Muddy sound... preamp caps?

[bluefoxicy]

my amp has a very muddy sound now... especially with low notes. I'm using this reference schematic:

http://www.sewatt.com/files/sewatt/Marshall.pdf

Following values have changed:

R4 - 270k 2%
R5 - 470k 2%
R6 - Tone dial
R7 - 270k 2%
R8 - 820 ohms 2%
R9 - 680 ohms 2%
R10 - 1k 25W 5%
R14 - Pot, 110 ohms
EL84 - EL34
C1 - .058uF
C2 - .047uF
OT - 16 ohm 25W non-reactive across 16 ohm output, to lower primary through a 16 ohm speaker to 2.5k

I'm thinking that I made C1 and C2 too big. I'm not sure what's happening, but I have two guesses:


Preamp gain stage 1 is emitting too much low end, muddying up gain stage 2. Correct with 0.022uF C1, 0.047uF C2.
Preamp gain stage 2 just needs to be filtered. Correct with 0.022uF on C2, 0.047uF on C1


If I can get the preamp to drive right, it has good, crisp overdrive; I really think I went nuts with C1 and C2 and need to scale those back.

[tubeswell]

Well you did alot of mods at the same time, and its hard to really tell which one is more responsible that the rest for the result. But if you want to persist, try changing C1 first. I'm guessing you might even have to go to 1nF depending on what sound you're after.

[bluefoxicy]

Well you did alot of mods at the same time, and its hard to really tell which one is more responsible that the rest for the result.

Yeah, I'm a fan of looking at the result and trying to determine "most likely cause" rather than iterating everything as if it's all equally likely. Most people hate this.

I'll try dropping C1 back to 0.022. I also need to check C4, because it's actually unknown to me... it's part of another circuit I dropped in from earlier, that I forgot about.

[black_labb]

the 820ohms for r8 would give it a bit of a fuzzier sound, which can be good, but i think having the large coupling cap after it gives it a bit too much undefined low end. either drop the following coupling cap down a buit or bring that cap back to 1200 or so. why not put the 1500 in there and then put a toggle switch to add a second resistor in paralell so you can switch between the different sounds? i find in simple amps a simple switch like this can make an amp really versatile.

[bluefoxicy]

Uh... 1500 cap? You mean the first gain stage cathode resistor? :P R8 is low for that warm-and-fuzzy sound, but I have a lot of switches in there as it is. I'll build a new amp some day with all my switches planned out for edge mounting.

also why a switch and not a ~1k pot? I have insanely small (1.5-2cm diameter) knobs, 820R into 0-1k works.

I'll drop that cap down to 0.022, and let the next cap keep its big value for now. If that doesn't help I'll clean up the resistor a bit.

[black_labb]

sorry, meant to say resistor. a pot will work as well

[Bruce / Mission Amps]

I'm going to go out on a limb here but I bet you didn't build this amp just like the schematic. For one thing, your filament voltage is going to be astronomically high with a FWB rectifier and 6.3vac across it.

I'd change both R8 and R9 to 1500 to maybe even as high as 2700 ohms and I wouldn't use a 220K plate load resistor (R4) especially with such small cathode resistors.
There is a VERY good reason most 12AX7s use 100K plate load resistors... especially in amps with a little EL84 (which requires much less signal on it's grid for full power output) and no other circuitry between the first and last gain stage,
R8 could have a 4.7uF to 10uF across it but I think R9 at 1500 to 2700 ohms should have 1uF to 2.2uF cap for starts.
Also reduce your plate coupling caps back to 22nF
Why is R14 a 110 ohm pot!?
And... I don't understand your 16 ohm across the output to make 2500 ohm...
You need at 5000 ohms across the SE El84 to get some hi-fi tone out of it.
That might be part of your muddy response, not to mention that the amp in that configuration isn't going to make more then 2 - 3 watts anyhow and half of that will go into a 16 ohm resistor.

[bluefoxicy]

I'm going to go out on a limb here but I bet you didn't build this amp just like the schematic. For one thing, your filament voltage is going to be astronomically high with a FWB rectifier and 6.3vac across it.


Nope. That's just reference. I twisted the filament leads around and used AC, 0.76A on one tube and 1.5A on another.


I'd change both R8 and R9 to 1500 to maybe even as high as 2700 ohms and I wouldn't use a 220K plate load resistor (R4) especially with such small cathode resistors.
There is a VERY good reason most 12AX7s use 100K plate load resistors... especially in amps with a little EL84 (which requires much less signal on it's grid for full power output) and no other circuitry between the first and last gain stage,


2.2k is too cold, 1.5k was too cold, I dropped those resistors to help cause overdrive. The 220k R4 was for pushing that tube harder too. It still doesn't overdrive as fast as I'd like...

There is no way to get an EL34 to overdrive at those gain levels. Which is fine, I can get the 12AX7 to overdrive and switch out R5 later.


Why is R14 a 110 ohm pot!?


To set the cathode dissipation to somewhere around 25W.


And... I don't understand your 16 ohm across the output to make 2500 ohm...

Zout = 3000

[bluefoxicy]

Awesome, it sounds good but I want it to overdrive a little faster. C1 = 0.022uF, less R6 now (0-200k instead of 0-1M), more R7 would help (it's 270k instead of 97k, might raise to 470k).

Power tube STILL won't overdrive, I might raise that grid leak (R5) to 1M. later.