i'm trying to make a 2 channel amp, will this work?

[pzung]

so this is just the start of a thought...

i want to switch between a high gain channel to a lesser gain or maybe even clean channel. i'm thinking of just using 2 of doug's relays powered off the heater wires.

i'm also trying out a cascode on the front for hopefully a better starting signal to noise ratio. my current amp, based on the plexi circuit with some soldano-esc bits, is the noisiest, so i'm also trying to figure out a different layout, kind of like a hiwatt with the high power lines down the middle and the small signal components between the preamp tubes and the control knobs.

i want to finish the power supply and get the switches working before i start filling in everything else. so my first question is really, does this general arrangement look like it would work as far as the switching is concerned? i have double pole,double throw switches, should i ground the hanging ends of the circuits when they're switched out?

thanks!

patrick

[PRR]

Your cathode-follower is drawn wrong. (You probably want to get output from the cathode.)

You probably want some heavy attenuation after your boost-stages. You want to turn-up going into the boost to force distortion and to work at good S/N. The result is too hot to go into further amplification, knock it down.

> cascode on the front for hopefully a better starting signal to noise ratio

How does a cascode help S/N in an audio amp?

It "helps" in the VHF radio band because plate-grid capacitance and tuned circuits destabilize a triode stage. A pentode makes P-G capacitance negligible but plate-screen partition hiss-noise degrades S/N. The cascode gives triode hiss with pentode stability.

Using tubes, do a standard Fender first stage. Bypass the cathode. Run a high supply voltage and high current. You may consider one of the higher-Mu "cascode" triodes to work at higher current. However the irreducable ~~5K pickup impedance (rising to ~~100K in a critical hiss-band) means heroic hiss-reduction design won't help much.

Hum and buzz of course need proper filtering and signal/ground routing. Magic circuits can be tried, but simple good-cleanliness practice does the job.

Chassis layout is similar to home plumbing sanitation. Keep your septic pit, sewage lines, and your pig-pen away from your precious water well. Ideally you keep all your pure-water lines short and tight to reduce the risk of contamination. With plumbing-pipe this is not so important, but we can't "seal" audio so neatly. Imagine you had no pipe and ran water through open troughs.... preamp and tone-pots will "catch the spray" from open dirty-water and sewage troughs (or high power sections). Separation is the best isolation.

[pzung]

thanks PRR,

re: cathode follower. oops, threw that drawing together while someone was giving me the "it's time to get off the computer eyes".

re: cascode. i thought i'd try it after thumbing through "designing valve preamps" by M Blencowe. i got a jan6922 tube for the purpose. somehow got the impression that it could be a quieter circuit if i filtered the crap out of it's plate voltage and possibly did dc heaters for it. if i could get it to be quiet than it's initial high gain would have less noise in it to be multiplied down the line. plus, i could mess around with something new that has the potential to sound either pentode like or triode like depending on where you land in the plate curves.

[pzung]

here's a photobucket link to the chassis i'm going to use. it's an old marshall chassis that got butchered and then given to me. i've moved the transformers and preamp tubes and i'm putting in more turret boards.



BORING ALERT: the following is a description of how i think the layout is going to work...

-the PT is all the way on the left with a rectifier tube just to the right. i had to put on a 5v filament transformer for the rectifier.
-the inputs are on the far top right.
-below the rectifier socket are the first main filter caps.
-the red board in the middle will be the bias circuit and the next set of mains filter caps. right above that board are the red and black wires coming from the choke.
-the long red board down the middle will be all the filter caps for the preamp stages with a heater reference voltage cap right after the driver supply. i'm using little axial caps with the positive leg towards the preamp tubes and plate resisters from the caps to the pins.
-the board at the top is from the first version of the amp i play now and still has some junk on it that didn't get transfered over. i'm going to run short wires from the preamp tube pins to that board to coupling caps, cathode bypass stuff, anything signal related. i was thinking about hanging the tone caps and slope resister right on the tone control pots, bad idea?
- i'm using KT88's so they're spaced out. the board between them is for the .1 caps that get the signal from the driver, and the grid stoppers, bias trim pots and 1ohm cathode resisters to check the bias draw.
-the OT is all the way on the right. the laminations for both the PT and OT are pointed up and down. i think that is the right way to keep they're magnetic fields pointed away from any of the tubes. so like a side with a bell is pointed towards where the tubes are. i placed both the OT and the 5v filament transformer using "the headphone trick" so they're not hearing the PT.
-the heater wires run are down the middle of the amp under the long red board laying flat against the chassis.
- i have the main ground to the chassis over by the plug where the house voltage comes in. there is a short buss under the main filters where i was going to bring all the ground wires to before sending them to the chassis ground. i sort of imagine a lot of ground wires right now. i'm not sure where to run them. i could also make the bottom edge of the long red strip into a ground bus.
-the signal wires from the driver tube have to pass by or over the mains supply. i thought maybe i'd shield them? it's not far to go just straight down to the yellow board.

is there an audible way to tell if a wire is throwing noise? like waving around the end of a chord plugged into an amplifier and wrapped with electric tape.

[PRR]

> the impression that it could be a quieter circuit if i filtered the crap out of it's plate voltage and possibly did dc heaters for it.

Do NOT say "noise"or "quiet".

Say hiss, hum, buzz.

In (non-audio) applications the Cascode may be lower HISS than alternatives. It can still hum as bad as any, and will buzz like a bee if your B+ has buzz or you have buzzy leads near its HIGh-impedance node.

For low hiss, the standard Fender first stage is as good as guitar needs. (Weed-out hissy tubes.)

For low heater-supply hum, twisted pair run well away from all audio points has been shown to be really darn good; however if you add too much gain for too-much distortion (or used PCB), you may need DC heat (and good DC, not single-cap filter).

Low B+ hum/buzz is trivial in this day of cheap electrolytics.

Yes, a cascode can be overloaded easier and with different flavor than a simple triode.

[pzung]

PRR,
you give a lot, i truly appreciate it.

 hum,hiss,buzz...

-so hiss is circuit/tube/component related
-hum caused by heaters
-and buzz from the dc supply?

how far away is far enough from heater wires? can my plate load resisters reach over them if the heater wires are against the chassis and the plate resisters are coming down off of a board that's about 5/8" up off the chassis?

do you think the cascode thing is a waste of time if truly my goal is more gain and less HHB? i guess i'm hoping to stumble upon something that sounds interesting.

well, actually i have 2 goals now.
1. i want to get the sound of my amp as gainy as it gets plus a tube screamer without the tube screamer  and without having to switch to standby to have a conversation with my drummer.
2. i want to be ably to foot switch over to a sound that is more marshall 70's rock inspired and i imagine being able to clean up this channel by diming the master volume and using the preamp volume for level.