cathodyne/LTP switch (for small P-P amp)

[deci belle]

My 5e3-thing has tweed and brown deluxe preamp controls (two volume/tone pots per channel~ it still does that interactive thing).

Would anyone please mention how I might be able to switch between the normal 5e3 PI and an LTP PI?

Installing another tube isn't a problem!

[pdf64]

Using the same 12AX7? If so, it wouldn't be pretty!

[deci belle]

pdf64 said:

Using the same 12AX7? If so, it wouldn't be pretty!

deci said:

Installing another tube isn't a problem!

c'mon Pete!! I got an itchy chassis punch!!

[pdf64]

Just need a 3 pole changeover switch then; one pole for the input, the other 2 for the 2 outputs.
As you'll be running another tube from the same B+ node, the dropper resistor will need reducing in value; maybe increase the bypass cap value, to keep the time constant the same.

[deci belle]

haha!! Yay~ thank you for that, Pete!

Hmmm, so that's a triple pole/double throw (I guess an ON-OFF-ON would be best in a toggle)?

I have room to put in another node too— and I better, because the 5e3 circuit normally has one measly 16mfd cap supplying the current for 4 triodes already.

It's really that simple?!!

[pdf64]

Hmmm, so that's a triple pole/double throw (I guess an ON-OFF-ON would be best in a toggle)

Why use an ON-OFF-ON? I guess the OFF could be a mute type of standby. But a regular 3 pole ON-ON may be easier to source and cheaper.
Also consider rotary type switches.

I have room to put in another node too— and I better, because the 5e3 circuit normally has one measly 16mfd cap supplying the current for 4 triodes already

Note that most later BF/SF designs have 6 triodes on the last B+ node, again 16uF on the DR.
For the cap value, an issue may be how many of those stages are in cascade; with say the DR AB568 http://bmamps.com/Schematics/fender/...b868_schem.pdf the V2 sections and one section of V4 are cascaded, so 3 12AX7 sections, whereas for a for the 5E3, even your proposed scheme, there would only be 2 cascaded stages off the last node, and one of those stages may only be a 12AY7.
And the LTP / CC+cathodyne wouldn't be handling signal simultaneously.
So in the first instance I wouldn't bother with another node, rather just reduce the 22k dropper to maintain the same node voltage, eg ~15k.
Note that the extra current through the 1st B+ dropper 5k will drop a few more V across it, which will affect the power tube conditions, so you may wish to tweak its value.

I'm assuming that in your scheme, the 'cathodyne' mode includes the preceding V2 common cathode gain stage, and 'LTP' mode would be fed directly from the volume controls?
If the LTP was being fed from the V2 common cathode stage, it would add a heap more gain, eg ~30x=30dB, and stability may be a problem.
Whereas with the 'switching the whole V2' arrangement, the LTP mode would only be about 0.5x=-6db down on the (CC+) cathodyne mode.
Of course if you don't want to lose any gain with the LTP mode, then the common cathode stage after the volume controls could be left permanently in circuit, and a potential divider be put at the LTP input (eg 1M / 33k).

Whatever, it should be fairly simple, though careful lead dress will be beneficial, and the extra stuff required for the switching option will prevent the end result being 'exactly' the same as dedicated implementations.

You may want to check that the overall system signal polarity doesn't get flipped when changing modes.

[deci belle]

Oh wow! This really great, Pete— thank you for considering all those aspects I need to work with!

Ok, a rotary sounds good, especially if a chicken-head will fit its shaft.

Thank you for the schematic… yes I see about the cascaded triodes and since my standard circuit is what you understand it to be, I should not add another node (I've added a number of them over the years and I much prefer the way it sounds with the original (minimum) amount of filter-stages.

Thank you for suggesting the dropped value of the second (22k) power resistor (when you mentioned to do that in the earlier post, I didn't understand where you meant). I have a bunch of those 3w IRC resistors with values below 5k to work with too.

If the LTP was being fed from the V2 common cathode stage, it would add a heap more gain, eg ~30x=30dB, and stability may be a problem.
Whereas with the 'switching the whole V2' arrangement, the LTP mode would only be about 0.5x=-6db down on the (CC+) cathodyne mode.

That's a LOT of gain …not what I was after in this amp! And that 6db drop is quite noticeable too!

Dang, switching the whole V2 arrangement isn't what I envisioned, but you are…

…assuming that in [my] scheme, the 'cathodyne' mode includes the preceding V2 common cathode gain stage, and 'LTP' mode would be fed directly from the volume controls?

Where the PIs are fed directly from the volume controls looks like I will need to come up with an extra triode (for the 6G3's 2nd stage)— But you've thought of that with the potential divider be put at the LTP input (eg 1M / 33k).

I had only considered sending the signal to the switch from the 5e3's common 2nd stage .02uF coupling cap, but I don't know what the 1M / 33k potential divider looks like.

You may want to check that the overall system signal polarity doesn't get flipped when changing modes.

Ya, if so, then another tube might need to be used somewhere anyway to flip it back. That's ok, because the whole thing will be going into a larger 6G3 chassis/cab anyway.

Please tell me more about the potential divider, so I understand that clearly, Pete; is that the same as a voltage divider?
If so, I'm not sure how to implement it at the input to the LTP.

I know my plate voltage is 159v at pin 3 of the common 2nd gain stage: I get either 5v or 154v out of the voltage divider, but I don't know what to do with it…

[pdf64]

By potential divider, think of a volume control at a fixed setting with fixed resistors http://www.mrcorfe.com/DAOS/Year11/E...ialDivider.gif
R1=1M, R2=~33k
This would sit at the input to the LTP, after the input pole of the mode switch.

Regarding signal polarity, as there's no NFB loop, it can be easily flipped as required at the phase splitter outputs.

[deci belle]

Ya, I used an online voltage divider calculator and saw the classic schematic per the link, Pete.

Regarding signal polarity, as there's no NFB loop, it can be easily flipped as required at the phase splitter outputs.

Cool!!

Thank you so much!

[deci belle]

I needed to have a picture that applies to this situation to understand how to insert the voltage divider.

I suspected one of the legs of the input (of the voltage divider) needed to be grounded and finally found this picture to prove it (to myself).

Thank you so much for the 1M + 33k values and for indicating which one feeds the LTP, Pete!

[deci belle]

inverter SW.pdf

I drew a layout of the volume inputs, switch, both phase inverters and the two tubes. Does this look right, Pete?

[pdf64]

Sorry, I meant to sketch something up for you.
But yes, that's what I envisaged.
It will work but there will be a thump when the switch is operated, due to the change in Vdc on the cap at the 'common' switch terminal.
That could be prevented by adding a blocking cap (eg 0.1uF or 0.047uF) at the cathodyne grid, together with ground referencing 10M resistors at the 'in' side of that cap, and the 'out' side of the coupling cap from the previous stage (= at the switch common terminal).

[deci belle]

Will you please check my layout, Pete? I added the cap and resistors where I thought I should…

[pdf64]

The 0.1uF cap and its 10M resistor on the PI1 grid need their own terminal, as the cap's being shorted out currently.

[deci belle]

I think this will do it, Pete.



Please let me know if this one looks correct!