EF86 MODERN CLASSIC--CHANNELS MIXABLE?

18watt-specific Tech Talk - Building, Fixing, Parts, Mods...

Moderators: CurtissRobin, colossal, zaphod_phil, Daviedawg, Graydon

Post Reply
ViperDoc
Occasional poster
Occasional poster
Posts: 195
Joined: Mon 05/18/20 8:09 pm

EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by ViperDoc »

I've never built an amp with an EF86. The EF86 Modern Classic from this site has an EF86 channel tapped into one side of the PI, and the Modern Classic channel into the other side of the PI. Assuming the EF86 plate signal is phase reversed from its grid input, does that mean these channels are mixable, or are they not mixable? Are these channels in phase if they both feed plate driven signals into opposite sides of the PI tube? Thank you!
0 x

User avatar
Bieworm
Verbose Moderator
Verbose Moderator
Posts: 2287
Joined: Mon 02/10/20 8:24 am
Location: Belgium

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by Bieworm »

If you look at the schematic you'll see that both channels have only 1 plate in the signal path. I'm not familiar with pentodes in the preamp, but to me it looks like both signals are in phase arriving at the PI.
I don't know if the signals can be mixed on both sides of the PI. Maybe someone with more knowledge can chime in.
If I was to do this and wasn't sure, I'd join both channels before the PI with mixing resistors and use only one half of the PI...
You can also cascade the MC channel into the EF86 channel.. I've done that with a normal MC and it was pretty great like that...
0 x
"THIS should be played at high volume..preferably in a residential area"

ViperDoc
Occasional poster
Occasional poster
Posts: 195
Joined: Mon 05/18/20 8:09 pm

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by ViperDoc »

That was my thought exactly. Wondering what the sages would say.
1 x

User avatar
TriodeLuvr
Frequent poster
Frequent poster
Posts: 616
Joined: Tue 10/16/18 11:19 am

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by TriodeLuvr »

ViperDoc wrote:
Sun 10/17/21 11:38 pm
Assuming the EF86 plate signal is phase reversed from its grid input, does that mean these channels are mixable, or are they not mixable? Are these channels in phase if they both feed plate driven signals into opposite sides of the PI tube? Thank you!
files/JMPGuitars_18_Watt_EF86_Modern_Cl ... ematic.pdf

Both channels invert. Each has a somewhat different amount of total shift, due to the tone controls and contouring, but I don't think that's the major concern. Essentially, if the same frequency is fed in phase to both channels, the outputs of the PI will null. This is the opposite of a resistive mixer in which in-phase components add.

It's not clear to me whether this will create an audible effect unless you drive both channels simultaneously from the same source. As long as both inputs are from separate sources generating complex waveforms, such as guitars and mics, frequency and phase will always be changing between the two inputs to the PI, and the outputs will never null.

Jack
0 x

ViperDoc
Occasional poster
Occasional poster
Posts: 195
Joined: Mon 05/18/20 8:09 pm

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by ViperDoc »

OK, super. That was what I wanted to know. So, a sine-wave fed from the same source to both channels fed into opposite sides of the PI tube will cancel. So a resistive mixer is the way to go. What about this?

[DELETED--SEE UPDATED FILE BELOW]
Last edited by ViperDoc on Wed 10/20/21 9:03 pm, edited 2 times in total.
0 x

User avatar
Bieworm
Verbose Moderator
Verbose Moderator
Posts: 2287
Joined: Mon 02/10/20 8:24 am
Location: Belgium

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by Bieworm »

You drew two different values for the mixer. I don't recall any schematic that doesn't use two resistors of the same value. I see no reason to use different values.
0 x
"THIS should be played at high volume..preferably in a residential area"

ViperDoc
Occasional poster
Occasional poster
Posts: 195
Joined: Mon 05/18/20 8:09 pm

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by ViperDoc »

Bieworm wrote:
Wed 10/20/21 3:05 am
You drew two different values for the mixer. I don't recall any schematic that doesn't use two resistors of the same value. I see no reason to use different values.
That was arbitrarily put in, I forgot to notate that on my print. I recall seeing an asymmetrical resistor mixer used to tame the mixed input of a high-gain channel, such as an EF86. That would have to be tested.
1 x

User avatar
TriodeLuvr
Frequent poster
Frequent poster
Posts: 616
Joined: Tue 10/16/18 11:19 am

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by TriodeLuvr »

There's an error in the revised schematic. The grid of the second triode is shown grounded. It needs to be bypassed to ground with a capacitor instead. A 0.1uF will work. Also, you might want to add a connection dot at the junction of the resistors above the tail.

Jack
1 x

ViperDoc
Occasional poster
Occasional poster
Posts: 195
Joined: Mon 05/18/20 8:09 pm

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by ViperDoc »

TriodeLuvr wrote:
Wed 10/20/21 1:18 pm
There's an error in the revised schematic. The grid of the second triode is shown grounded. It needs to be bypassed to ground with a capacitor instead. A 0.1uF will work. Also, you might want to add a connection dot at the junction of the resistors above the tail.

Jack
WRT the second PI triode, you are correct, thanks for catching that. I will correct it. Is the .022 uF cap appropriate from the trem channel into the resistor mixer?
0 x

User avatar
TriodeLuvr
Frequent poster
Frequent poster
Posts: 616
Joined: Tue 10/16/18 11:19 am

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by TriodeLuvr »

ViperDoc wrote:
Wed 10/20/21 3:33 pm
TriodeLuvr wrote:
Wed 10/20/21 1:18 pm
There's an error in the revised schematic. The grid of the second triode is shown grounded. It needs to be bypassed to ground with a capacitor instead. A 0.1uF will work. Also, you might want to add a connection dot at the junction of the resistors above the tail.

Jack
WRT the second PI triode, you are correct, thanks for catching that. I will correct it. Is the .022 uF cap appropriate from the trem channel into the resistor mixer?
That cap is driving a very high impedance, so .022uF is quite a bit larger than it needs to be. I'd use a .01 in that spot.

Jack
0 x

ViperDoc
Occasional poster
Occasional poster
Posts: 195
Joined: Mon 05/18/20 8:09 pm

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by ViperDoc »

How does this look?
EF86 MODERN CLASSIC R-MIX 2.0.pdf
You do not have the required permissions to view the files attached to this post.
0 x

User avatar
TriodeLuvr
Frequent poster
Frequent poster
Posts: 616
Joined: Tue 10/16/18 11:19 am

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by TriodeLuvr »

I think that's all correct.
1 x

ViperDoc
Occasional poster
Occasional poster
Posts: 195
Joined: Mon 05/18/20 8:09 pm

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by ViperDoc »

TriodeLuvr wrote:
Wed 10/20/21 4:27 pm

That cap is driving a very high impedance, so .022uF is quite a bit larger than it needs to be. I'd use a .01 in that spot.

Jack
I frequently encounter things I don't understand, such as high vs. low impedance at certain points in a circuit. How do you see that? I'd like to know more.
0 x

User avatar
TriodeLuvr
Frequent poster
Frequent poster
Posts: 616
Joined: Tue 10/16/18 11:19 am

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by TriodeLuvr »

ViperDoc wrote:
Thu 10/21/21 1:51 pm
TriodeLuvr wrote:
Wed 10/20/21 4:27 pm

That cap is driving a very high impedance, so .022uF is quite a bit larger than it needs to be. I'd use a .01 in that spot.

Jack
I frequently encounter things I don't understand, such as high vs. low impedance at certain points in a circuit. How do you see that? I'd like to know more.
Well, the input impedance of the LTP itself is much higher than its 330K input resistor (R16) because the cathodes are bootstrapped by the signal current through the two triodes. This high input impedance of the LTP is effectively in parallel (relative to AC ground) with the 470K mix resistor from the EF86 channel. So, the total impedance at the input to the LTP is probably about 250K. The cap is driving this through the 330K mix resistor, for a total driven impedance of 580K. That's more than high enough for a 0.01uF cap to conduct the lowest frequency of a six-string without loss. Bear in mind, these are all rough numbers off the top of my head. I would need to run this through a simulator to be more exact, but rest assured the 0.01uF will work well.

Jack
3 x

ViperDoc
Occasional poster
Occasional poster
Posts: 195
Joined: Mon 05/18/20 8:09 pm

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by ViperDoc »

I see a filter capacitor as the last device in the circuit before injecting into the phase inverter on most schematics. Would it be any different to resistor-mix both channels before heading into such a filter cap and THEN into the PI, or keep it as in the current schematic as above? Would each filter cap BEFORE the resistor mix protect each channel?
0 x

User avatar
colossal
Moderator
Moderator
Posts: 450
Joined: Thu 05/10/07 2:00 am
Location: Moving through Kashmir

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by colossal »

ViperDoc wrote:
Tue 10/26/21 8:15 pm
I see a filter capacitor as the last device in the circuit before injecting into the phase inverter on most schematics. Would it be any different to resistor-mix both channels before heading into such a filter cap and THEN into the PI, or keep it as in the current schematic as above? Would each filter cap BEFORE the resistor mix protect each channel?
You mean coupling capacitor in that context. That cap is coupling one stage to another, not filtering the ripple on the DC supply. It is blocking DC from the plate from the next section and is critical for the correct operation of the LTPI.

The design choice of one channel per side of the phase inverter allows the user to A/B and so gives two distinct tonal voicings. This makes for a killer sounding and versatile amp. Dial one side for your cleans and one for your overdrive. Set up effect chains on each. Simple and easy. Each stage is capacitively coupled to the phase inverter.

You can also resistively mix the channels which allows jumpering to run them in parallel, much like a Plexi. The resistors are not one-way check valves, however. That is, when you turn channel B down, it affects A, or if you turn A up, it affects B; they are interactive. You can also cascade the channels with a relay or switch. This might be useful for some wanting to get their clean from say a 12AX7 channel and then cascade that into a pentode channel for their lead tone. You have coupling capacitors "upstream" of your Volume/Gain controls, otherwise you would have DC on your pots which would be both noisy and unsafe.
3 x

ViperDoc
Occasional poster
Occasional poster
Posts: 195
Joined: Mon 05/18/20 8:09 pm

Re: EF86 MODERN CLASSIC--CHANNELS MIXABLE?

Post by ViperDoc »

Many thanks as always, Colossal.
1 x

Post Reply