Reverb Issue in Homebrew Amp

[ChrisM]

Can't figure this one out on this amp I have built.
Here is a portion of the schem where the troubles are (click to enlarge).

[IMG:160:71]http://img.photobucket.com/albums/v610/kurtlives/th_CM-50.jpg[/img]

The issue is when the reverb is switched on the overall volume drops considerably. The weird thing is messing with the control labeled Volume affects how loud or quiet the reverb is. And its not a linear thing like as you turn the Volume there is loud spots and quiets spots throughout the rotation. Does that make sense?

I have checked things over a bunch and can't find any glaring mistakes. Any insight or help is appreciated.

Thanks...

[ChrisM]

Unplugged the tank and the problem is still there.

Now I'm getting confused this isnt making a ton of sense to me. Is there a fundamental design error?

Also was checking voltage. The triode with the plate resistor sits at 205V, ok. The transformer triode is at 300V....hmmm

EDIT: Moved the reverb return wire from the Treble Wiper to the Volume Wiper. I no longer have the Volume control making the reverb volume jump all over the place. I still have a weakened reverb singal though and the Volume control does affect it. If you max out the Volume control the reverb comes through strong and sounds fine.

[Super_Reverb]

Looking at your schem, there is one question:

How can the DC plate voltage on your first gain stage V1A and the drain of the N-channel MOSFET source follower both be 190V? Are you intending to say B+5=190V?

If were was debugging this, I'd drive the amp with a sine wave around 300mV P-P (~100mV RMS). I have used a guitar tuner out with a volume pot in the past. You need to verify the circuit is doing what you intend.

If you don't have a scope, use your multimeter in AC volts mode and find something to drive the input to determine if the AC signal is being gained up through the triode stages, attenuated through tone stack, etc. Your first look at a finished circuit, you are verifying circuit design *and* electrical connections and layout. Suggest checking all your preamp plate and cathode voltages and report back. You can determine cathode current using Ik=Vk/Rk. Then as a sanity check, your plate bias voltage should be Va(DC)=B+ - Ik*Ra. This neglects the small grid current. This will help you understand if your gain stages are working.

What is your intent with the 220K/500p in the grid ckt of 2nd gain stage? If you are doing a grid stopper, the 500p is not needed. You should have a pretty sparkly top end already with the volume pot bypass caps. I'd suggest a few k Ohms series without parallel cap.

The reason is good design practice. Tubes have some grid to cathode capacitance of a few pF: when you have a few k Ohms series resistor driving the grid, you are creating a low pass filter which should allow all guitar audio frequencies to pass (unless R is too big) and which will help prevent parasitic oscillation.

The DC plate voltage on your reverb driver, V4A will be at a fairly high voltage, because the bias current won't create a voltage drop like you get with a 100K plate resistor.

cheers,

rob

[Ian444]

Chris, without being able to see the whole schem I'll have to ask a dumb question, is the reverb output signal in phase with the original guitar signal? Another possibly dumb question, maybe irrelevant to your problem, why is the reverb circuit bypassed with a 100K resistor?

[10thTx]

There is some information and schematics about the one tube reverb right here:

http://www.el34world.com/Forum/index.php?topic=7957.0

Here is a schematic of how I use that reverb with an LTPI amp.

With respect, 10thtx

[martin manning]

Chris, "clean out" and "to phase inverter" are opposite in phase. That might be a problem if the "clean out" signal doesn't get flipped before it reaches the PI, as it is feeding forward through the 100k.

[jjman]

Is there a need for the 100k path from the reverb's input to it's output? That's a volume reducer via phase cancellation as mentioned, although the full schematic is needed to verify. Without it all the reverb's output will be "wet" and therefore not in phase with anything else.

[ChrisM]

I got the single tube reverb from here http://www.glaswerks.com/schematics/reverbcir.jpg
I have used this reverb before with success in a Rocket type amp.

I got the 100K resistor from there, its the mix resistor.

Do you want a full schem? Other than that there is just an ODS style OD section a LTPI and output tubes o and PSU.

Any ideas on how to flip the phase

[martin manning]

Any ideas on how to flip the phase?

You could just move the point you are using for the reverb input over to the "clean out".

[ChrisM]

Any ideas on how to flip the phase?

You could just move the point you are using for the reverb input over to the "clean out".

Did this and the phase issue is resolved thanks.

However there is a huge boost in gain. Like huge! ???

[snyder80]

Hi Chris!

Actually i think you got the mixing resistro (100k) wrong.

Have a look at the working 1 tube reverb schematics floating around.

Its like this:


Stage1 ----Mixing Resistor----Stage 2----Next Stage or PI
_______I______________I
_____Reverb IN_____Reverb Out (Wet)


You had it this way:


Stage1 ----Mixing Resistor----Stage 2-----Next Stage or PI
_______I________________________I
_____Reverb IN______________Reverb Out (Wet)


So basicly you bypass "Stage2". This "doubles" the phase of the Reverb and Dry phase hence giving you a boosted signal.

Simply cure: flip the primary Reverb OT´s connections and connect the reverb circuit In/Out incl. the mixing resistor between the same gain stages.

have fun,
snyder80

[ChrisM]

Hmm I dont quite understand you. Comparing my schem and the one tube reverb schem the mixing resistor looks the same to me.


What if I reverted back to putting the reverb in at the tonestack and flipped the reverb xformer's wires. Would that fix my phase problem, I would think so...

[snyder80]

me,too

Or use your circuit drawn above (cm50) and leave out the 100k mixing resistor!
its not that complicated.

I´ll try to explain:

Basicly Reverb circuit adds the "dry" and the "wet" signal in the same phase.

The mixing resistor and the "reverb" pot defines the mixture between the dry and reverberated/slightly delayed signal.
There are two typical types of adding the reverb signals.

Voxy by mixing the dry signal and the reverb at the 2 PI inputs OR Fender type by adding a "mixing resistor" between the same 2 Preamp-stages.

Why ?

1. Each triode-system (aka. gainstage) flips your signal in 180° if its coming from the plates. So you have the mixing resistor between the reverb driver stage and the reverb recovery-stage.
So your signal gets flipped 1x180° @ the reverb driver and 1x180° at the recovery stage.This flips your signal going to reverb 360° (in phase) with the dry signal. As adding similar waveforms the signal sounds louder and hence the "boost" appears. To avoid this you have to do some tweaking to the mixing resistor and the recovery-stage.
Usually, the further at the end of your preamp your reverb is, the more "boost" you can expect. Why because at the end its getting gainier and louder to the PI.

2. Feeding the wet signal to the PI flips the signal out of phase (i think) to the negative LTPI-Input. That way its getting mixed and flipped 1x180° in the LTPI so its in phase with your dry signal. But i think you would need a gainier driver stage.

3. If you mix the voxy-type and fender-type like you did (mixing resiostor AND feeding the Wet to the PI) your reverb signal gets 3x180° wich makes it out of phase and thus shitty sounding.

4. Too complicated? Just compare the schematics of a fender deluxe reverb reverb circuit (or the 1 tube reverb circuit drawn above in the tweedy amp) with the voxy type.

[ChrisM]

snyder!!

Thanks so much. I see what you mean now.

I took out the 100K resistor and kept eveything else stock. Seems to be just fine now.

Thanks...