Audio worthy 3pdt latching relays?

[tokyoburns]

Hey if you've been following the posts Ive been writing lately. First of all let me say thanks for all the great advice. This forum has been extremely valuable. Also you may know now that I am trying to design a pretty big project. If you read my robotic potentiometer post you might have figured what comes next. AUTOMATIC TRUE BYPASS! After reading up on PICs for the stepper motors in my pots which I feel I have a good basis for my design now I am moving on to the true bypass bank which will operate on the same principal of being a non-digital 3pdt that can be controlled remotely. So I figure I would use a latching relay controlled by a PIC. 1 pole for audio. one for an LED and one for voltage feedback to the PIC so it make its decisions about what to do. should be a lot simpler than the potentiometers. But I am wondering if there are any latching relays that are truly audio worthy. Obviously chatter is a huge issue. There won't be any point in having a true bypass system if the relay lowers my fidelity or colors my sound in any way and especially if there is any chatter on the relay. Does anybody know of any specific parts that would be worth trying?

Thanks for input!

[kurtlives]

Does it have to be 3PDT?

You can use a DPDT relay for true bypass and have the LED indicator across the relay's coils. I have used this many times and it works great!

[Hides-His-Eyes]

If it can flip a relay it can light an LED!

[tokyoburns]

well id prefer to use a latching relay. I think it might make it easier to use with a PIC also I'm thinking that if I can keep my audio signal away from an energized coil Id feel better. So if I use a latching relay I wont have a constant voltage across the coil and I couldnt light an LED that way. If I wasnt using a latching relay I would consider it. it probably makes finding parts easier. what kind of relay do you use? whats your set up like id be interested to see it.

Thanks for the suggestion.

[slacker]

You could just use another output from the PIC to drive the LED.

[Hides-His-Eyes]

The LED isn't in the audio path... It'd be purely part of the control system. The extra pole is superfluous.

[defaced]

DC to power the coils doesn't really affect audio because its low current and doesn't couple well. The Omron G5V series are very commonly used for signal switching.  I've never noticed a noise issue with them and I use them in high gain tube amps.

[composition4]

Hi,

I have about 50 4PDT latching relays designed specifically for audio - NAIS DS4E-SL2-DC5V

Just size a suitable current limiting resistor and there you go.  Really good for audio, no big clicks or anything either

They're normally not cheap (> $10 USD) but I got about 50 of them for pretty cheap on surplus.  If you're interested then let me know and we can work something out, keep in mind I'm in Australia though so I would have to ship them

Jonathan

[tokyoburns]

@hides his eyes

I dont know why you thought I thought Id put in LED in my audio path. In a latching relay the relay does not need a constant voltage to the coil so I cant light an LED across the coil.

@slacker

yeah I could do that but it would just add to the programming. Id rather just use an extra pole.

@composition
I would definitely be interested. that is exactly the product Id be looking for. but it might be a while before I actually acquire some. Shoot me an Email with some ordering information.

Thanks for the reply guys!

[R.G.]

Latching relays trade complexity for lower coil power. To make a latching relay work and also get the current savings, you have to turn on the correct one of two coils (or coil contacts) and then turn it back off after the right amount of time passes. Not a huge deal if you're already programming a controller, but it does mean that instead of turning on an output, you have to turn it on and off, and then when you flip the relay to the other state, you have to turn a different output on and off again, instead of just having the relay be on when one of the pins is high (or low).

Clicking in the audio path from relays comes from two effects. When you turn a coil off, the inductance causes a voltage spike. That spike can be quite large, and very fast, so it couples well through the inter-element capacitances in the relay itself. Some signal relays have internal shields for this very purpose. Also, the change in magnetic field can couple through to the audio path.

I've used latching relays, but I prefer to not deal with the complexities of two coils or two directions for actuation even when I use a uC. I reserve them for cases where saving the power of constantly having a relay coil active is worth the extra complexity. It's all a design tradeoff.

[tokyoburns]

I guess part of the reason is that I wont have a default command on start up for my uC. So if I unplug my pedal and plug it back in I will lose whatever my last setting was. I would prefer my bypass to remain in whatever position it was last left like my robotic pots would be. so if I unplug my pedal board and plug it back in then I'll most likely lose my setting. That is a big part of using a latching relay. Also From the guts of latching relays that ive seen (and im not an expert) the inductive coil is separated by more space from the switch path than non-latching relays that I have seen. So I was hoping this would help preserve my audio from the spike that would occur with the rising and falling current in the off and on of the coil. I suppose that would depend on that particular relay. which is why I posted this thread looking for just that very thing.

So those were the things I was thinking when I decided on latching relays.

Thanks again guys I appreciate it.

Also if you have suggestions for relays audio samples or links to them would be a huge help in my decision making. thanks!

[Hides-His-Eyes]

@hides his eyes

I dont know why you thought I thought Id put in LED in my audio path. In a latching relay the relay does not need a constant voltage to the coil so I cant light an LED across the coil.

Misinterpreted your reply. Why not use a bistable flip flop to control the LED? Simple circuit, trivial with CMOS 4000 ICs available.

EDIT: and now I've read your first post for the third time I finally get it. Sorry. If your PIC doesn't have any spare outputs, you could still use the flip flop, or even build a discrete one if you're feeling cheap: http://www.falstad.com/circuit/e-multivib-bi.html

[tokyoburns]

@hides his eyes

I guess there are plenty of options for lighting an LED. To me it seems like the easiest one is to just give one its own pole. I appreciate the suggestions. Do you have any good reason why a flip flop would be better than that? Perhaps Im missing something.

[R.G.]

I guess part of the reason is that I wont have a default command on start up for my uC.

Why not? Almost all the uCs and certainly all the PICs have electrically erasable programmable read only memory (EEPROM) in them. This is a small section of nonvolatile memory that you can write on the fly. One thing you could do is to write the settings into EEPROM as you go along, then go load them from EEPROM at startup every time. The uC always comes back up just as you left it.