Changing Marshall bass to one channel only? Shared/Split Cathode

[shocki]

Hey guys,

I would like to build a 50W Marshall Bass but with just one channel.
Basicly I will use the following schematic.

http://www.drtube.com/schematics/marshall/1986u.gif

1) I think I can get rid of the two 440K mixing resistors
2) I am not really shure about the shared cathode first stage. Should I just use a part of the first preamp tube (so actually split) or should I keep the shared cathode thing? I think I have to go for 1.5k to archive the same bias with split cathode.

I think soundwise it will be actually the same thing?

Thanks alot

[J M Fahey]

Build it as-is
Otherwise, what will you do with the now unused triode?
You are saving nothing, both in $$$ and in time.
Ok, you are saving 2 jacks and a pot, big deal

[shocki]

It is not about the money. I want to use a smaller chassis and I have no need for the second channel. Also I want to understand it theoretically

[nashvillebill]

I'm not quite sure you understood what Mr. Fahey is saying...

In reality, there is not a "second channel", there's not an entire second half of an amp sitting there. There is simply an extra set of input jacks. The first tube (valve) is a dual triode--in essence, two tubes combined into one envelope. The first set of input jacks uses one half of the dual triode. The "second channel" just connects to the second half of the dual triode and that second half's signal combines with the signal from the other half afterwards. So removing this so-called "second channel" merely means that you aren't connecting anything to the second half of that tube, so you aren't installing two input jacks, a potentiometer, and a few resistors and caps. The amp still would have the same number of tubes--only now, one tube will be only using one triode instead of both triodes-- and it would still take the same amount of chassis space--eliminating the two jacks and the pot isn't going to free up an appreciable amount of real estate. You'll save all of about $5 (the cost of two jacks, one pot, and a few 12 cent resistors).

[g1]

As to the theoretical part, the only real difference would be that the V1 cathode resistor would need to be changed. It is a shared cathode resistor so with only one triode running you would have to adjust it to properly bias the single triode. For the two 470K's, I think you would want to leave the one in the channel you use as it probably helps prevent blocking distortion.

[nashvillebill]

.... I've taken the schematic and eliminated the "second channel". I kept the Bright "channel, the one with the cap across the pot.


edit: I didn't change the cathode resistor value on that schematic, somewhere around 1500 would be typical I believe.

[g1]

As to the cathode resistor, I was thinking of if the plate was disconnected.
For nashvillebill's example, if the triode is sitting there idling unused, I guess the cathode resistor would be fine as is?

[shocki]

In reality, there is not a "second channel", there's not an entire second half of an amp sitting there. There is simply an extra set of input jacks. The first tube (valve) is a dual triode--in essence, two tubes combined into one envelope. The first set of input jacks uses one half of the dual triode. The "second channel" just connects to the second half of the dual triode and that second half's signal combines with the signal from the other half afterwards. So removing this so-called "second channel" merely means that you aren't connecting anything to the second half of that tube, so you aren't installing two input jacks, a potentiometer, and a few resistors and caps. The amp still would have the same number of tubes--only now, one tube will be only using one triode instead of both triodes-- and it would still take the same amount of chassis space--eliminating the two jacks and the pot isn't going to free up an appreciable amount of real estate. You'll save all of about $5 (the cost of two jacks, one pot, and a few 12 cent resistors).

At least I thought I understood it I know I still need the same amount of tubes and it will not safe too much space. You guys are right about that.

.... I've taken the schematic and eliminated the "second channel". I kept the Bright "channel, the one with the cap across the pot.

What I still don't understand when I look at the schematic. Couldn't I also get rid of the 100K resistors on the "left" part of the dual triode and not connect the cathode of the "left" tube. So basicly using just the "right" triode and using a 1.5k resistor on the cathode?

In a "2 channel" amp. If both channels would be voiced equal would there be difference between 2 channels split cathode (2 x 1.5k and 320uF) and shared cathode (820 and 320uF)?
I think the reason they actually did this back than is to safe some cost? (one resistor and one cap)

[soundguruman]

Hey guys,

I would like to build a 50W Marshall Bass but with just one channel.
Basicly I will use the following schematic.

http://www.drtube.com/schematics/marshall/1986u.gif

1) I think I can get rid of the two 440K mixing resistors
2) I am not really shure about the shared cathode first stage. Should I just use a part of the first preamp tube (so actually split) or should I keep the shared cathode thing? I think I have to go for 1.5k to archive the same bias with split cathode.

I think soundwise it will be actually the same thing?

Thanks alot

you are going to use 1X 470K w/ 500 p in parallel instead of the 2X mixing resistors.
You want that in series w/ the V2 input grid.
This will keep the input grid of V2 at a more ideal voltage and prevent some undesirable distortion.
Without the series resistor, turning the volume up and down, V2 will not be sounding as clear...

You want to use 1 side of the first V1 stages, and leave the other half un-used.
This is purely for noise consideration.
There is a noisy side of that 12AX7, and a quiet side...you will get much less filament hum from 1 side only.
In a bass amp, this design can have quite a lot of filament hum amplified with the audio.

Some of this filament hum is coming from V1, but most of it comes at V2.
You have to carefully arrange the filament connections at V2, to cancel out this hum problem...
The design is inherently noisy, and all of these type (plexi based design) preamps have a pretty loud filament buzz, mostly from V2.
But as I said, it's possible to cancel most of it out from the layout of the wiring.

As far as cathode of 1/2 V1, you are probably better off with 1500 bypassed with 22 uF instead of 820/ 320 uF.
320 uF is going too subsonic in frequency response, and would sound too muddy/ too much gain for a bass amp.

For bass, I think .047 instead of .022 for inter-stage coupling capacitors.

[J M Fahey]

That's probably what shocki had in mind.
Let's see.
Now we have an unterminated triode with an open grid, happy picking radar signals from Mars or whales talking, it's messing with the other triode's bias, we have unexpected +6 dB gain because we eliminated a resistive mixer and the sound of the remaining channel is like none of the original ones.

Here's one proper way to do it, at least keeping the bright channel:


EDIT: well, while I was busy editing the schematic other answers clicked in.
The mod I suggested is the one most faithful to the original circuit and sound , not getting into discussions whether it's good or bad.
Pity is, both channels do not sound the same, one is bright and punchy, the other is mellow, dark.
Now you lost the option to plug into one or the other.
Even worse, you lost the unique Marshall option of plugging into the bright one, jumpering the unused input into the dark one, and mix both sounds at will to get other flavours.
Personally I would save a little panel space by using just *one* jack, wired to both triode inputs, and keep the dual volume controls.

[nashvillebill]

Of course! I did not even think about having that grid sitting there unterminated.

[shocki]

Ok. Now I am totally confused What I mean is this. Of course with a 1.5k resistor.



I have build an "original" 50W Bass already. And I found I just was using one "channel" (the bass one). In a Marshall lead amp you are mixing the channels and it really makes sense. In the bass amp I did not find it necessary. BTW I just will keep the bass channel not the bright channel.

And I am not going to use this as a bass amp but guitar amp. It sounds extremly good and this is what Mike Landau played for some time.
To sum it up. I want the amp to sound excatly like it did in my first build. But just with the bass "channel". Maybe the question was too easy (or stupid). If yes sorry for that and thanks for your patience

The thing with leaving the 470k in makes sense. This is clear now. Also interesting that there is a noisy side of the V1.

BTW I got almost no hum in my first build. The little hum I got was from using a bit lower filter caps.

[g1]

No dual triode is perfectly matched, so one side will always be a bit noisier. So what does SGM do, rewire his sockets every time he changes preamp tubes? Just one of the latest crazes he is on, like the new one where everyone else here is just a tech, but he is a musician too, like no one else is. (and Jim and Leo were NOT )

So the 470K addresses the level difference as JM pointed out.
Your latest schematic is what I thought you meant originally when I mentioned changing the cathode resistor. So you will need to adjust it so you can get the plate voltage where it was in the original circuit.
But one other concern if you want to be just like it was originally, all your supply voltages will change a bit because you don't have that little extra drain through V1. So to get it perfect, you would need a little load in parallel with V1 so the supply voltages would line up with what was there originally. Usually triodes in this scenario are running around 1mA, so a resistor across that last 50uf cap that would give around 1ma load there would probably work.

[pdf64]

Perhaps SGM meant that the lack of full cathode bypassing on the 'high treble' input triode may result in increased susceptibility to picking up heater hum (on that triode)?

Regarding the 2 triode sections of V1, why not parallel them up?
eg 47k plate resistor, 430 ohm cathode resistor, plate linked to plate, grid to grid, cathode to cathode.
As per various Matchless models http://schems.com/manu/matchless/matchless.htm
This may reduce noise a little.
Pete

[shocki]

So to get it perfect, you would need a little load in parallel with V1 so the supply voltages would line up with what was there originally. Usually triodes in this scenario are running around 1mA, so a resistor across that last 50uf cap that would give around 1ma load there would probably work.

So I put an resistor parallel to the last 50uF to ground? Let's say I am measuring 300V at this point I will use R = 300V/0,001A = 300k ? So I am simulating what the other triode will draw?