ECC81 instead of ECC83 which modifies is to be done as to achieve it ?

[kagliostro]

This circuit is proposed by Mr. Merlin

He specifies that also an ECC81 or ECC832 (12DW7) will be ideal for this circuit

but my knowledge is not enough and the use of the Led and of a virtual-earth confuses me further

I would like to know which modifies are to be done

as to mount an ECC81 instead of the ECC83 that is shown on the schematic

Many Thanks

K

[sluckey]

I would not modify anything. Just plug'em in. As for the 12DW7 (which is basically half a 12AX7 and half a 12AU7), I'd wire the circuit so the 12AU7 side (pins 1, 2, and 3) will be the cathode follower.

[kagliostro]

Hi Sluckey

Thanks for your council

My confusion is also due to the fact that the ECC81 is used as CF for his high current performance, so I was wondering if inserting it in the circuit without mods will give an increase in the current available at the CF output

The proposal of Mr. Merlin of a good CF send circuit using ECC81 is very different from the arrangement on the previous schematic (in which an ECC83 is show)

you can see it in this schematics (two different versions)
(that are claimed to "extract" the better performance from the ECC81 as send)

Grazie ancora

Franco

p.s.: Merlin's books are rich of many ideas and my council is to have both and read with attention

[HotBluePlates]

My confusion is also due to the fact that the ECC81 is used as CF for his high current performance, so I was wondering if inserting it in the circuit without mods will give an increase in the current available at the CF output

What will you do with extra current?

Trick question, because you don't really care about the circuit's current output. What you do care about is the cathode follower having a nice low output impedance to drive long cables easily. An ECC81/12AT7 has a lower plate impedance and more transconductance (Gm) than an ECC83/12AX7, so it will result in a cathode follower with a lower output impedance. No circuit changes are necessary (though I suppose you could change some things if you needed ultimate performance from the ECC81, but it's really not necessary in a guitar amp).

The proposal of Mr. Merlin of a good CF send circuit using ECC81 is very different from the arrangement on the previous schematic (in which an ECC83 is show)

It's only different in the details, after you know what you're looking at.

Look at the first circuit you posted.

Pretend the cathode follower is a Fender split-load inverter without a plate load. The 1M grid resistor is returned to the bottom of the "cathode bias resistor" instead of ground, with the cathode load between that point and ground. The cathode load in this case are the 10k + 47k resistors, in series.

The "cathode bias resistor" is an LED. When enough voltage appears across the LED, it will turn on and conduct, and will flow as much current as required to maintain that voltage. I don't know enough about LEDs to tell you without looking what voltage a red LED conducts at, but we can guess it's between 1-2v. The fact it will maintain this voltage for any reasonable current means it acts like it is a cathode bias resistor with a bypass cap; so you use 1 cheap part instead of 2 slightly more expensive parts.

Now this effects loop is probably placed late in the preamp circuit, maybe right before the phase inverter. The signal level is big compared to a guitar's pickup, and you're probably using effects pedals in the loop. So along with a low output impedance, you need to drop the signal strength. This is done by dividing the cathode load in 2 parts; the output in this case is about 1/6th the input to the cathode follower, and then has a pot to allow you to reduce the signal further if needed. A coupling cap keeps d.c. at the cathode load out of your pedal's input circuit.

On the Return side, there is a virtual earth mixer, as you noted. Both signals are mixed only if you have selected the parallel loop setting. In essence, this tube is being run as an operational amplifier (opamp), and the gain of the stage is determined by the series resistor and the feedback resistor. The feedback resistor in this circuit is 1M; The series resistor for the Return jack is 10k, with 100k for the parallel path. The gain is 1 + feedback/series resistor, or 1 + 1M/10k = 101 for the return jack (really just the gain of the ECC83 at around 50-60, or essentially no feedback) and 1 + 1M/100k = 11 for the parallel path. That's because the return jack has an output at pedal-level, and needs to be boosted back up.

Which is why Merlin recommends a ECC83 or ECC832/12DW7 for the loop; you need the extra return gain, though the low output impedance of the 12AU7 as a cathode follower might be helpful.

[HotBluePlates]

The proposal of Mr. Merlin of a good CF send circuit using ECC81 is very different from the arrangement on the previous schematic (in which an ECC83 is show)

you can see it in this schematics (two different versions)

So if you look again, knowing what you now know about the complete loop, these ECC81 circuits aren't that different.

1. They don't use an LED as a cathode bias component, but use a regular resistor.
2. They use a smaller cathode load, because the ECC81 can, and that results in a smaller output impedance.
3. Loop #1 has the same pot to reduce the signal, but doesn't split the cathode load to get a reduction of output signal. This assumes you want to be able to get the full output signal size (which will be only slightly smaller than the full input signal size). Loop #2 moves the pot to a point ahead of the send jack because the resistor and pot in Loop #1 is in paralle with the cathode load and reduces the size of the maximum signal output of Loop #1. Again, this assumes that you want to be able to get the biggest output signal possible from the cathode follower.

As I pointed out, the complete loop you originally posted makes a different assumption: that you don't want the maximum signal output possible, because you need to drop a line-level signal down to pedal-size. Doing that with just a pot as in your later examples may mean you keep it turned down to 1-2, where it's harder to control the usable output range. The later examples may be more suitable to feed to the input of a device that needs line-level signals.

[kagliostro]

Many Thanks for the explanations HotBluePlate

There is one thing I don't understand well

You say that what is important is to have a low output impedance as to drive correctly the external load, but for that purpose isn't really requested also a high current ?

Or a high current is more useful only for rack effects units that are far from the amp and so require long cables ?

---

I'm wondering about the use of a 12dw7 as the best tube for this purpose, something like the loop #2 as send circuit with the 12au7 part of the 12dw7 and the return circuit taken from the first schematic (the complete one with ecc83 shown) with the 12ax7 part of the 12dw7

do you think this way it will be possible to have an FX loop who is suitable for pedals and line level units (like rack effects)

may be it will be better to insert in the send part of the circuit a voltage divider as to have two separate outputs one for each kind of load (pedals and line level units) ?

K

p.s.: Merlin indicates red Led at 1.6v and red Super Bright Led at 1.8v

[HotBluePlates]

There is one thing I don't understand well

You say that what is important is to have a low output impedance as to drive correctly the external load, but for that purpose isn't really requested also a high current ?

Or a high current is more useful only for rack effects units that are far from the amp and so require long cables ?

Ohm's Law.

If you have a set amount of voltage, and you move from a high resistance (impedance) to a lower resistance (impedance), what happens to current?

I'm wondering about the use of a 12dw7 as the best tube for this purpose, something like the loop #2 as send circuit with the 12au7 part of the 12dw7 and the return circuit taken from the first schematic (the complete one with ecc83 shown) with the 12ax7 part of the 12dw7

do you think this way it will be possible to have an FX loop who is suitable for pedals and line level units (like rack effects)

may be it will be better to insert in the send part of the circuit a voltage divider as to have two separate outputs one for each kind of load (pedals and line level units) ?

Sounds good.

Overall, you could probably use a relatively low-value total cathode load (say 15-22k), while splitting it into 2 resistors. Use a switch to select whether the pot is fed from the cathode output or the junction of the split cathode load to give 2 different total output levels.

It's hard to make these kinds of decisions without deciding exactly what you're starting with (where in the amp you're inserting this loop) and what you'll send the signal to.

[Willabe]

Ohm's Law.

If you have a set amount of voltage, and you move from a high resistance (impedance) to a lower resistance (impedance), what happens to current?

It increases because of the lower resistance?


      
                  Brad     

[tubeswell]

Ohm's Law.

If you have a set amount of voltage, and you move from a high resistance (impedance) to a lower resistance (impedance), what happens to current?

It increases because of the lower resistance?

I think ^this^ is a simple effective way to remember the E=I x R, I = E/R, R=E/I relationship

[HotBluePlates]

Ohm's Law.

If you have a set amount of voltage, and you move from a high resistance (impedance) to a lower resistance (impedance), what happens to current?

It increases because of the lower resistance?

Right.

So it's not that we necessarily care to think about the "higher current apabilities" of the 12AT7/12AU7, but rather the lower impedance available with these tubes which is a byproduct of their lower internal plate resistance.

[kagliostro]

It's hard to make these kinds of decisions without deciding exactly what you're starting with (where in the amp you're inserting this loop) and what you'll send the signal to

In this days I'm drawing (for a guy in an italian forum) a modded 5F1 + FX Loop

in that amp we planned to use the active FX Loop used by Tubenit that is something like a Dumbleator

in the modded 5F1, due to the presence of the NFB I moved the FX Loop between V1a & V1b

But talking about the Merlin's FX Loop, I'm thinking to a different schematic, where the FX Loop will be between the preamp and the PI or 1/2 (the send) on a stand alone preamp and 1/2 (the return) on a PA

with this architecture (stand alone preamp) I'm interested to have both signal levels for pedals and for rack effects units (or line level units)

K

[HotBluePlates]

I'm interested to have both signal levels for pedals and for rack effects units (or line level units)

1. Start with the schematic you first posted.
2. Disconnect C3 from the junction of the 47k and 10k cathode load resistors.
3. Connect the center (switched) lug of a single-pole double-throw (SPDT) switch to the free end of C3.
4. Connect one outer end of the SPDT switch to the junction of the 47k and 10k resistors, where C3 used to go.
5. Connect the other outside lug of the SPDT to the junction of the 47k resistor and LED.

You can now switch between full output signal and pedal-level signal, with each having the ability to be adjusted with the Send level pot.

If the pedal-level maximum signal is too high or low, adjust the ratio of the 47k and 10k resistors while keeping the total resistance of the pair at ~60kΩ.

[kagliostro]

OK, that is a nice solution see "Line - Pedal Levels" image, I like it ! Thanks


K