A simple Relay Bypass

[JonesXavier]

so when i first fired up the circuit it was immediately apparent that my layout (above) had the bypass/active routing backwards (LED on in bypass).

revised layout (for 12V circuit, 520R/9V coil):
View attachment 57902

verified
View attachment 57901
fitted it to a CE-2 build running @ 12V. was quite a squeeze.
but it works!
it's an awesome switch. i love it.
thanks for sharing this circuit @Chuck D. Bones

Nice! Congrats! I Have 2 relays here, are OMRON G6S-2 4,5 VDC

I looked for this information in the datasheets but I didn't know how to interpret it. Do you know if I can use this Omron G6s-2 relay in this layout you posted? Or some appropriate layout? The ones I've tried so far have had a very loud pop! Thank you very much for your attention! Brazilian greetings!

 

[Chuck D. Bones]

That relay will draw 31mA coil current, which is pushing the limits of the 555's drive capability if you power the 555 from 5V. You get a little bit of margin if you power the 555 with 9V. Or you could go with the MOSFET drive option, then you have no issues driving the relay. Omron says the coil is polarity-sensitive, so make sure pin 1 goes to the +5V rail.

I do not know why you are getting a loud pop, that is not normal. Is the pop sound coming from your speaker or from the relay itself?

 

[owlexifry]

View attachment 75846

Do you know if I can use this Omron G6s-2 relay in this layout you posted?

as @Chuck D. Bones has pointed out, it appears this relay is polarity sensitive.
unfortunately it doesn’t look like this relay will work with this layout because the + supply will end up on pin 12 instead of pin 1.

 

[temol]

New layout.

 

[aquataur]

I noticed that all of those relay bypass circuits uses a shottky diode in parallel to the relay coil to tame the kickback voltage.
Somebody further up mentioned that he experienced a lag upon relay turn-off, which could well be attributed to a shottky diode.

I had recently done a circuit with a fairly big contactor, which I had fitted with the usual flyback diode (a 1N4007), and the thing just would not turn off in time (upon hitting the end stop switch). I found out this way, that this kind of protection is hardly ever used in current systems. The reason is that there is (in large contactors) a huge energy charge stored in the activation coil, which has to be destroyed in order to remove the charge. Logically, if the voltage is clamped to 0.7V, the charge can only be removed slowly.

This will apply to a small relay only to a smaller extent, but it does, and we do not know what the OP has used.

On some of my builds, like on a speaker turn-on delay and protection, I want the relay to shut off as quick as possible, so I use a (reverse) diode with a 12V Zener in series. If I monitor the mains failing, such a relay turns off before the mains switch can make a pop in the power amp. This makes a remarkable difference in turn-off speed.

Again, this may be an overkill for such a small relay, but using a shottky diode that clamps at 0.25V is not justified. Nothing is endangered with 0.7V of reverse voltage. Unless there were some other reason I fail to recognize.

BTW for the above contactor system I finally used a 50V MOV, and this shuts down the contactor lightning fast.

 

[Chuck D. Bones]

It's true that the time to unload the energy stored in an inductor, in this case the relay coil, is inversely proportional to the coil voltage. With large inductors, this is a real problem*. The relay datasheets usually don't specify coil inductance, so I measured a couple. An EA2-5NU relay coil measured 184Ω + 35mH. An AS5W-K measured 175Ω + 47mH. The current in an inductor is related to voltage & inductance by:

V = L * di/dt

Where di/dt is the current rate of change in Amps/second.

At 5V, the nominal coil current is 28mA.

Rearranging the equation,

di/dt = V/L

If we clamp the coil voltage at 0.2V with a 2N5817, when we turn the relay off, the coil current will drop at a rate of 4.25 A/s or 4.25mA/ms. The relay will drop out somewhere around 20% of the nominal coil current, in this case around 5.6mA.

It will take 5.3ms for the coil current to drop from 28mA to 5.6mA. This is comparable to the time it takes the armature to move if the current were to drop to zero instantaneously. I think this is plenty fast for a pedal.

BTW, the lag was caused by the time it took for the "memory" capacitor to recharge. We adjusted the Rs & Cs to bring this down to under 200ms.

* I once had a job working on a large physics experiment that included a superconducting magnet. The inductance was about 11H. The operating coil current was 2000 Amps. It took about 1/2 hr to ramp the current up to 2000A. The magnetic field could stop a mechanical watch and erase the magnetic strip on a credit card at a distance of a few m.

 

[Silver Blues]

Yeah those large superconducting solenoids are no joke. We have an instrument in my lab like this as well, there is enough stray field that when the magnet is charged to maximum field (7 T, I don't remember the exact current needed but probably in the 100s of amps) my (quartz) analog watch(es) stop within a few m of the magnet and you can't have any ferrous metal tools anywhere in the vicinity...

 

[aquataur]

So it looks the shottky flyback diode was chosen for convenience rather than a technical purpose.
Glad I don´t have any of those super solenoids in my pedals.

Great little 555 squircuit by the way.

 

[Chuck D. Bones]

A 1N4148 will also suffice.

 

[aquataur]

Just for the record...
I had used a similar circuit in the vicinity of a coil in the audio path (a wah). Using an OMRON G5V-2 seemed to induce a spike into the coil, and I had to resort to all sorts of trickery to bring this down to a palatable level.
I got hold of a box of NAIS (= Panasonic, Matsushita) DS2E-S DC5V relays and tried this one with a series resistor.
For some reason, they are totally quiet. They are supposedly top quality, albeit not as minute as the ones mentioned here, but then not much bigger than the G6-S2 above.

 

[Brett]

For some reason, they are totally quiet.

It's probably safe to assume that when you're switching contacts at half the current (36mA vs. 72mA) with the high-sensitivity type (denoted by the "S" in the part number) it's going to result in less noise.

Edited to say that I have had zero switching noise that gets into the audio path using cheap low-profile relays, at least so far.

 

[aquataur]

It was a wah coil that was picking up the noise, and I have an application pending using a coil.
The ones I have are termed DS2E-DC5V, lacking the "S" or "M" suffix which the current production (?) units have.
I found a reference that claims that they are equivalent to the "S" type, which I believe is true because of the lower current, as you say. I found several hundred of them brand new waiting to be thrown away in the near future, so I might as well make good use out of them.
I don't know if they are screened, but as you say, they are dead quiet, where the OMRONs weren't.

 

[Brett]

I don't know if they are screened, but as you say, they are dead quiet, where the OMRONs weren't.

I haven't used the high-sensitivity type yet, but I do have an unfinished bypass design (based on one of Chuck's designs) where a high sensitivity type (or lower current type) is called for. The reason a low current relay is needed on that design isn't because of current spikes bleeding into the audio circuit though, it's to reduce the system's overall current draw to fall within the acceptable operating range of an SMD voltage regulator.

I'd take a guess (and this is not verified) that the larger relays may provide better isolation against current spike emissions simply because the housing is larger. I'd assume that the internal mechanisms (coils/contacts) are similar between the larger relays and the miniaturized versions. Following this train of thought, it's possible that the material used to pot the relays internals is more substantial and therefore provides better insulation on the larger relay.

 

[Chuck D. Bones]

Relay coils are solenoids and throw a wide magnetic field. Many Wah inductors are pot cores, which should provide good shielding. No shield is 100% effective. Distance is the best thing to minimize magnetic pickup. Orientation can be helpful too because the magnetic field from a relay is directional. If the inductor has an E-I core (Xicon or Triad transformers) or RM core (SBP ME-6 wah inductor) then there is much less shielding. Basically, if you can see the winding, it's vulnerable to magnetic pickup. Unless it's a toroid. Toroids have the winding on the outside, but the magnetic field is highly contained by the internal core. I hand-wound a toroid inductor for my CryBaby many years ago. My brother has it now.