Cheating Relays?

[KarenColumbo]

On a side note: I should have marked this thread as "Off Topic", since it touches the main theme, stompboxes, only in a hypothetical way. For this I apologize!

[FiveseveN]

Here's a different approach: mute the signal that feeds the poweramp when you power off. You can even do it with another lovely relay! 

[KarenColumbo]

Here's a different approach: mute the signal that feeds the poweramp when you power off. You can even do it with another lovely relay! 

Heh. I'm in. But there must be no logic circuitry involved.
In my head this looks a bit complicated. A time-cascade when switching off.

Main power off = Mute audio -> delay for relay (1 second) -> cut power. Hm.

I really don't want to involve ATTINY & Co (5 V required -> yet another power supply)

[KarenColumbo]

This looks promising: https://www.analog.com/en/technical-articles/flexible-power-supply-sequencing-monitoring.html#
Meh. 5 V supply needed, 5 V trigger output.

[FiveseveN]

But there must be no logic circuitry involved.

Unless you absolutely insist on using a non-latching relay, it's dead simple:



You don't even need the transistor driver, just energize the coil from a (stepped-down) poweramp rail. See other options in Rod's article on muting.

[KarenColumbo]

Thanks for reminding me of this one! Problem(?) is: I'd like to switch it all off with a single "Mains" switch. I'd like this sequence at mains power off: Audio Mute -> Pause -> Power off, so that remaining "loads" don't release themselves into the speaker(s) because some effing relay thinks it has to be faster than the power amp.

[duck_arse]

I don't think I'm paying enough attention to this thread/problem, cause it strikes me that you would just reverse the states. have the relay-unpowered-state as was your relay-powered-state [language, don't fail me now!], so when your power supply collapses at off-switch, the relay is already unpowered and in your preffered channel X state.

clear? not to me, either.

[amptramp]

The effect of D2, the zener across the coil is to collapse the current more quickly than if it wasn't there.  You have a certain coil current.  After the current supply is shut off, the current will collapse at a rate determined by the inductance and voltage.  The current flows through the flywheel diode and the zener diode, so the voltage will be large.  Replacing the zener diode with a short circuit will result in the voltage being smaller, so the reduction in current will be slower and the relay will remain on longer.  The formula is

V = L(di/dt)

If the V around the coil/diode/zener loop is reduced, the di/dt is reduced and the current drops more slowly.

Remove the zener and just leave the diode there and see if that works.  If not, a capacitor from the base of the transistor to ground will hold the transistor on and prevent the current from switching off rapidly.

[KarenColumbo]

I don't think I'm paying enough attention to this thread/problem, cause it strikes me that you would just reverse the states. have the relay-unpowered-state as was your relay-powered-state [language, don't fail me now!], so when your power supply collapses at off-switch, the relay is already unpowered and in your preffered channel X state.

clear? not to me, either.

Yeah, it's (quite) clear what you're suggesting. But the two states are - from the user's point of view - of equal importance. So I'd rather avoid to set all switches to a "safe" position before I power off the amp. I'd like to leave'em where they are, and that could be a position in which an almighty "BUMP" at switch-off is possible.

[KarenColumbo]

The effect of D2, the zener across the coil is to collapse the current more quickly than if it wasn't there.  You have a certain coil current.  After the current supply is shut off, the current will collapse at a rate determined by the inductance and voltage.  The current flows through the flywheel diode and the zener diode, so the voltage will be large.  Replacing the zener diode with a short circuit will result in the voltage being smaller, so the reduction in current will be slower and the relay will remain on longer.  The formula is

V = L(di/dt)

If the V around the coil/diode/zener loop is reduced, the di/dt is reduced and the current drops more slowly.

Remove the zener and just leave the diode there and see if that works.  If not, a capacitor from the base of the transistor to ground will hold the transistor on and prevent the current from switching off rapidly.

Thanks mate! I already tried with the diode - to no avail, I'm afraid. But I will try the one with the cap to ground! Haven't thought of this one yet

[KarenColumbo]

Just tried it - it sure delays the relay (tried it with 10 uF and 100 nF), but there's still a spike at power shut-off.
I made a short video - at the very end you can see the red LED (which is connected to the normally-closed position) flashing AFTER I shut down the power supply.
http://www.mrscolumbo.com/uploads/relay.mp4

[KarenColumbo]

So in the long last it's probably not really a "Cheating Relays" question ... seems to me it's about the general power supply chain in the amp as a whole.

[KarenColumbo]

Very nice link to Mr. Elliott's muting article, by the way, thx! (https://sound-au.com/articles/muting.html). The "Delayed Start & Mute" circuit (section #13 in the article) looks a bit like what I'm searching for. I will try this at home.
In reality I think I should put this right before the phase splitter, I guess.
And this: https://sound-au.com/project104.htm

[iainpunk]

what kind of power switch are you using? if you can get one with an extra throw, you could just disconnect the speaker out from the power amp when you switch it off.

this will work given its not a trafo/inductor loaded power amp or a bridge configuration.

cheers, Iain

[KarenColumbo]

what kind of power switch are you using? if you can get one with an extra throw, you could just disconnect the speaker out from the power amp when you switch it off.

this will work given its not a trafo/inductor loaded power amp or a bridge configuration.

cheers, Iain

Ah! That is new! Hm, I guess I could look for such a switch - won't there be mains noise coupled into the speaker line in this way? Or did you mean I just cut the power to a relay that switches the speaker in/out?

[iainpunk]

i don't think the capacitive coupling in a switch in anywhere near big enough to drive a low impedance like a speaker, so don't worry.

cheers, Iain

edit:
ill demonstrate this:
if the capacitance is unusually high (100x what im used to measuring in most switches) well use 1uF
if you use a 16 ohm speaker, the cutoff frequency is about 10kHz, that means that the 50Hz from mains will be attenuated by 46dB. and that is with a 100 times larger switch capacitance. if we take a more realistic 10n we attenuate the 50hz mains with 86dB

[KarenColumbo]

Just to get this right: The mains switch only has to switch the live wire (as shown right at the top of http://www.valvewizard.co.uk/standby.html) since i plan of using a detachable cable ("Kaltgerätekabel"). If that's correct I won because all the power switches I bought are DPST, so there's set of poles free for a very physical mute
It occured to me my questions must read like some kinda moron asks them. But those little details are something my (quite creative but also quite chaotic) brain always puts aside with a "ah, we're solving this on the go as soon as it's acute" and instantly forgets. Like the pinout of a LED - so I really want to go the hard way (most hard on your patience, I admit) to finally cram some "systematicality" into myself. Especially as in a tube amp there's a bit more than 9V wallbugs involved. So: sorry for my 9-year-old-ness. I'll eventually come around.

[DIY Bass]

I have always loved the power supply in my bass amp.  There is the usual toroidal tranny and really big caps.  I am not sure what else is there, but I have had someone kick out the power from the wall and then plug it back in without losing sound.  When it turns off it waits for a few seconds and then a tone that starts high and gradually gets lower in frequency.  No pop.  I have no idea how they did it but it's pretty cool.  There must be some type of RC network I guess.

[KarenColumbo]

Sounds like HUGE reservoir caps - almost sorta backup battery pack

[amptramp]

I have always loved the power supply in my bass amp.  There is the usual toroidal tranny and really big caps.  I am not sure what else is there, but I have had someone kick out the power from the wall and then plug it back in without losing sound.  When it turns off it waits for a few seconds and then a tone that starts high and gradually gets lower in frequency.  No pop.  I have no idea how they did it but it's pretty cool.  There must be some type of RC network I guess.

I'm not sure they did that deliberately as it might be cool for you but not for every application.  This is probably the reaction of power regulator circuitry responding to the loss of input while still trying to keep the output caps charged.  Most regulators need protection against turnoff since at that point, the input voltage will be lower than the output voltage and there should be a diode to discharge the output caps and prevent the output from remaining at a higher voltage than the input, which would kill the series pass transistor in the regulator.

When the input power drops out on a switching regulator, you don't get the same behaviour as you would from a linear regulator.  A linear regulator simply drops out of regulation and the voltage declines.  A switching regulator will attempt to maintain output voltage by drawing more current off the input caps and you will notice something here: if the current increases when the voltage decreases, you have a negative resistance and it can be made to oscillate with timing set by the input L-C filters but the conditions for oscillation are there.