Questions

[THChrist]

Hi, i have this very odd schematic and i wanna know just 3 things. first, what the hell is this symbol?

L7


second, how much is the total resistance of the input, so i can calculate the total gain (just for the second op-amp)



i'm guessing, first, with the switch open, is:
Parallel R52 & R53,
Parallel R50 & R51, in series with R92
and both values in parallel, right?

third, wanna know the gain of the first and second amp (asuming pot is at max)




and final, any of you knows 'bout any good poweramp (just the IC), i think is called chip-amp, which can delivers 100 watts?

[teemuk]

L7 is some sort of an inductor (inductance 250 micro henries) to suppress RF frequencies. Looks like a dual one (two inductors) but I'm not sure whether it's a so-called "common mode choke" (inductors share the same core) or just two individual inductors.

The second question (about) gain. You are correct in how you derive the gain but you also need to consider the AC coupling. At first, I think the branch that is in series with the 1Meg resistor can be pretty much ignored (due to its high resistance). In the upper branch the gain is mostly that defined R53 but at higher frequencies (when C103) begins to couple the signal the parallel connection of the resistors begins to matter more and more. So, the gain is not constant throughout the bandwidth. The maximum amplitude is naturally limited by the LEDs.

Third question: Again the gain of the first amplifier stage varies throughout the bandwidth. Mostly it's R57 / R43 +1 but at higher frequencies the parallel branch of R44 again begins to have effect and increases the gain. At very high frequencies C81 in turn decreases the gain.

The gain of the following stage varies according to switch position. In clean mode (and pot at max) it's about R65 (270K) / R80 (33K). I think the switch labels should be reversed (it's clean in the portrayed position). In distortion mode the gain is approximately R65 (270K) / R88 (470R) but the maximum amplitude is naturally limited by the clipping LEDs that are no longer in series with the high-value resistance. All these are naturally just approximate values that are somewhat modified by the lesser parallel branches.

If you really want to completely figure out the behaviour then I feel it's best that you get acquainted with some kind of circuit simulator software. I, for one, hate calculating all the stuff involved with pen-and-paper method.


Concerning the final question: I can't give you an answer but I recommend that you search for a while and see what different manufacturers offer. The TDA2794-might be an option but circuits based on it are rather complex and I have read some stuff about that chip being quite unreliable in many of them. TDA7294 is what many Marshall amplifiers use (that including the concerned MG100 model) but Marshall amps also have a lot of issues due to these concerned chips. I don't know if other chips in the 100W power range exist but you can always use some lower power ones in bridge mode or something.

[THChrist]

Third question: Again the gain of the first amplifier stage varies throughout the bandwidth. Mostly it's R57 / R43 +1 but at higher frequencies the parallel branch of R44 again begins to have effect and increases the gain. At very high frequencies C81 in turn decreases the gain.

The gain of the following stage varies according to switch position. In clean mode (and pot at max) it's about R65 (270K) / R80 (33K). I think the switch labels should be reversed (it's clean in the portrayed position). In distortion mode the gain is approximately R65 (270K) / R88 (470R) but the maximum amplitude is naturally limited by the clipping LEDs that are no longer in series with the high-value resistance. All these are naturally just approximate values that are somewhat modified by the lesser parallel branches.

If you really want to completely figure out the behaviour then I feel it's best that you get acquainted with some kind of circuit simulator software. I, for one, hate calculating all the stuff involved with pen-and-paper method.

i thought when in clean mode, gain should be:
parallel VR1, R85 and R86 / parallel (R87 + R88) R89

and when in distortion mode, just omit R86 and R87.

but i have no idea on how to calculate gain with that network (i can't even calculate ProCo Rat's Gain bcos of the resistors in parallel)

lets just omit the leds, and it will be easier to me.

BTW, thanx for your answers and your patience.

[teemuk]

Yes, you are right (except that you should also add the LEDs in series with R86). However, I already said that the equations I stated were only approximate. When you consider what is the total resistance of, for example, R85 in parallel with the gain pot (1Meg at max setting) you see that high resistance values do not really alter the total resistance much. Same thing is true also with R89 versus the series connection of R87 and R88 and with the 2M2 resistor in series with the LEDs. Why make things more difficult than what they are especially since these figures do not even consider all the coupling capacitances involved, which also effect the gain a lot. If I want to know this stuff in detail I rather resort to circuit simulator than to a series of complex equations.


I also browsed through some of my Marshall schemas: The dual inductor is indeed a common mode choke, likely CM08RB used by many other Marshall amplifiers. The capacitors with "angled wires" going to them are three-terminal EMI suppression filters. Likely type DS306. Basically they're just "dual capacitors" but due to the three-leaded structure they have lower series inductance, which makes them better for the concerned application.