3 stage overdrive channel

[Lynxtrap]

I'm a little frustrated at the moment.
I'm modifying this old amp. What I've got is 3 gain stages (12AX7), a Fender two-knob tone stack, a gain pot and a master volume into a cathodyne.
With these ingredients I'm trying to make a lead channel (preamp overdrive), in the Marshall camp soundwise. Like, how hard can that be? Quite hard it seems.

Are there any known good sounding circuits of this kind and within the limitations mentioned above that I might want to look at?
2203 as a basis doesn't work great without the cathode follower it seems. That was sort of my starting point.
Didn't sound great, not enough gain from the "cold clipper" to push the third stage, so I started to change component values around.
It's been all downhill since

[tristanc]

Check out the Comanchero: http://ax84.com/associateprojects.html

[roberto]

Easiest way is to replicate channel III of an Engl preamp, just don't trust the ones you can find with google, they are totally unaccurate in terms of coupling caps (680 pF is the right one for most of their lead channels).
100k 1k5//22u
680pF into 250k log pot
100k 1k5//22u
22n (try even lower down to 1n) into 100+100k voltage divider
100k 1k//22u
plate drive tonestack
the concertina (don't economize on the value of the grid stopper)

[Lynxtrap]

Thanks!
I ended up with this.
The main challenge is that the tone stack sucks a lot of gain. This is not a 60W amp anymore, but I think it puts out enough for most situations.
I still thought that placing it after the third gain stage got me the best overdrive tones so I decided to commit to that and work around the problems as best I could.
If anyone has suggestions for improvements, you are welcome

Impact 60_TF_rev2.jpg

[pompeiisneaks]

Thanks!
I ended up with this.
The main challenge is that the tone stack sucks a lot of gain. This is not a 60W amp anymore, but I think it puts out enough for most situations.
If anyone has suggestions for improvements, you are welcome
Impact 60_TF_rev2.jpg

tone stacks almost always do because they load down the next stage. This is why on a lot of amps that don't have that problem they do a cathode follower before the tone stack, this sets up the right current source to keep it from loading down the next stage. Vox AC30 uses it as to many others I'm sure.

If you can add in another triode, that would allow for half a stage to do the cathode follower, and maybe another half stage to change from a cathodyne to a LTP PI. (but that's also entirely up to your needs)

~Phil

[pdf64]

Do you really need to have the normal channel in use at the same time as the overdrive channel? Bear in mind they're in opposing polarity.
It's just that the 220k - 470k channel mixer resistors are losing you ~1/3 of the max available signal level from the overdrive channel, ~2/3 from the normal channel.
A switch to select one channel or the other would get around that, perhaps relay operated, or vactrols if you can get them.
The OD tone stack may be more lossy than necessary, due to the 100k slope resistor; consider bringing it down a bit (eg 68k, 56k) and retweaking all the other stack component values as required.
And as there's not much gain after it, a linear taper master vol control may provide a better 'user interface'.

It would be a lot better to put the bias supply feed resistor BEFORE the standby switch, as the power tubes will be having a very rude awakening in it's current arrangement

[Lynxtrap]

Do you really need to have the normal channel in use at the same time as the overdrive channel? Bear in mind they're in opposing polarity.
It's just that the 220k - 470k channel mixer resistors are losing you ~1/3 of the max available signal level from the overdrive channel, ~2/3 from the normal channel.
A switch to select one channel or the other would get around that, perhaps relay operated, or vactrols if you can get them.
The OD tone stack may be more lossy than necessary, due to the 100k slope resistor; consider bringing it down a bit (eg 68k, 56k) and retweaking all the other stack component values as required.
And as there's not much gain after it, a linear taper master vol control may provide a better 'user interface'.

It would be a lot better to put the bias supply feed resistor BEFORE the standby switch, as the power tubes will be having a very rude awakening in it's current arrangement

Now you're getting me ready to pull out that chassis once again...
So you're saying that the mixer resistor from the clean channel is a path to ground for the lead channel...back to where? The plate?
This mixing solution is seen in loads of old two channel amps, I've never regarded those mixers as voltage dividers.

RE the slope resistor, I have learned that a larger value = less load (higher impedance)? I deliberately went up from 39K to 100K with that in mind.

Well, I had thought of a channel switching option. Where I would tap the signal after the clean channels 1st stage, into the two stages of the lead channel, and back into the clean channel 2nd stage. Maybe I should have done that after all...

@pompeiisneaks: Yes I would put 1-3 more tubes in there if I wanted to make the ultimate overdrive amp.

Think I'll have to show a little history of this amp. It's an English amp made in 1969, Partridge transformers, two identical channels and looks about as cool on the inside as it does on the outside.

When I bought it a couple of weeks ago it was basically untoched, down to electrolyctics dated 1968.

The stock circuit was so bright it was unusable. As I thought there were Marshall-like tones hiding in it, I started digging them out.
I don't want to ruin the appearance, and I like to keep as many original components as possible on the tagstrip board.
So I wanted to stay within those confines. But swapping a switch for one of the input jacks would be fine.

Someone had already drilled a hole in the back where I put the master volume.
I thought I could do channel switching with my ABY box.

Here is the original schematic and some pictures.

Impact 60 A-S.JPG

Impact 6.jpg

Impact inside_2.jpg

Impact 3.jpg

Impact_1.jpg

[Reeltarded]

As said, dropping that slope gets more shoulders in the mids. The "too bright" is not a problem then. 47-56k is where I start to like it on the way up.

Switching is fne, but cascading is better.

Cascaded, I would make it normal channel stays the same, other channel switches in series like Dumble ODS. Leave the second stack, that one is where to drop the slope. The second set of input jacks become drive controls, the eq is broken out to TMB on that channel. Will be sort of like my blonde Bandmaster was previously, but more stages.. yikes!

[martin manning]

So you're saying that the mixer resistor from the clean channel is a path to ground for the lead channel...back to where? The plate? This mixing solution is seen in loads of old two channel amps, I've never regarded those mixers as voltage dividers.

Time to start! A DC voltage is as good as ground to AC.

[Lynxtrap]

So basically, the lead channel sees 570K to ground?

I don't think I will reinvent the circuit right away, but we'll see.
At least I will try to lift the clean channel just to see how much more volume I get from the lead channel to the output, and vice versa.

I guess a temporary solution would be to put a switch between the mixer resistors just to prevent the bleeding to ground?

[Xander8280]

I recently did a jcm800/2203 style preamp with the tone stack plate driven to a master volume in a 5E3 phase inverter/power amp. I am really surprised and satisfied, I think a larger value middle pot is good for getting back to 5E3 tones with using the low sensitivity input.

[pdf64]

So basically, the lead channel sees 570K to ground?...

Do you mean after its 220k mixer, at the LTP input?
If so, close but not quite there, it looks like you are adding the normal channel's 100k plate resistor (which goes to HT = ac 0V due to the big decoupling caps) to the normal channel's 470k mixer; that's part of the picture, but ignores the plate impedance, around 60k.
The output / source impedance of a regular 12AX7 stage is the plate impedance // plate resistor, so ~40k total. Aiken goes through it all here http://www.aikenamps.com/index.php/desi ... amplifiers
So the normal channel's total source impedance, ie looking back from the LTP input, is 470k + 40k = 510k

Whereas the output impedance of the overdrive channel is dominated by the 1M pot; at very low settings it will be pretty much set by the resistance from the wiper to 0V common, at around the electrical mid point it will be ~500k, above that it will fall again, down to its level at max setting, which will be complex due to the tone stack. That will be dependant on the treble control setting, and, to keep this from going on much more, assuming non-extreme settings, will probably be in the range 50k - 200k.

So the degree of attenuation that the normal channel gets from the overdrive channel varies a bit, according the the tone and master settings.

...a temporary solution would be to put a switch between the mixer resistors just to prevent the bleeding to ground?

If there's a switch to select channels, I can't see any point in leaving the mixer resistors in there?

[Lynxtrap]

So any two-channel Fender BF amp loses almost half it's preamp gain the way the channels are joined? I can't believe I have never encountered, not to mention realized, this basic fact before.

And yes, the mixer resistors are unnecessary if I install a switch.

[pdf64]

It’s only a 6dB loss; most folks barely notice that same difference between the 1 and 2 jacks of the same channel

[Lynxtrap]

My dilemma is that I'm hesitant to install relays/optos for channel switching. I thought I'd just keep two separate channel and be able to switch between them with an A/B pedal. If changing channels still requires a switch, that idea is sort of defeated.

And it's not like it doesn't work the way it is, I'm even quite happy with the sound. It's just that there could be more gain into the output stage. I suppose this is more like a 10-15W amp right now with almost no capability to overdrive the EL34's.

Perhaps I should wire in a 3-way switch with the options of each channel separately and both through the mixer resistors for a "low power mode"