Best way to combine two dissimilar channels into PI?

[dehughes]

Hey all,

So, below is an (older, but close enough) schematic for the amp I'm (still) tinkering with. I find that a 500pf bright cap on the 6SN7 channel (switchable, and installed across the 220k mixer resistor...have yet to update the schematic), when engaged, actually TAKES high end from the OTHER channel (12AX7). This even happens when the 6SN7 volume is completely turned down! Weird. Conversely, I find that the tone pot setting on the 12AX7 channel does NOT affect the 6SN7 channel. This seems to tell me (I think) that the way I have the two channels mixed at the PI is causing some bleed over from one channel to the next, so I'm considering putting each channel into separate sides of the PI.

1) Am I right in assuming the mixing of channels as I've done can cause some signal bleed from one to the other?

2) Is there any appreciable benefit/difference in putting each channel into its own PI input?

Thanks!

[IMG791]http://i452.photobucket.com/albums/qq24 ... ematic.jpg[/img]

[paolojm]

I'm at work so can't see the schematic but if I'm picturing it correctly then this is exactly what I'd expect to happen. The 220k mix resistor acts as a load to ground via the other channel (6SN7) volume control. This is forming a low pass filter filter which is attenuating highs. There should be a by pass cap on the mixing resistors (again if I'm imagining your circuit correctly) if you want to avoid this.
You could have the bright control installed between the top lug and wiper of the volume control like a traditional brite switch if you wanted.

[dehughes]

Thanks man. Actually, I just wired both channels into their own PI input yesterday, and totally re-worked the bright switch. I'll give it a run today and see how it goes.

[dehughes]

Okay, lesson learned: Putting two channels into each side of the PI works great if they are OUT OF PHASE with each other, but when they are in phase....it's, well, interesting, to say the least. Lots of different tones available via mixing the two channels and such, what with reversed tone knobs for the opposite channel when both are engaged, but there's a lot of canceling when both channels are running. Fun, but not really practical.

I think I'm going to put them back on to one side of the PI and try to adjust coupling caps in order to get the 6SN7 side a bit fuller. I like running that tube into a .047uf at the PI, but then the 12AX7 channel (Brown Deluxe) really digs the .01uf...but that's too small for the 6SN7, and a .047uf would be too much low end for the Brown Deluxe side.

I'll tinker some more...

[oldhousescott]

Putting a bypass cap across your mix resistor will look like a low-pass filter to your other channel. Go back to your two 220k mix resistors. You don't really need the cap at the input to the PI since you have coupling caps isolating each channel. This will let you pick an appropriate coupling cap for each channel without having to make a compromise. Just tie the junction of the mix resistors to the input to the PI.

The tone circuit doesn't affect the other channel since it's isolated by the make-up gain stage following it.

[dehughes]

Now that's a thought...so I'd just put both channels (the mixer resistors, that is) right up to the grid on the PI? Would that allow me to keep the Brown Deluxe channel unaffected by the low pass created by a 500pf cap across the 220k mixer on the 6SN7 side?

Two things I've found in tinkering about with this amp:

1) I find the 6SN7 channel needs the cap across the mixer resistor, as it clears things up a bit. As well, the PI input cap value for this channel is best set at about .047uf.

2) When both channels are mixed into the same PI input cap, the tone pot on the 12AX7 channel is pretty dark/muted until nearly all the way up. HOWEVER, when each channel was given its own PI input side, the tone pot came alive! It had a much more usable, full, sweep, and the channel overall was more clear and usable than when both channels were mixed into one side of the PI. The downside of course was that there was (obvious) tonal/sweep/phase issues when the two channels were used together (I'm assuming this is because they were in phase before hitting the PI?).


What does this indicate? What should I be gleaning from this? Is there a way to have my cake and eat it too? That is, I want to have both crazy awesome Brown Deluxe as well as that nice, full, chimey 6SN7 thing...

THANKS.

[oldhousescott]

No, you'll have to lose the cap across the mix resistor to prevent the high-end rolloff. The mix resistors do load each channel so you won't get as strong a signal hitting the PI. If you want to brighten things up a bit, you can put a bright cap across the volume pot on the 6SN7 channel. You can also put that 470p cap across the 220k coming from the tone stack/volume pot on the 12AX7 channel leading to the make-up stage grid. If the 470p is too strong, try 220p. If that's still too strong, try putting a resistor in series with the 470p, something in the range of 22k to 220k.

[dehughes]

No, you'll have to lose the cap across the mix resistor to prevent the high-end rolloff. The mix resistors do load each channel so you won't get as strong a signal hitting the PI. If you want to brighten things up a bit, you can put a bright cap across the volume pot on the 6SN7 channel. You can also put that 470p cap across the 220k coming from the tone stack/volume pot on the 12AX7 channel leading to the make-up stage grid. If the 470p is too strong, try 220p. If that's still too strong, try putting a resistor in series with the 470p, something in the range of 22k to 220k.

Hmm...okay. Thanks!

So, given these two channels, what's your recommendation as to how I would best mix them? I find that when the 12AX7 shares the same PI input cap as the 6SN7, the tone pot on the 12AX7 side is so much muddier and unusable than if each has their own PI input...

[oldhousescott]

Try this. I marked up your drawing showing recommended changes. If it still doesn't get you what you want, you could add an anode follower to one channel to swap the phase and then take each channel to a separate PI input.

[dehughes]

Cool! Thanks man! I added a bright cap across the 6SN7 volume and like that much better. As well, I tried installing a .047uf cap for the sake of deepening the 6SN7 side. The bright switch on the volume and the .047 on the PI input helped out the 6SN7 side, but the 12AX7 is even darker than before (surprise). Attached is what I tried....

I'm able to get each channel sounding great on their own, but it's combining them that's proving to be very difficult. I'd try the anode follower thing, but I don't have any more space for tubes on my chassis...

[dehughes]

Attached is the updated version....I tried this schematic as well, but again, the channels cancel each other out. If I could get them in phase (or out of phase...depending on the PI setup), I think this would be close to good...

[Wayne]

Forgive the silly question, I'm late coming to the party.

In your AX7 channel, is there a typo, or are you padding the signal level after the 2nd triode with the 15k/100k resistors? If the 7:1 split load is intentional and you do have that much gain to spare, how about this - reduce the cathode bypass cap to provide partial bypass/treble boost, then make up for the lost gain by either increasing the 15k, reducing the 100k, or both.

W

[dehughes]

Forgive the silly question, I'm late coming to the party.

In your AX7 channel, is there a typo, or are you padding the signal level after the 2nd triode with the 15k/100k resistors? If the 7:1 split load is intentional and you do have that much gain to spare, how about this - reduce the cathode bypass cap to provide partial bypass/treble boost, then make up for the lost gain by either increasing the 15k, reducing the 100k, or both.

W

Hey Wayne! Thanks for checking this out...much appreciated.

I've basically copied the 6G3 Brown Deluxe's "Normal" channel, save for not putting a cap across the 220k plate resistor on the first triode. The 100k/15k plate resistor configuration is just in keeping with that...as is the .01uf cap off the juncture of those two resistors on the plate.

Now, whether or not this is essential in getting the 6G3 tone, I'm not sure. What I do know is that, when I had the PI closer to what the actual 6G3 uses (6800 into a 15k tail...82k/100k on the plate caps), this 12AX7 channel sounded PHENOMENALLY cool in the gain department, albeit way to dark in terms of the sweep of the tone control. Putting both channels into a more Voxy PI (47k tail, no feedback, 100k/100k on the PI plates) was good as well, and seemed to "feel" better as far as working with both channels.

So, I'm still not sure how to best set this all up. I'd like to keep the Brown Deluxe side as true to the real thing as possible, yet still able to blend well with the 6SN7 channel. When both share the same PI input I tend to get a harsher overall midrange response from the amp, as well as an overly bassy 12AX7 side. When they each have their own PI input, the midrange edge goes away for the most part and the tone pot is way more usable, yet the 12AX7 side seems thinner in the low end and the channels cancel each other out when used at the same time.

[Wayne]

I dunno. I suppose you could try to find some "non-brown-deluxe" tone control values to compensate for the change in tone that the extra channel brings - I don't even know what direction to suggest you go in that department.

Thinking waaaay outside the box, I believe there's a 12 pin tube that has 3 'AX7 type triodes in it. The third one could provide phase inversion for the octal triode channel. No new chassis holes, but would require enlargement of an existing hole.

Another option might be to lose the rectifier tube in favor of some "sand", thereby freeing up a spot for another tube but again, I don't know how crucial the recto tube is to the "brown sound".

At any rate, looks like a cool & unique amp - you've probably got some tones in there that you can't get off a shelf.

W

[dehughes]

Cool man. Thanks for the suggestions. I'll land somewhere...