Screen and Plate Switch pop in AC15 build

[shaun]

Heyo. I have a sneaking suspicion there is no quick fix for this pop noise, but I bow to superior wisdom.

So, I built the Sluckey AC15R a while back, and it's a great amp - reverb to the max. It does seem to have early break up though. In an attempt to create more headroom, I traded the EF86 1M screen resistor to 470k, and I changed the 220k plate resistor to 100k. This tamed the break up nicely without sacrificing too much tone.

Then I thought it would be nice to make it switchable, so I added a DPDT switch and wired it so when it's in one position, it runs the high gain option as per the original 1M and 220k resistors. In the other position, it runs the low gain option as follows:  I ran the B+ to a second 1M resistor and a second 220k resistor, so that when the switch is thrown, the screen and plate resistors kick into parallel mode with their twin, thus halving their resistance. It works fine, except for the nasty pop in one direction and snap in the other.

I've researched pop noise issues up the wazoo, and many builders seem to have a pop issue with cathode switching. But not so common - possibly because it's a bad idea -  is that I'm trying to switch DC current at the plate and screen. I guess the DC means a capacitor to ground won't help, as it would simply block DC current. I've tried a variety of resistor placements to no avail. So I'm starting to conclude that I cannot do what I want. Again 

Should I abandon this idea and try something else?

Thanks as always for the help.

[tubeswell]

Maybe short the 1M screen resistor with a DPDT that also cuts the screen bypass cap (on the other pole*). The 1M resistor would thus be wired permanently into the circuit and perform a dual function as a switch noise snubber.

(* the two poles of the DPDT could be wired so that one pole is open while the other pole is closed. That way you could also put 2M2 across the other pole to snub any switching noise from the cap).


See also attachment excerpts from Merlin Blencowe's 1st edition Designing Tube Pre-amps book (book no longer in print)

[shaun]

Wow. Great stuff. Thanks Tubeswell. Looks like gold in them thar pages. And that's an interesting suggestion - I had considered the screen bypass cap but didn't see it as part of the solution.  Much appreciated.

[pdf64]

A sudden change to DC operating point will pretty much inevitably have some sort of audio consequence - pop, click, thump etc. Commercial amps might use a momentary mute circuit. If you find the switch function useful but can't be bothered with mute circuits, just turning the appropriate vol control down would probably work pretty well too 

[shaun]

Actually, pdf64, that thought crossed my mind .

But because I was unsure of the damage I might be doing in the circuit due to current possible surge (just guessing here), I though it would be best if I fully understood what I was doing. When this comes up in a future build - which I'm sure it will - I won't be quite so in the dark.

That snippet of info you offer about the momentary mute circuit has got my interest, especially as I've haven't encountered one before. A quick bit of research shows they can be quite complex or quite simple. If I want to avoid adding a relay, do you have any suggestions as to where I might start if I wanted to add an MM circuit? ie., do you, by any chance, happen to know of a schematic for a simple and effective MM circuit? There's a bunch I found online for various purposes, although they get quite complex. But perhaps that's the nature of the beast.

Thank you!

[pdf64]

Haha, you've put me on the spot, I'm afraid it's something I've not taken much notice of.
Offhand I don't know for sure, but it may be worth having a nose at the 6505 or almost any Blackstar amp.

[shaun]

Maybe short the 1M screen resistor with a DPDT that also cuts the screen bypass cap (on the other pole*). The 1M resistor would thus be wired permanently into the circuit and perform a dual function as a switch noise snubber.

(* the two poles of the DPDT could be wired so that one pole is open while the other pole is closed. That way you could also put 2M2 across the other pole to snub any switching noise from the cap).


See also attachment excerpts from Merlin Blencowe's 1st edition Designing Tube Pre-amps book (book no longer in print)

Hi Tubeswell. I was a little confused about your wiring idea - but that's my fault for not loading a schematic to reference. The one attached is my first attempt at using Jschem. I can't figure out how to erase my wiring mistakes, but hopefully it makes sense.

Does your suggestion mean a second DPDT switch?

[tubeswell]

Maybe short the 1M screen resistor with a DPDT that also cuts the screen bypass cap (on the other pole*). The 1M resistor would thus be wired permanently into the circuit and perform a dual function as a switch noise snubber.

(* the two poles of the DPDT could be wired so that one pole is open while the other pole is closed. That way you could also put 2M2 across the other pole to snub any switching noise from the cap).


See also attachment excerpts from Merlin Blencowe's 1st edition Designing Tube Pre-amps book (book no longer in print)

Hi Tubeswell. I was a little confused about your wiring idea - but that's my fault for not loading a schematic to reference. The one attached is my first attempt at using Jschem. I can't figure out how to erase my wiring mistakes, but hopefully it makes sense.

Does your suggestion mean a second DPDT switch?

No. Just one DPDT. I’ll knock up a schematic when I get back to my desk.


(To undo lines/shapes etc in JSchem, use the pointer function to click on the line/shape, which highlights the thing, and press the delete key)


Edit: Here ya go (Disclaimer: there may still be some pop)

[HotBluePlates]

Does your suggestion mean a second DPDT switch?

You can go much simpler:  AC10

I've got a 1965 AC10, and the Vibrato Amplitude control (R12) acts like a Gain/Headroom control for the EF86.  You could copy that circuit, but simply don't have a wire running from the EF86's cathode out to some other tube.

No switch, no pop, and continuously-variable control.  Even with the indicated 470kΩ plate resistor, the EF86 will have a setting where gain/output is lowest and headroom is highest.

[shaun]

Haha, you've put me on the spot, I'm afraid it's something I've not taken much notice of.
Offhand I don't know for sure, but it may be worth having a nose at the 6505 or almost any Blackstar amp.

No worries, pdf64. There's so much info available online. In fact, someone is discussing this very thing in one of the other recent threads in this forum. I doubt I'll accomplish it with this build, but it's very interesting idea.

[shaun]

Thanks, Hotblueplates and Tubeswell. You seem to have delineated the best ways to add adjustment capabilities to the EF86. I hadn't seen the AC10 circuit before. I might try that if I can find room to mount another pot. Much appreciated. Thanks for drawing that schematic, Tubeswell - it made me laugh, seeing how nice it looks compared to my pitiful attempt. But I suppose I can only get better at Jschem.

[shaun]

So, I installed Merlin's  figure 3.26b from "Switchable Pentode Designs," and it pops very loudly, despite the explanation to the contrary.
Disappointing.
Perhaps the AC10 is the better way to go - no switching.

EDIT: I installed the 3.26b from Merlin's book, and it pops like crazy too. However, sonically it is pretty great, so if it pops, it pops. Pops be damned, I say.

It seems to me that anytime there is a sudden opening or closing of a relatively high current DC line, this issue will arise, especially near a signal per tube. I guess Vox figured this out and adopted their vibrato amplitude in the AC10, using the slow adjustment of the pot to avoid the surge created by a switch. Thanks again for all the help.

[HotBluePlates]

It seems to me that anytime there is a sudden opening or closing of a relatively high current DC line, this issue will arise, especially near a signal per tube. ...

I agree:  it's the sudden-change of d.c. voltage, throwing a transient in the system.

... I guess Vox figured this out and adopted their vibrato amplitude in the AC10, using the slow adjustment of the pot to avoid the surge created by a switch. ...

Nah, they just used a pot like everyone else does for trem intensity.

The only thing unusual about their approach is coupling the trem into the cathode of a pentode, and them figuring out they can control the effective strength of the trem by manipulating the pentode's screen voltage.

The way I'm using that control is entirely an unintended side-effect.