Single/Parallell switch pop & NFB switch pop

[Albi]

Hi guys,

I have been designing a circuit and am really happy with the end result and its getting positive responses locally. My final hurdle is a pop when switching. It's a clunky kind of DPDT switch which switches the first 2 stages to parallel or series mode. Before pulling it a part and redesigning the preamp I'm looking for help here to see if anything jumps out. Here's the area im looking at. Any advice greatly appreciated.

I'm having a similar problem with the NFB switch. I've added a cap and a resistor to add some presence to the NFB and I have gone with a switch where it can be switched off for an open amp. This isn't such a deal breaker just putting it out there.

see screen grabs for each section.

Thanks

Albi

[Stevem]

Put a 1 meg resistor across your switch / relay closure points .

[Albi]

Thank you Steve! I did try that but let me go back and look again and report back.

[B Ingram]

... a pop when switching. ... the first 2 stages to parallel or series mode. ...

If there is a change of DC volts, there will likely be a "Pop."

• Stage 2 has a switch changing which plate load resistor it uses (which also changes the direct current through Stage 1's cathode resistor). Removing this switch seems desirable from a "don't change DC Volts" perspective.

• There is a switch changing the Grid-to-Ground resistance used by Stage 2. Even though grid current should be zero, this seems undesirable from a "don't change DC Volts" perspective.

So what would I try?

• Provide a Grid-to-Ground resistor for Stage 2 (like 1MΩ or so).

• Provide a single plate load resistor for Stage 2 (probably the 100kΩ it uses when the stages are cascaded in-series).

• Place a coupling cap between Stage 1 and the switch: this already exists in the 0.022µF cap, if you simply move the wire to the "upper part" of the switch to the other side of the cap, where it connects to the "top" of the Gain pot.

• Place a coupling cap between the common terminal of the "lower side" of the switch and Stage 2's Grid/Grid-to-Ground resistor. This section of the switch determines whether Stage 2's Grid is feed by the Input Jack or the output of the Gain pot Voltage Divider. This part of the circuit would then look like Switch --> Cap --> 1MΩ resistor to Ground --> Stage 2 Grid.

If you make those changes, I bet the pop is very substantially reduced, or eliminated entirely.

[Albi]

Thank you B! That's a great response and I appreciate you taking the time to do this.

I was thinking the huge voltage/current changes wouldn't be making it easy.

So I'm gonna grab an iron and make the changes. Just to be certain: The plate resistor on V2 will be moved to the actual plate side of the switch and that part of the switch will now be in the OFF position and when ON it will be connecting the coupling cap of V1 to the plate of V2 for parallel mode operation. I am assuming that the plates of V1 & V2 need to be tied together for both stages to work in parallel. Do I have this correct?



Thanks!!

Albi

[Albi]

Hi - After making those changes there is no difference in the pop when switching, it's still there loud as ever. I was surprised because on paper it all makes sense.

The common terminal at the bottom section of the switch now has a cap to grid of second stage and a 1m resistor to ground. 100K plate resistor going directly to plate on stage 2. upper part of switch common terminal going to plate of stage 2 and on the other side is either OFF or connecting to stage 1 coupling cap.

Am I missing something? Any other ideas?

Thanks!

[Albi]

So when I loose that "Upper side of the switch" thats joining the plates of Stage 1 & 2 the pop is a lot better. It is hard to hear the difference but am I correct in thinking I am not hearing both stages in Parallel now and that is is in fact only the second stage when the switch is in 'parallel mode?"

[B Ingram]

• Place a coupling cap between Stage 1 and the switch: this already exists in the 0.022µF cap, if you simply move the wire to the "upper part" of the switch to the other side of the cap, where it connects to the "top" of the Gain pot.

... upper part of switch common terminal going to plate of stage 2 and on the other side is either OFF or connecting to stage 1 coupling cap.

I missed that my first recommendation would shut Stage 2's plate volts to the voltage divider. So you need:

Stage 1 Plate ---> Coupling Cap ---> "top of the Gain pot" ---> 2nd Coupling Cap 10x bigger than the first (but a 0.1µF is close enough if that what you have handy) ---> Stage 2 Plate.

In other words, we needed to block DC Volts from Satge 1's Plate, and also block DC Volts from Stage 2's Plate. And I think disconnecting that part of the switch proved it was the remaining bug when the popping settled down.

[Albi]

That makes total sense! Shame on me for not seeing that myself, of course the B+ is being fed into the voltage divider and right into the grid of stage 2. Well spotted. So does it matter if the bigger coupling cap is before or after the switch, I would imagine it doesn't matter? I could attach it directly to the plate of stage 2 and the common terminal of the switch on that side.

[Albi]

You are a genius B! Thank you so much for your help. This has almost completely dialed out the loud pop. It's much quieter now. I had a .22uf on hand. Out of interest, to make it even quieter, where would I adjust things further on the bottom of the switch that connects stage 2 grid to stage 1 grid or the voltage divider? I have installed the 1m resistor from grid to ground of stage 2 and a .1uf cap from the common point of switch to the grid.

Also how would you tackle that NFB switch? This pops too but there shouldn't be any DC on a speaker return right? Wondering what your words of wisdom are here. I've tried some basic stuff but to no avail.

[B Ingram]

... So does it matter if the bigger coupling cap is before or after the switch, I would imagine it doesn't matter? ...

You had a 0.22µF, and so I would do 0.022µF from Stage 1's Plate, and use the larger 0.22µF after.
The only reason for using the large-value cap is that 2 caps in-series lower the total capacitance (like 2 resistors in-parallel). Therefore, we set the bass roll-off of Stage 1 with the smaller (normal) cap, and use the larger cap after to avoid further rolling-off low end.

Also how would you tackle that NFB switch? This pops too but there shouldn't be any DC on a speaker return right?

The speaker won't have DC Volts, but the rest of the circuit (phase inverter?) where the 1k8Ω connects is something other than 0vDC.

Since your present diagram shows the 0.1µF cap connected regardless of switch-setting, I would just wire that cap across the 1k8Ω, so that it is always in-circuit.

Your selections then appear to be "56kΩ" or "22kΩ", with the latter being about half of the former.

• I would wire it Speaker --> 22kΩ --> 1k8Ω (paralleled with 0.1µF) --> Ground. This gives your lower-resistance setting (but see below before wiring).

• Place a 33kΩ in-series with the original 22kΩ above. When active, the total resistance is 55kΩ.

• Place a Single-Pole-Single-Throw switch (or one side of half your present switch) across the "series-33kΩ resistor". When the switch is closed you get 22kΩ, when open you get 55kΩ.

There will still be some pop here, because there's always a minor DC-volts shift. But I suspect if we simplify how many things are connected/disconnected, the popping will be reduced.

[Albi]

Thanks for explaining that about the bigger vs. smaller cap, really appreciate that!

In the NFB circuit that switch also has an OFF position in the centre so it is ON-OFF-ON. I really like the option to have an open amp with no NFB and that was why I wired the cap to ground on its own switch so when the NFB is off it is taken out of the circuit and doesn't bleed the high's to ground. Otherwise I really like the method you propose of having the resistor in series. However if the .1uf is complicating things I can loose it for good. I would like to be able to switch off NFB and also have 2 different levels of it. Do you think a switch is the best idea here or should I be looking at another method of controlling NFB. A pot perhaps? I prefer a switch if possible. Instead of the .1uf to ground I could introduce more sparkle earlier in the preamp and leave the switch purely for NFB and not a sudo presence type thing.

[B Ingram]

Switch is fine here. Just accept that if there is a switch, and especially an option that opens the feedback loop, that there will be popping to some degree.

And given that you're using On-Off-On, you could leave things as they are.

If you were inclined to tinker, you might set up a series string of "1MΩ --> 33kΩ --> 22kΩ" and then use a 3-position switch to

• Have every switch-contact open, which uses the 1MΩ series resistor for "No Feedback"

• Close the 1st switch contact to short the 1MΩ, giving 55kΩ and "Low Feedback"

• Close the 1st and 2nd switch contact to short both the 1MΩ and 33kΩ, giving 22kΩ and "High Feedback"

I think you would get there with a DPDT On-On-On switch (and Alternate Source), which would also be more intuitive by having "No Feedback" at one outside setting rather than in the middle.

[Albi]

Hi B - I can't thank you enough for your help with this problem and for taking the time out of your day to help me. Really really appreciate that and I have learnt a bunch here!

The amp switches silently and as smooth as butter from parallel mode to cascade. The NFB advice is good too. For now I just put that 'accent' cap on the lower level of feedback with the on-off-on switch. Maybe there is an off-on-on switch I could use to lay it out as you suggested.

[Albi]

One more question if you don't mind.

For the next build if I want loose the cascade option and just have a switch to use V1 as parallel or just a single stage. In other words what would you do to simply remove the second stage from the circuit? Would it be enough to shunt the cathode to ground?