Multiple inputs - question

[EFK]

I have a two-channel amp which has 1 input for each channel, w/ 68K input resistor, and a 3rd input which blends the two channels and which puts the two inputs in parallel for approx. 33-34K. What I would like to do is add the option to have this 34K input resistance on either channel alone, not blended-only. Kind of like a standard bright and normal input layout for each channel. Is there any easy way to do this? I have the room, so I thought about just adding an extra input jack for each channel and mounting a 68K on each in parallel which would then give me a bright and normal input for each channel. My main question, however, is: will this effect the normal 'blend' input? I'd like the keep the @ 34K resistance on that, not take it down even lower. The current 68K input resistors are board-mounted. If I add a second input for each channel, and jumper ground to ground and use a second 68K as a jumper between hots, will this work? This seems pretty darned simple to me but as a novice I'm not sure if it would work or screw something up. Anyone? Thanks!

[Gaz]

are you sure those resistors you mention aren't grid stoppers? so by two channels do you mean bright and dark, or clean and dirty (or high and low)?

what amp do you have?

i'm still kind of confused by what you wrote, but if it's any help i can tell you that a SUNN Model T has bright and dark inputs, which both have high and low inputs as well as a 'blend' input that combines both high inputs. it is the most complex input scheme i've seen. i think that's what your after so you might want to check out that schematic.

[tubeswell]

Take a look at many of the typical classic Fender schematics. The 'hi-lo' inputs give you 34k 'grid stopper' (the two paralleled 68k) and 1M to ground on the 'hi-gain' input, and 68k 'grid stopper' with 68k to ground on the 'lo-gain' input (the 1M is bypassed). Its all achieved with tip switching input sockets. Of course, the way the 'grid stoppers' are wired onto the main board in the classic '50s amp layouts, they are not as effective as grid stoppers could be if they were mounted right at the grid pins of V1.

Maybe you could have a go a posting a schematic of what you think you would like and ask for opinions about how it might work.

[EFK]

It's two separate channels, both are somewhat differently voiced as each channel has it's own 12ax7 and each has it's own tone stack. Each channel has one input but there is a 3rd input which blends the two channels. The 68K are indeed grid stoppers, board mounted. When I use the mix input, these two are in parallel so I'm getting about 34K grid input resistance, correct? What I would like to do is set up a 'classic' Marshall type 4-hole input scheme that doesn't affect that mix input that's already there. Right now, with only 1 dedicated input for each channel, I'm stuck w/ 68K grid stopper only unless I want to blend channels. What I'm trying to do is add the possibility to gain a 'bright' input, or 34K grid stopper, on each channel individually without complicating the current mix input (I want that to remain 68k paralleled with 68K, not 34K X 34K...)

Alternately, there is probably a way I could just add another 68K grid resistor to the board in parallel with each that is already there and put the ground leg on a switch, correct? Then I'd have to add two switches rather than inputs. Is this a viable option?

I will take a look at the Sunn schematic you mention - thank you! I think that might be what I am after as well. This amp is a Gibson GA42RVT although there is no accurate schematic currently available for this amp, unfortunately. Believe me I've tried!

[tubeswell]

It's two separate channels, both are somewhat differently voiced as each channel has it's own 12ax7 and each has it's own tone stack. Each channel has one input but there is a 3rd input which blends the two channels. The 68K are indeed grid stoppers, board mounted. When I use the mix input, these two are in parallel so I'm getting about 34K grid input resistance, correct?

Gibson amps were all over the place with their input arrangements and I can't find a schematic for a GA42RVT either so its hard to say without seeing the schematic, but what you're saying it sounds intuitively right.

What I would like to do is set up a 'classic' Marshall type 4-hole input scheme that doesn't affect that mix input that's already there.

That would be a classic Fender arrangement that JTM copied of course

Right now, with only 1 dedicated input for each channel, I'm stuck w/ 68K grid stopper only unless I want to blend channels. What I'm trying to do is add the possibility to gain a 'bright' input, or 34K grid stopper, on each channel individually without complicating the current mix input (I want that to remain 68k paralleled with 68K, not 34K X 34K...)

See Fender 5E3, 5G9, 5F6A etc input arrangements

Alternately, there is probably a way I could just add another 68K grid resistor to the board in parallel with each that is already there and put the ground leg on a switch, correct? Then I'd have to add two switches rather than inputs. Is this a viable option?

Can you draw a schematic of what you mean?

[HBamps]

i don't believe that you will achieve a "bright input" by means of a lower grid resistor alone. The difference in tone from a 68k and a 34k input (grid) resistor is negligible.

Maybe I am mis-understanding what you are asking.

[pdf64]

I was thinking that also.
From Aiken:-
for a typical 12AX7 stage, 'the total input capacitance would be:
Cin = 2.3pF + (61+1)* 2.4pF = 151.1pF
The input capacitance and output resistance of the previous stage form a first-order low-pass filter with a -6dB/octave (-20dB/decade) slope and an upper -3dB frequency point that can be calculated as follows:
f = 1/(2*pi*Rout*Cin)'

For a 34k grid stopper, this gives a -3dB point at 32kHz, with the 68k at 16kHz.
Any difference between the 68k and 34k inputs is going to be at the limits of human perception.
A couple of weeks ago, while I had it out on the bench, I noticed that the shorting contact on 2nd input of my Fender BF build wasn't working, meaning that the #1 input was on 68k.
I bent the contact to get the switch working, and confirmed that the total resistance was now down to 34k.
The amp didn't sound any brighter, but I was able to button it back up with the warm glow of satisfaction that there was now less resistance in my signal path. Peter.

[EFK]

Hmmm. So there really is not much difference between the hi and low inputs on the same channel (using a classic Fender or Marshall for example...)? I'm aware of the difference between say a 'bright' channel w/ treble cap, cap across mix resistor and/or different cathode bypass cap, but I was under the impression that even given the same channel, the 'hi' and 'low' inputs allowed for a distinctly hotter or less driven input.

To take a step back for a moment then - and go easy on me as I'm still trying to understand all of this - if I have two separate channels both set to same volume (individual volume controls) I can plug into one or into the other and achieve a certain level of perceived loudness and certain level of perceived drive. Now if I plug into the mix input, I am blending those two channels; without touching any of the volume or tone knobs, why does the amp sound somewhat louder (not hugely so, but definitely louder) and possess a noticeably higher level of drive? There is no cascading going on here - all the mix input does is parallel the two channels. My first thought was that the difference could be ascribed to the input going from 68K to @ 34K with both grid stoppers in parallel, i.e the mix input in this situation essentially becoming a 'hi' input while the other two standard inputs are acting as 'low' inputs. But here I'm being told that this really would not make a difference?

[loudthud]

How are the grid stoppers in parallel in EFK's amp when the both input is used? This would require a connection between the grids of the respective channels tubes. Does EFK understand the basics of input impedance and resistive voltage dividers?

Depending on how the two channels are mixed within the amp, feeding both channels could result in a 6dB increase in apparent gain. There is a mod for a 4 input Marshall where one or both volume controls are replaced with a volume control with a switch such that when the volume is all the way down, the channel is disconnected. This eliminates the attenuation at the mixing resistors.

[tubeswell]

My understanding is that the reason there is difference between hi & lo inputs in a classic Fender design is that the 'lo' input has much more attenuation because of the lower signal-to-ground reference resistance (as well as the slightly bigger reduction in high frequencies from the 68K, as opposed to the 34k). Plus the Fender clean stage in V1 is set up to allow a full audible bandwidth of frequencies to be amplified (so if the input is changing that in any way, you notice it more).

(Well anyway that's my theory for now, and I'm sticking to it. I can definitely hear a difference on all my Fender clones).

[EFK]

EFK most definitely does not understand anything you mention, loudthud, and you are way over EFK's head! What I generally understand is 'solder a to b and you will get c.'

Here is a clip from the schematic which Gibson sent me which is for either an earlier version of the amp or just a different amp altogether; however, the input segment seems accurate. Maybe this would explain where the added perceived gain is coming from on the mix input?

Following the 68K grid stopper and the 1M resistor to ground, the input is straight into the 12ax7; V1 for input 'J1' and V2 for input 'J3.' Both 12ax7 are currently biased the same w/ 100K plate, 1.5 shared cathode and 22uf bypass.

[Enzo]

Well, don't confuse this with the classic hi/lo Fender input. I think that is what you are doing.

In the classic hi/lo jack circuit, if you plug into the hi jack, you have the two 68k resistors ending up in parallel with each other and in series with your signal. And there is a 1M to ground.

What I don;t see you taking notice of is that when you plug into the lo jack, your signal runs through ONE 68k resistor to the circuit inoput, but the other 68k resistor runs from that input point to ground. The result is a voltage divider. There are now two 68k resistors in series. One end gets the guitar signal, the other end goes to ground, and the signal for the amp input comes from the point between them. That is a classic voltage divider, and the signal to the next grid will be 1/2 the voltage of the guitar signal.

SO the difference in the hi/lo jacks is not about the 34k versus 68k grid stopper. The difference is the signal level. At 1/2 voltage, it is 6db reduced.

So if you have one jack, and thus just one resistor, it could be 68k or 34k and not make much of a difference. As with so many other things, it is not the value of one part, it is the whole circuit that we must consider.

In your first post you mention the combiner input jack places the two 68k resistors in parallel for a 34k total. But your diagram posted just above does not do this. The classic hi/lo jack DOES do that, but in this diagram, all you have is two independent 68k grid stoppers. They are NOT in parallel.

Why does it sound a little louder when both channels are in use? Same reason two people singing is a little louder than just one. You can set the knobs the same, but unless you build laboratory equipment, the two channels are not going to be identical. Their little differences add up.

[EFK]

Thanks guys, and thanks Enzo - that was very clear and I understand it. The two singers is a perfect analogy.