain for only certain frequencies and acts as a high pass filter when in parallel with the cathode resistor. When the AC signal is above the filters 3dB or half power point, maximum gain of the stage is provided. When the AC signal approaches the half power frequency of the filter, the impedance rises, and the valve no longer amplifies the signal at its maximum rating since that AC signal no longer sees a path to ground. Remember that a valve has an idle bias and an AC signal on the grid modulates that bias point up and down, where the modulated signal is then amplified beta times the impedance looking out of the plate. Conversely, without the capacitor, the cathode resistor attenuates all aspects of the frequency spectrum equally and the valve stage amplifies based on its bias and beta rating accordingly.

Technically your stage’s amplification factor is limited by the beta rating of the stage, the plate impedance of the tube, the plate resistor, and also where the quiescent DC bias point is located set by the cathode resistor. Let’s be clear that the cathode resistor’s primary job is to provide the quiescent DC operating point of the stage which it does so by making the cathode a more positive DC voltage than the grid, which is the same as making the grid more negative than the cathode.

So in summary there’s nothing magical happening in a triode stage that fundamentally is different than how we bias transistors today. There’s just additional specifications we have to watch for like heater to cathode voltage requirements to prevent internal arcing and needing to provide power to heater filaments to heat the cathode so that valence electrons are able to leave the cathode to travel to the plate through the grid.
 
Last edited:
Read the article below. This is how I’ve always understood it to work and I still believe it’s correct.

https://www.tropicalfishvintage.com/blog/2020/4/18/on-cathode-bypass-capacitors
Here’s a quote from the article that supports my response:

“However, the presence of a cathode resistor creates a local negative feedback loop. This is called cathode degeneration, and despite the negative connotation of its name, it could be desirable (or not) based on the designer’s goals for the gain stage. Negative feedback decreases gain and promotes linearity. In this case, it also increases the stage’s output impedance. If the circuit benefits from these features, the designer may omit a bypass capacitor from one or more gain stages.”

And here is another from a different article that also agrees with my statements:

“Before we dive in, let's talk a bit about what exactly the cathode bypass cap does, and how changing it can affect the tone of your amplifier. We need to begin with the concept of "cathode degeneration," which is a normal part of tube operation, but can be seen as an unwanted byproduct -- it is also a form of "negative feedback." Negative feedback can notoriously cancel out frequencies in your signal, so in order to combat this cancellation of frequencies, a "cathode bypass capacitor" is implemented.”
 
Last edited:
Degeneration is not a concept limited to tube stages and is simply saying there’s a cathode resistor present to make the cathode more positive than the grid, which is the same as making the grid more negative than the cathode. It actually is used as a linearity concept in MOSFETs as source degeneration. You’re sacrificing gain for linearity, hence the need for a capacitor that is frequency dependent to compensate. It is a common application in cascode topologies and common emitter stages or better yet known as cathode follower stages of vacuum tubes. In source follower or cathode follower stages yes, there’s a loop from plate to grid that is made, but in a normal triode stage, there’s no negative feedback loop to be drawn. It may act like a feedback concept, but there’s no feedback in a class A triode normal amplification stage.
 
That's way above my level of understanding but let me throw this out there...if something "acts like negative feedback" then isn't it negative feedback by definition? To me the term 'negative feedback' sort of describes what it is...how it behaves. It can manifest itself in all sorts of ways. If it's behaving like NFB then it is NFB.

Just a very simpleton way to look at it.
 
SpiderWars said:
That's way above my level of understanding but let me throw this out there...if something "acts like negative feedback" then isn't it negative feedback by definition? To me the term 'negative feedback' sort of describes what it is...how it behaves. It can manifest itself in all sorts of ways. If it's behaving like NFB then it is NFB.

Just a very simpleton way to look at it.
Click to expand...
I think this is an area that terminology gets thrown around with little care and people need to be cognizant. This is an area of speciality for me because it’s by day job. You actually have to be able to complete a Norton or Thevenin loop within the feedback path and it be a complete physical loop in the small signal model, meaning supply and ground are both shorted as small signal ground, and a loop still needing to exist, for it to be true feedback. Just because you can degenerate a gain stage doesn’t mean feedback exists.
 
glpg80 said:
I think this is an area that terminology gets thrown around with little care and people need to be careful. This is an area of speciality for me because it’s by day job. You actually have to be able to complete a Norton or Thevenin loop within the feedback path and it be a complete physical loop in the small signal model, meaning supply and ground are both shorted as small signal ground, and a loop still needing to exist, for it to be true feedback. Just because you can degenerate a gain stage doesn’t mean feedback exists.
Click to expand...
OK, that may be the whole crux of any disagreement. To me, just two wires being close to each and interacting such that they cause a decrease (NFB) or increase (PFB) in gain...which then changes the interaction...which then changes the FB...aso.
 
SpiderWars said:
OK, that may be the whole crux of any disagreement. To me, just two wires being close to each and interacting such that they cause a decrease (NFB) or increase (PFB) in gain...which then changes the interaction...which then changes the FB...aso.
Click to expand...

Taking a signal that is quantifiable and out of phase, injecting it back into a previous stage, is in fact negative feedback if your gain and phase margin say it is stable. If it has a zero phase margin causing positive feedback then you’ve just created an oscillator.

It’s not a hard concept but once again biasing concepts aren’t feedback concepts. To get a NFB local loop you’d actually need to apply a capacitor from plate to grid and inject a signal out of phase back into the grid of the same stage for a local feedback loop to exist.
 
Last edited:
glpg80 said:
Taking a signal that is quantifiable and out of phase, injecting it back into a previous stage, is in fact negative feedback if your gain and phase margin say it is stable. If it has a zero phase margin causing positive feedback then you’ve just created an oscillator.

It’s not a hard concept but once again biasing concepts aren’t feedback concepts. To get a NFB local loop you’d actually need to apply a capacitor from plate to grid and inject a signal out of phase back into the grid of the same stage for a local feedback loop to exist.
Click to expand...
My limited understanding is that if there is no cathode bypass cap, the signal exists on the cathode (in phase with grid) and that's the NFB. It affects that grid-to-cathode voltage and thus current flow and thus gain.

I think that in my case, you are just thinking of it at a higher level. I'm trying to dumb it down to a little box that I can understand.
 
FWIW, you can have actual negative feedback on a preamp tube stage. It will act to stabilize gain and to increase input impedance of the stage. I don't recall seeing it in a guitar amplifier, but I'd expect to see it in a tube hifi preamp or a tube mic preamp, so as not to load passive input devices. See below. This is what I'd think of when referring to a negative feedback loop in a preamp gain stage. (This is for information purposes only, as it doesn't really apply to your Soldano circuit question.)

1648133255461.png

Preamp schematic courtesy of the Valve King website.
 
Last edited:
rstites said:
FWIW, you can have actual negative feedback on a preamp tube stage. It will act to stabilize gain and to increase input impedance of the stage. I don't recall seeing it in a guitar amplifier, but I'd expect to see it in a tube hifi preamp or a tube mic preamp, so as not to load passive input devices. See below. This is what I'd think of when referring to a negative feedback loop in a preamp gain stage. (This is for information purposes only, as it doesn't really apply to your Soldano circuit question.)

View attachment 112773
Preamp schematic courtesy of the Valve King website.
Click to expand...
Dumble does something similar. It's often something like 22M resistance and I think he used a dedicated cap for it. Definitely in the Return stage for a Dumbleator.

The Silver Jubilee (and possibly Wizards too) have something almost if not identical to the above schematic.
 
You must log in or register to reply here.
Back
Top