Output backup resistor to chassis or circuit ground?

[joulupukki]

I've got a rotary switch (from Mojotone) for selectable 4/8/16 ohm output impedance for my speaker jacks. I also want to use a backup resistor so that the amp is protected if it's run accidentally for very long without a speaker plugged in (or if the connection to the speaker is interrupted).

Right now, all components ground to the chassis at the input jack. My question is whether this backup resistor should ground to the chassis or ground to the circuit? In my mind it's a question between noise reduction (when the speaker is properly connected) and effectiveness of the resistor when there's no connection to the speaker.

Option 1: Ground to Chassis

Option 2: Ground to Circuit Ground

 

[dan40]

I would ground close to the jack myself. If you find that this causes unwanted noise, simply move it to the main grounding point.

 

[Peegoo]

@joulupukki


Hmmm

One way to make it foolproof would be to use only one speaker jack of the switching type.

When the speaker plug is removed, the switch closes and engages the dummy load across the tip and sleeve contacts of the jack.

 

[joulupukki]

@joulupukki


Hmmm

One way to make it foolproof would be to use only one speaker jack of the switching type.

When the speaker plug is removed, the switch closes and engages the dummy load across the tip and sleeve contacts of the jack.

Right. Well, the only problem with that is what happens when someone plugs a speaker cable in with no speaker actually attached? The backup resistor wouldn’t be available because it would be switched out.

 

[chas.wahl]

^^This is why I've thought about, but haven't bothered to, use such a protection resistor. The idea of having it perennially connected, which is the other approach (470 Ω in parallel with whatever speaker load there is), doesn't appeal, like having one of those parasitic bleeder resistors on the 1st filter cap(s).

While I'm not very good at following sage advice, I do remember having my ears pinned back once by @muchxs, whose wry advice was something like "the component you're not going to have trouble with is the one you leave out."

 

[tubedude]

You should not get any noise either way. I'd attach it at the jack.

 

[joulupukki]

^^This is why I've thought about, but haven't bothered to, use such a protection resistor. The idea of having it perennially connected, which is the other approach (470 Ω in parallel with whatever speaker load there is), doesn't appeal, like having one of those parasitic bleeder resistors on the 1st filter cap(s).

While I'm not very good at following sage advice, I do remember having my ears pinned back once by @muchxs, whose wry advice was something like "the component you're not going to have trouble with is the one you leave out."

Yeah, this definitely has me second-guessing. Yeah, 470R in parallel with an 8 ohm speaker would be abt 7.8R. Nice advice. I'll make the decision later and probably test out a couple different ways to see if I actually notice any difference in output volume.

 

[2L man]

Option 2 is better.

 

[GPoint]

@joulupukki


Hmmm

One way to make it foolproof would be to use only one speaker jack of the switching type.

When the speaker plug is removed, the switch closes and engages the dummy load across the tip and sleeve contacts of the jack.

Very good idea (in case of sclerosis - it is some 90% !!! Except if the problem occurs inside the connector, loudspeaker, or speaker wire...

 

[Peegoo]

Very good idea (in case of sclerosis - it is some 90% !!! Except if the problem occurs inside the connector, loudspeaker, or speaker wire...

That is true. But consider how many amp makers install three separate speaker jacks (4, 8, & 16 Ohms): it is up to the user to properly configure the amp when plugging in a speaker. Screw it up and you screw up your amp.

It's like having your fingers touching the prongs on an AC mains plug when you connect it to the receptacle. Some things are just up to the user to get right.

There's a point at which the "fool-proof-ness" ends because the engineering to accomplish it becomes complicated and a is liability unto itself.

You know what they say: "make something foolproof, and along comes a better fool."