Stereo guitar amp 2x6L6

[Moonage]

Thinking about building a stereo guitar amp. I have some ideas, and did some drawings. I am posting a schematic for output section that I designed. Any critics, ideas and suggestions are welcome. (Do not hit too hard though  )

[Jack_Hester]

Take a look at the Gibson GA-78 schematic.  I have one of those and it works very well as a true stereo, on mono amp. 

Jack

[Paul1453]

Another thing you might consider is to buy a stereo tube amp on Ebay and modify that.

I'm quite cheap, and can usually get something like that for much cheaper than I could buy just the PT and OTs for.   

[Moonage]

Jack, thanks for the tip, I did not see that schematic before.  Now, my approach is different, I am not looking to get amp with the stereo input, it is more about pushing Fx loop all the way to the end of the amp...at the speaker. That way Fx signal is not distorted by the output tube, and is separated from the main signal. That is the idea of my design. One input, two outputs, dry and wet.

[jjasilli]

Jack, thanks for the tip, I did not see that schematic before.  Now, my approach is different, I am not looking to get amp with the stereo input, it is more about pushing Fx loop all the way to the end of the amp...at the speaker. That way Fx signal is not distorted by the output tube, and is separated from the main signal. That is the idea of my design. One input, two outputs, dry and wet.

You are not getting many responses.  Why? 


It is important to develop a clear concept and design goals for the project.  It is not clear to me what you wish to accomplish.


Do you want the "wet" FX signal to be undistorted by its output tube; but the dry signal to get output tube distortion?  Do you want single ended (Jack's suggestion is PP)? 

[Moonage]

Usually signal chain looks like this:
                                               
Amplification - Distortion - Fx - Drive/PI - Output
                         

So, wet and dry signal get mixed, then amplified, and distorted in Drive/PI and output stages.
Let’s say you have echo effect in the Fx loop.  If last two stages are overdriven, you will hear distorted echo. 
What am I trying to do looks like this:
                                                           
Amplification - Distortion - Drive/PI - Output 1 - Fx - Drive 2/PI 2 - Output 2

This way, dry and wet signal are separated, and effect does not get distorted ( it could be distorted if wanted by adjusting drive pot.)

[jjasilli]

Ok.  A few  things.  Distortion is a form of FX.  So to separate distortion from FX may cause confusion.  Yes some effects like reverb, delay, chorus, should go last in the FX chain, with no distortion after that.  See, e.g.:  http://www.jamplay.com/articles/5-guides/182-signal-chain-a-look-at-effects-routing-part-1


FX goes, both out & in, before the PI or the driver of a power tube.  Stereo is not necessary for this to happen, BUT. . .

it is possible, maybe desirable, to overdrive the PI (or the driver of a power tube).  It is possible, maybe desirable, to overdrive the power tube(s).  Is this why you want stereo?

Do you want PP or SE?

Do you want FX in a preamp loop?  Or pedals before the preamp? 

You need to describe the whole plan from guitar chord to speakers.

[HotBluePlates]

Your 6L6 cathode resistors seem a bit small.

I noticed you're using essentially data sheet conditions for 350v plate, 250v screen. The 6L6 sheet claims 54mA plate and 2.5mA screen current, with a bias of -18v. That implies a cathode resistor of ~318Ω, or the closest standard value.

Do you know what level your effects attached to the line out will need? Power output may be reduced a bit from the data sheet's 10.8w, but let's roll with that. Secondary voltage will be √(10.8w * 16Ω) = ~13.15v RMS, reduced to 4.1v RMS by the 22kΩ/10kΩ divider. If the effects typically see a guitar-level input, perhaps swap a 1kΩ resistor for R152. Or since that may still be too hot (about 570mV RMS), perhaps a 1kΩ resistor in series with a 1kΩ pot for a Line Out level.

R131 and R137 may be unnecessary. 100Ω is probably too small to be an effective grid-stopper, but a grid-stopper isn't really needed here as the connection formulate to grid on these stages will be made either by bending the socket tags over to solder together, or through a very short hunk of hookup wire.

I didn't draw any loadlines, but everything looks reasonable at a glance. V105 1 shows a different current going into the plate than what comes out the cathode, but I assume that's just a typo.

[HotBluePlates]

Ok.  A few  things.  Distortion is a form of FX.  So to separate distortion from FX may cause confusion.  Yes some effects like reverb, delay, chorus, should go last in the FX chain, with no distortion after that.
...
Do you want PP or SE?...

"Stereo" may have been an unfortunate word choice in the original post.

What I think I see is Moonage wants a SE 6L6 amp which will generate the primary sound (with distortion from the 6L6, if desired). The output of this amp is applied to a speaker, but also to a Line Out which feeds effects to process the sound from this amp's speaker. Those effects output a signal which is then amplified by a 2nd SE amp, and output to a "wet" speaker.

This is so not different in concept from the Fryette Power Station (though that unit has its own clean power amp to boost the effected sound and power the "wet" speakers).

[jjasilli]

Yes, that may be. I'm very literal; you're better at getting at the meaning behind the post.  I'm trying to draw-out Moonage so his issues can be addressed.

[shooter]

FWIW I did build a *super Champ* stereo amp 2 ax7, 1 kt88 per side, for some of the reasons mentioned, the guitar guys *request* 2channels that 2 players could use into a common cab, or 1 player with a clean channel, one as an FX channel.  The *head* weighed in at 42+ lbs, add 20-30 for an oversized 2X12 and, well, practicality and reality met.  Upside, lots of parts for then future builds

[Moonage]

Do you know what level your effects attached to the line out will need? Power output may be reduced a bit from the data sheet's 10.8w, but let's roll with that. Secondary voltage will be √(10.8w * 16Ω) = ~13.15v RMS, reduced to 4.1v RMS by the 22kΩ/10kΩ divider. If the effects typically see a guitar-level input, perhaps swap a 1kΩ resistor for R152. Or since that may still be too hot (about 570mV RMS), perhaps a 1kΩ resistor in series with a 1kΩ pot for a Line Out level.

HotBluePlates, thank you for going in dept with my design. This is first time I am doing design with line out, and your input helps a lot. I also like the idea of line out level pot.

[HotBluePlates]

You're welcome.

I'm thinking the 150kΩ loads for the 12AY7's may be higher than is really necessary. Then again, I hadn't drawn loadlines or tried to figure out the impact of the direct-coupling.

[Moonage]

What I think I see is Moonage wants a SE 6L6 amp which will generate the primary sound (with distortion from the 6L6, if desired). The output of this amp is applied to a speaker, but also to a Line Out which feeds effects to process the sound from this amp's speaker. Those effects output a signal which is then amplified by a 2nd SE amp, and output to a "wet" speaker.

Thanks HotBluePlates, exactly what I am trying to do.

Question:  If there is no Fx attached to the line out, signal from line out is fed to the 2nd output amp, and essentially, just duplicated. Can I expect to get fairly undistorted  and same level signal from the second SE amp?

[Moonage]

You're welcome.

I'm thinking the 150kΩ loads for the 12AY7's may be higher than is really necessary. Then again, I hadn't drawn lordliness or tried to figure out the impact of the direct-coupling.

Here is the graph with loadline:

[sluckey]

I didn't draw any lordliness

Then again, I hadn't drawn lordliness

That's twice, my lord. Better sharpen your pencil! 

[DummyLoad]

Jack, thanks for the tip, I did not see that schematic before.  Now, my approach is different, I am not looking to get amp with the stereo input, it is more about pushing Fx loop all the way to the end of the amp...at the speaker. That way Fx signal is not distorted by the output tube, and is separated from the main signal. That is the idea of my design. One input, two outputs, dry and wet.

why SE? just curious.


--pete

[Moonage]

why SE? just curious.


--pete

It was easier to make schematic...less components.
Seriously, no reason, beside the number of components, and the preamp if built I initially designed is already too complex, and I might end up with two chassis.
At this moment, this is more a concept. 

[HotBluePlates]

I didn't draw any lordliness

Then again, I hadn't drawn lordliness

That's twice, my lord. Better sharpen your pencil! 

Ha! "Your Majesty!" 

Yep, got a new computer whose browser does auto spell-correct (instead of just telling you the word is spelled wrong), and it doesn't recognize any electronics terms. Now you also know I can't touch-type (so didn't see it changing the words as I wrote).

Question:  If there is no Fx attached to the line out, signal from line out is fed to the 2nd output amp, and essentially, just duplicated. Can I expect to get fairly undistorted  and same level signal from the second SE amp?

Was my earlier guess that you'll be using guitar-level effects at the Line Out correct? If so, and the Line Out can knock the signal down to guitar level (or similar), then the 2nd amp circuit should amplify a dry signal cleanly.

I'll analyze the circuit in more detail later tonight or tomorrow, to get an idea of proper signal levels throughout.

[HotBluePlates]

Did the analysis; unless your effects output a strong signal, you may need a gain stage between the Line In and Volume pot of the Wet Amp.

At the operating point you chose, the 12AY7 has an internal plate resistance of about 37kΩ, and with a 150kΩ plate load will have a gain of about 32 if the cathode resistor is fully bypassed. Mu at the operating point is 40. Gain with the unbypassed 3kΩ resistor will be about 19.4.

The a.c. load of the cathode follower is (I think) immaterial in this circuit. The a.c. load of the common cathode stage feeding the follower is appreciably same-as 150kΩ. So no drawing of a.c. loadlines needed.

As I mentioned before, since you're basically copying the data sheet condition for 350v plate and 250v screen (albeit with a different load), I'll assume the correct bias voltage will be -18v. I've annotated 18v on the cathode and a bias resistor of 320Ω. Those are available in a 3w wire wound from Vishay Dale (and perhaps others).

The output tube will probably make all its clean output and be breaking into serious distortion by the time the peak grid signal equals the bias voltage. So following this rule of thumb, our required peak signal for full output power will be 18v peak at the 6L6 grid.

There is a loss of about 0.8 at the voltage divider formed by R139 & R142, so signal at the cathode follower output must be about 22.6v peak for max clean output.

Gain of a cathode follower is roughly µ/(µ+1), so gain of the 12AY7 cathode follower will be about 0.98. The must be a signal input to the cathode follower of about 23v peak.

As noted earlier, the gain with a fully bypassed cathode resistor is 32, but only 19.4 if unbypassed. Assuming you add a bypass cap (and delete C122, which would then be a capacitive voltage divider and drop input signal to nothing), a grid input on the 1st 12AY7 stage of 0.72v peak is needed for max clean output.

There is a volume control ahead of the 12AY7, and assuming you want all the clean headroom of the Wet Amp to run out at about "9" on this volume pot, you will need double the input from the Line In as we just previously calculated. So the Line In should receive 1.44v peak (a bit over 1v RMS) for max clean output from this Wet amp. That's a pretty hot line level, consistent with pro audio gear.

All that said, you can run it just fine with weaker inputs, and everything will work. You won't really max out the clean power of this Wet Amp, but then againYou may be looking to ensure it never distorts. Eliminating R139 will increase the sensitivity of the Line In to about 1.15v peak (0.81v RMS) for maximum clean output.


Adding gain would then give your Volume control a range of adjustment, instead of being full-up all the time. Or you could just eliminate it and replace with a resistor to ground.

[Moonage]

Did the analysis; unless your effects output a strong signal, you may need a gain stage between the Line In and Volume pot of the Wet Amp.

At the operating point you chose, the 12AY7 has an internal plate resistance of about 37kΩ, and with a 150kΩ plate load will have a gain of about 32 if the cathode resistor is fully bypassed. Mu at the operating point is 40. Gain with the unbypassed 3kΩ resistor will be about 19.4.

The a.c. load of the cathode follower is (I think) immaterial in this circuit. The a.c. load of the common cathode stage feeding the follower is appreciably same-as 150kΩ. So no drawing of a.c. loadlines needed.

As I mentioned before, since you're basically copying the data sheet condition for 350v plate and 250v screen (albeit with a different load), I'll assume the correct bias voltage will be -18v. I've annotated 18v on the cathode and a bias resistor of 320Ω. Those are available in a 3w wire wound from Vishay Dale (and perhaps others).

The output tube will probably make all its clean output and be breaking into serious distortion by the time the peak grid signal equals the bias voltage. So following this rule of thumb, our required peak signal for full output power will be 18v peak at the 6L6 grid.

There is a loss of about 0.8 at the voltage divider formed by R139 & R142, so signal at the cathode follower output must be about 22.6v peak for max clean output.

Gain of a cathode follower is roughly µ/(µ+1), so gain of the 12AY7 cathode follower will be about 0.98. The must be a signal input to the cathode follower of about 23v peak.

As noted earlier, the gain with a fully bypassed cathode resistor is 32, but only 19.4 if unbypassed. Assuming you add a bypass cap (and delete C122, which would then be a capacitive voltage divider and drop input signal to nothing), a grid input on the 1st 12AY7 stage of 0.72v peak is needed for max clean output.

There is a volume control ahead of the 12AY7, and assuming you want all the clean headroom of the Wet Amp to run out at about "9" on this volume pot, you will need double the input from the Line In as we just previously calculated. So the Line In should receive 1.44v peak (a bit over 1v RMS) for max clean output from this Wet amp. That's a pretty hot line level, consistent with pro audio gear.

All that said, you can run it just fine with weaker inputs, and everything will work. You won't really max out the clean power of this Wet Amp, but then againYou may be looking to ensure it never distorts. Eliminating R139 will increase the sensitivity of the Line In to about 1.15v peak (0.81v RMS) for maximum clean output.


Adding gain would then give your Volume control a range of adjustment, instead of being full-up all the time. Or you could just eliminate it and replace with a resistor to ground.

Thanks a lot...

So, by replacing the 12AY7 with 12AX7(and adjusting quiescent point to -2V), will give me , for the same 150 ohms plate resistor, a voltage gain of about 63. That way, I can have an option to have clean amplification (in case I run signal through the fx and do not want additional distortion after) or use it as an overdrive (for example, running output 1 fairly clean, and then pushing the output 2 in overdrive).

[HotBluePlates]

Yes, you can go that route. The key is to solidly define what capabilities you want the final circuit to have, then evaluate whether the circuit you're working on delivers those capabilities.

For example, if you want only clean when using the 2nd amp for effects, but the ability to overdrive it when cascaded directly from the 1st amp, you could simply use switching to change the voltage divider and deliver a stronger signal to the input of the 2nd amp.