How to find the peak decibel gain of the highest gain channel of a high gain amp?

[thinkingchicken]

Hi,

According to this comment https://www.thegearpage.net/board/index ... t-30886030 the peak decibel gain of Marshall JVM410H highest gain channel which is the OD2 Channel with the red mode turned on and the gain knob at max is 115 dB. How did he get that calculation? Is it possible to get the peak decibel gain from the highest gain channel from schematics alone? If it is possible, how? Anyway, this is the link to the preamp schematics of Marshall JVM410H https://www.drtube.com/schematics/marsh ... 1-iss2.pdf

[martin manning]

Answer is here: "There are a bazillion Dumble schematics out there. This is just one simulation, with everything maxed (EQ, gain, boost on, internal trimmer and overdrive"
In other words, model the circuit, and let the software calculate the gain in dB.

[thinkingchicken]

Answer is here: "There are a bazillion Dumble schematics out there. This is just one simulation, with everything maxed (EQ, gain, boost on, internal trimmer and overdrive"
In other words, model the circuit, and let the software calculate the gain in dB.

Thanks, can you recommend the software?

[martin manning]

LTSpice is probably the most accessible.

[R.G.]

The other way is to set the controls to however you want to test the gain, apply a sine wave signal at some chosen frequency, then turn the input DOWN until the output is also an undistorted sine wave. Measure and record the output voltage and the input voltage when you have it set up correctly.

The gain is then V(output)/V(input). The gain in DB is then db = 20 * log10(gain).

This may take some doing, as the gain will usually be different at each frequency. Guitar amps are anything but flat frequency responders. Each control affects things too, and this is true for both simulated and measured responses. And it can be hard to measure. A gain of 20db is 10x. 40db is 100x; 60db 1000x. 115db is a gain of G = 10 ^ (db/20) or 562,341 if I punched the right buttons on the calculator. An output of 10V at a gain of 115db means the input is 18 microvolts or so.

On second thought, use a simulator.

[romberg]

Is it possible to get the peak decibel gain from the highest gain channel from schematics alone? If it is possible, how?

It is possible to calculate the theoretical gain of a circuit from the schematic alone. Merlin shows how to calculate the gain of a single gain stage here in the first chapter of his book:

https://www.valvewizard.co.uk/Common_Gain_Stage.pdf

One would calculate how much gain the first stage produces (in theory) then account for the expected gain reduction in the voltage divider or tone stack that connects stage 1 to stage 2. Then rinse and repeat. This would not be impossible to do. But it would be tedious. Thus the recommendation to use a circuit simulator which will do all this math in the blink of an eye. Although entering the circuit into the simulator may be tedious .

Mike

[roberto]

Calculating the gain manually is very time consuming, because at every frequency the AC loadline of each stage is different, and every voltage divider is different (considering the bright caps, miller capacitance, etc...).

[thinkingchicken]

Thank you for these responses. I will definitely try LTSpice as calculating manually will be very time-consuming. Not to mention my math is bad

Is it possible for solid-state amps, especially in the distortion channels, to reach similar level of peak decibel gain like those tube amps or even exceed it?

[wpaulvogel]

The other way is to set the controls to however you want to test the gain, apply a sine wave signal at some chosen frequency, then turn the input DOWN until the output is also an undistorted sine wave. Measure and record the output voltage and the input voltage when you have it set up correctly.

The gain is then V(output)/V(input). The gain in DB is then db = 20 * log10(gain).

This may take some doing, as the gain will usually be different at each frequency. Guitar amps are anything but flat frequency responders. Each control affects things too, and this is true for both simulated and measured responses. And it can be hard to measure. A gain of 20db is 10x. 40db is 100x; 60db 1000x. 115db is a gain of G = 10 ^ (db/20) or 562,341 if I punched the right buttons on the calculator. An output of 10V at a gain of 115db means the input is 18 microvolts or so.

On second thought, use a simulator.

Why do you use an output voltage of 10 volts? I don’t know what the OP is attempting by any of this unless they are wanting to build something with similar gain but wouldn’t the output voltage be more like 40 volts @ 16 ohms (100 watts) or 50 volts @ 16 ohms (156 watts). I’m looking to get a better idea.

[thinkingchicken]

The other way is to set the controls to however you want to test the gain, apply a sine wave signal at some chosen frequency, then turn the input DOWN until the output is also an undistorted sine wave. Measure and record the output voltage and the input voltage when you have it set up correctly.

The gain is then V(output)/V(input). The gain in DB is then db = 20 * log10(gain).

This may take some doing, as the gain will usually be different at each frequency. Guitar amps are anything but flat frequency responders. Each control affects things too, and this is true for both simulated and measured responses. And it can be hard to measure. A gain of 20db is 10x. 40db is 100x; 60db 1000x. 115db is a gain of G = 10 ^ (db/20) or 562,341 if I punched the right buttons on the calculator. An output of 10V at a gain of 115db means the input is 18 microvolts or so.

On second thought, use a simulator.

Why do you use an output voltage of 10 volts? I don’t know what the OP is attempting by any of this unless they are wanting to build something with similar gain but wouldn’t the output voltage be more like 40 volts @ 16 ohms (100 watts) or 50 volts @ 16 ohms (156 watts). I’m looking to get a better idea.

Sorry for not telling the point of my question, just want to see the comparison of peak dB gain between tube amps and solid-state amps.

[roberto]

Alot of SS amps relies on diode/zener clipping for distortion, so the comparison will be unfair.
On top of that, it is not the amount of gain per-se that gives a specific tone, but how it is shaped in frequency, and how hard every stage is hit, together with other stages.

[roberto]

Why do you use an output voltage of 10 volts? I don’t know what the OP is attempting by any of this unless they are wanting to build something with similar gain but wouldn’t the output voltage be more like 40 volts @ 16 ohms (100 watts) or 50 volts @ 16 ohms (156 watts). I’m looking to get a better idea.

looking at what he wrote, I'd say that he (but everyone here) is talking about the preamp only.
The reason why he used 10V on output is, I guess, that above that it is easy to have folse information due to compression/saturation of some gain stages.

[thinkingchicken]

Alot of SS amps relies on diode/zener clipping for distortion, so the comparison will be unfair.
On top of that, it is not the amount of gain per-se that gives a specific tone, but how it is shaped in frequency, and how hard every stage is hit, together with other stages.

So when the gain knob on the highest channel, which is the OD2 channel with the red mode turned on if the amp we're using is Marshall JVM410H, on maximum, actually other gain stages other than OD2 also get affected that they will add another decibel increase to the peak decibel gain of OD2 channel?

[R.G.]

Why do you use an output voltage of 10 volts? I don’t know what the OP is attempting by any of this unless they are wanting to build something with similar gain but wouldn’t the output voltage be more like 40 volts @ 16 ohms (100 watts) or 50 volts @ 16 ohms (156 watts). I’m looking to get a better idea.

To prevent clipping and compression from lowering the measured gain. In general, an amp is a chain of gain stages, with attenuation (tone and volume stages, mixers, etc.) between them. If one stage in the middle overloads or clips, the output in that stage is limited to that level, and the voltage from there on will not increase. It gums up the works of measuring what the actual, non-clipped/limited gain really is.

I did not know what the OP was asking either. I suspected that he might have been asking something like how much distortion he could get in an amp. There is this long-standing confusion in the pedal effects world confusing gain and distortion. It's common for someone new to the electronics to ask how to get a higher gain pedal when what they want is actually more distortion, having heard the common description that way on the facts-lite internet. It takes a while for people to learn that gain is not equal to distortion.

Sorry for not telling the point of my question, just want to see the comparison of peak dB gain between tube amps and solid-state amps.

That's another of those seemingly simple questions with only complex answers if you want real, accurate answers. The real answer is that the gain of all amplifiers of any kind is limited by their signal-to-noise ratio, bandwidth, and construction. If you just keep piling on gain stages, there comes a point where some of the output signal leaks back into the input(s) and makes the thing oscillate. Oscillation is infinite gain - an output but with zero input. The common wisdom among professionals is that everything will oscillate if you just keep adding more gain stages.
So construction becomes ever more critical in keeping unintended feedback from happening by isolation and shielding. Radio amplifiers (tube and SS) very commonly have much larger gains than audio amps because the signal from the antenna may be as much as -140db that has to be amplified back up to usable levels. They can only do this because they are very, very carefully constructed to limit unwanted feedback, and because they filter their input down to a limited bandwidth.
Signal to noise gets into this because every real electronic device has inherent noise. Every resistor has some thermal noise, and the resistive part of every inductor and capacitor has thermal noise too. It's everywhere.
Every amplifying device has built-in noise as a side effect of its operation. If a stage has a gain of for instance 30db, its internal noise is
Every amplifying stage amplifies the noise it is being fed on its input, so any input noise is amplified by the same amount as the signal through the rest of the amplifier. Eventually, you lose small input signals in the noise. The resistors and amplifying device noise puts a limit on how small an input you can amplify and still have anything resembling the signal at the output. This can be blocked out to some extent by limiting bandwidth, blocking noise above and below the desired signal, but it's not much help with audio and its wide bandwidth.
So adding more real gain (that is, signal out divided by signal in) means adding pickier and pickier construction to be done, and paid for. Adding more gain is simpler and cheaper with smaller solid state devices. It's more likely to be POSSIBLE to build higher gain amplifiers with solid state than tubes, but whether it actually gets done in production guitar amplifiers is ... speculative. It's complicated.

looking at what he wrote, I'd say that he (but everyone here) is talking about the preamp only.
The reason why he used 10V on output is, I guess, that above that it is easy to have folse information due to compression/saturation of some gain stages.

Actually, I meant the whole amp, out to the speaker terminals, but it could be done either way. I used 10V because I was pretty sure the speaker terminals could swing that much, and it was handy. Depletion mode tubes (that is, all the ones used in guitar amps) in general can only amplify grid signals as big as the swing from 0V to the grid cutoff voltage. There are exceptions, but this is a general rule. For the common output tubes, this is in the range of forty to fifty volts swing at the grid. For P-P, the grids can swing up to about twice this, but in practice, they're biased to overlap some, so it's more like a 60V p-p signal at the power tube grids before the output tube clips. This is kind of constant, and the OT makes up the rest of the gain/loss to the speakers. Each stage, going backwards from the power amp, you're limited to the grid swing for getting actual gain, not "gain", meaning distortion. Also going backwards from the power amp, you run into the gain lowering effect of volume controls and tone controls. Tone controls often mean a 16-24db loss, so you have to insert this much gain after the controls to make up for the losses in the chain, too.

All this stuff has to be considered in the answer to "how much is the gain of a tube amp?"

[thinkingchicken]

Sorry for not telling the point of my question, just want to see the comparison of peak dB gain between tube amps and solid-state amps.

That's another of those seemingly simple questions with only complex answers if you want real, accurate answers. The real answer is that the gain of all amplifiers of any kind is limited by their signal-to-noise ratio, bandwidth, and construction. If you just keep piling on gain stages, there comes a point where some of the output signal leaks back into the input(s) and makes the thing oscillate. Oscillation is infinite gain - an output but with zero input. The common wisdom among professionals is that everything will oscillate if you just keep adding more gain stages.
So construction becomes ever more critical in keeping unintended feedback from happening by isolation and shielding. Radio amplifiers (tube and SS) very commonly have much larger gains than audio amps because the signal from the antenna may be as much as -140db that has to be amplified back up to usable levels. They can only do this because they are very, very carefully constructed to limit unwanted feedback, and because they filter their input down to a limited bandwidth.
Signal to noise gets into this because every real electronic device has inherent noise. Every resistor has some thermal noise, and the resistive part of every inductor and capacitor has thermal noise too. It's everywhere.
Every amplifying device has built-in noise as a side effect of its operation. If a stage has a gain of for instance 30db, its internal noise is
Every amplifying stage amplifies the noise it is being fed on its input, so any input noise is amplified by the same amount as the signal through the rest of the amplifier. Eventually, you lose small input signals in the noise. The resistors and amplifying device noise puts a limit on how small an input you can amplify and still have anything resembling the signal at the output. This can be blocked out to some extent by limiting bandwidth, blocking noise above and below the desired signal, but it's not much help with audio and its wide bandwidth.
So adding more real gain (that is, signal out divided by signal in) means adding pickier and pickier construction to be done, and paid for. Adding more gain is simpler and cheaper with smaller solid state devices. It's more likely to be POSSIBLE to build higher gain amplifiers with solid state than tubes, but whether it actually gets done in production guitar amplifiers is ... speculative. It's complicated.

Ok, so regardless of how much the sound is distorted, can we achieve a peak decibel gain of say 115 dB or more on the highest gain channels of solid-state amps? If it is possible or impossible, what is the reason?