relay Snubber Questions

[trixdropd]

I am building a circuit simiilar to this, to make true bypass via relays.

I've ready "A remote indicating bypass" on www.geofex.com about a million times.  I've never been able to get the "snubbing" portion right. So i breadboarded what's on the circuit in the pic i provided and it works. (Does it pop I don't know)
What i wanna know, is that if the two wires that form the "switch" have a common for me to pigtail several of these circuits together? For instance, i wanna make a 6 loop rackmount looper with each channel sharing this circuit (or close to it),
and I wanna use an 8 conductor ethernet wire to attack the rack to the remote footboard that houses the switches.
I have been making relay based true bypass boxes for a couple years now, but now I wanna get the snubber section proper so My pedals switch silently..

Anyone?

BTW:
The tranny a 3904, the resisters are 4.7k, cap 0.1 ect.

[slotbot]

im not sure if i understand you correctly but... you want to make an 8 conductor wire control 6 of these circuits?

yes you can do it

the "common" wire would be the one going to + power "SW1".

sorry if i misread you and thats not what you were asking.

[trixdropd]

im not sure if i understand you correctly but... you want to make an 8 conductor wire control 6 of these circuits?

yes you can do it

the "common" wire would be the one going to + power "SW1".

sorry if i misread you and thats not what you were asking.

That is what I'm asking. I guess another snag i wanna get around is having the + be the common between the seperate loops. Though that would be fine in the rack example I mentioned, I am gonna build other units that need the - to be common as to allow a standard t button footswitch with a common sleeve to switch one of these circuits and perhaps change an amp function too, which are usually-. Does this require an npn tranny (assuming this one is pnp) to accomplish?

Also, thanks for replying.

[slotbot]

Hi

i think you could do it with NPN and a ground reference.

add a resistor, say maybe 1k, from SW1 to SW2 (collector of NPN to base).

then connect your switch, 1 lug at SW2 and 1 at power - (ground).

connect the switch one lug to power - adn one to the base of the transistor. 

basically what will happen is when the switch is open the relay will be on. when the switch is closed, the base will be grounded and the transistor will turn off, which means the relay will turn off.

try it and see. i thinkit will work though.

[trixdropd]

Hi

i think you could do it with NPN and a ground reference.

add a resistor, say maybe 1k, from SW1 to SW2 (collector of NPN to base).

then connect your switch, 1 lug at SW2 and 1 at power - (ground).

basically what will happen is when the switch is open the relay will be on. when the switch is closed, the base will be grounded and the transistor will turn off, which means the relay will turn off.

try it and see. i thinkit will work though.

Nice logic there. I'm new to this end of the spectrum but that really makes sense. I'll breadboard it up and see what happens. I really appreciate you taking the time to help me out.

Rock on!!

Jeremy

[trixdropd]

This indeed works flawlessly. I haven't hooked up the audio lined to listen to see if the pops are eliminated but I'm onto that next.
Thanks again!!

Jeremy

[slotbot]

alright cool.

good luck.

[trixdropd]

Well, unfortunately for me, I found out too late that i had my wire going to ground on the wrong end of the resister... On both sides it bypasses the relay, but on the one side it also short circuits the power supply, resulting in me frying my trusty Ibanez wallwart I've been using since 1998. Luckily I bought a second recently, but nonetheless a bummer...    The circuit works it seems though, so i'm happy for that!!

[slotbot]

oh whoops you are right. the control should not not go to sw2 but the base of the transistor.

sorry about that again. 

[trixdropd]

oh whoops you are right. the control should not not go to sw2 but the base of the transistor.

sorry about that again. 

Well, it's truly a small price to pay for the knowledge I gained. This has opened up my understanding a bit more which is a great thing.



Thanks! 

[trixdropd]

Well I breadboarded this and the pop is ever present, and this is just using jumpers for the loops rather than pedals, so it's mechanical click. Any ideas?

[slotbot]

what are you using for a power supply?

also i would suggest disconnecting the signal ground (from the 'S' inputs) from the power - of the relay circuit and ultimately running the relay power off an isolated power supply from the pedals or whatever is in the signal chain.

[trixdropd]

what are you using for a power supply?

also i would suggest disconnecting the signal ground (from the 'S' inputs) from the power - of the relay circuit and ultimately running the relay power off an isolated power supply from the pedals or whatever is in the signal chain.

I will look into keeping the power ground and signal ground isolated. I remember RG mentioning that they could be connected with a 100ohm resistor or left open, whichever is more quiet.
I'm using an Ibanez ac109, and it's always been a great power supply for me, stable 9v all day long. never had noise issues either after many builds.

[slotbot]

hmm. well hopefully someone comes along who has actually done this and might have some better insight....

I would suggest trying to disconnect the signal and power grounds. do an experiment just with the send and return jumpered like you were doing.

[trixdropd]

hmm. well hopefully someone comes along who has actually done this and might have some better insight....

I would suggest trying to disconnect the signal and power grounds. do an experiment just with the send and return jumpered like you were doing.

So far it seems to be the exact issue with the grounds isolated. I'll have to mess around a bit more. The circuit as is is reported to work, and i even have the same relay as the designer. They originally called for a 470r resistorfrom sw to tranny, but i put a 4.7k instead and it worked. i then realized all my 4.7r are in fact 4.7k. maybe this is my issue?

[trixdropd]

hmm. well hopefully someone comes along who has actually done this and might have some better insight....

I would suggest trying to disconnect the signal and power grounds. do an experiment just with the send and return jumpered like you were doing.

So far it seems to be the exact issue with the grounds isolated. I'll have to mess around a bit more. The circuit as is is reported to work, and i even have the same relay as the designer. They originally called for a 470r resistorfrom sw to tranny, but i put a 4.7k instead and it worked. i then realized all my 4.7r are in fact 4.7k. maybe this is my issue?

Alright,
I used the circuit snippet with the relay and snubber on rg's Geofex site for the cmos switcher. very simiilar to this circuit. I get the exact same popping regardless of how logic and audio ground are tied together, as well as wether or not I  create a ground reference. The circuit physically functions fine, but the pops are still there just as bad as without the r/c network

[trixdropd]

Bump,  anybody else wanna chime in?

[R.G.]

There is a capacitance between the relay coil and the signal path. This varies with the mechanical construction of the relay. When the relay coil is switched, the voltage on it necessarily does some very quick movements. You can't totally suppress these voltage changes. I have seen relays in guitar amps with 1uf, as big as 22uF capacitors across the relay coil.

The capacitance to the audio path plus the impedance of the audio path form a high pass filter. The higher the impedance of the audio path, the lower the frequencies this filter lets through.

Slowing down the coil voltage transition lowers the frequencies present in the voltage changes on the coil. It helps and may be enough for some relays.

However, the higher the impedance of the signal path, the worse things are.

There are some relays which shield the contacts from the coil with a grounded shield. This helps a lot, and only degrades the audio with the capacitance to ground of the signal to the shield.

Suppressing relay pop will always be a combination of finding the right relay, slowing down the coil, and shielding the signal path, in some combination.

[slotbot]

could you use an RC series circuit in parallel with the coil instead of just a capacitor? this would slow things down quite a bit no?

if you used the right values maybe you could even make a tremolo

[trixdropd]

There is a capacitance between the relay coil and the signal path. This varies with the mechanical construction of the relay. When the relay coil is switched, the voltage on it necessarily does some very quick movements. You can't totally suppress these voltage changes. I have seen relays in guitar amps with 1uf, as big as 22uF capacitors across the relay coil.

The capacitance to the audio path plus the impedance of the audio path form a high pass filter. The higher the impedance of the audio path, the lower the frequencies this filter lets through.

Slowing down the coil voltage transition lowers the frequencies present in the voltage changes on the coil. It helps and may be enough for some relays.

However, the higher the impedance of the signal path, the worse things are.

There are some relays which shield the contacts from the coil with a grounded shield. This helps a lot, and only degrades the audio with the capacitance to ground of the signal to the shield.

Suppressing relay pop will always be a combination of finding the right relay, slowing down the coil, and shielding the signal path, in some combination.

I messed with the circuit some more. I made the circuit in the schematic in the "A remote indicating Bypass" article. I measure with a meter that indeed the pop is supressed quite a bit. I put a 220uf cap across the coil.
I don't have a 47uf available to test. My relays are hi sensitivity. Are there any values i should experiment with resister wise?

Also, in your circuit RG, does establishing a ground reference and using ground as my common going to be an issue at all?