GeoFex Relay Driver - Popping When Energising

[stalefish]

Hi everyone,

So I've been putting together my own variant of RG's Programmable Effects Switcher. So far, I've got the foot-switch going (mom switch into CD40106 inverter into a 74HC573 to generate the +on signal, as per GeoFex's excellent article), no worries. I'm running on +5V thanks to the 74HC573, by the way.

However, I'm stuck on the relay and driver bit. I used the circuit below from GeoFex to drive a DPCO 12V relay using +5V out of the 74HC573 as an "+on/off" signal but I'm getting popping and clicking on energising. De-energising is more or less quiet. I messed around with C1 and C2 from the circuit, slowing the on/off to the point that I could actually see the slow switching on the indicator LED and hear a distinct signal cut off as the switch breaks, then makes, but I still get a pop when the relay energises.

Interestingly, I get the best results (almost no pop on de-energise, but popping on energise) when I remove C2. I also added a rather massive cap (100uF) parallel to diode D2, which seemed to help a little. Still significant popping on energise, though.

Any suggestions? I'm wondering if I should order up some new relays (admittedly, I went with the "lowest bidder" approach  ).

Thanks!!

-Shaun

[R.G.]

Are the relay/coil/transistor grounds connected to signal ground?

[stalefish]

Yeah, they are. From the sound of your post, I take it that wasn't correct? 

Thanks!

[R.G.]

Correct versus not correct gets kind of foggy when high gain effects are involved.

Try isolating the ground to the two effects jacks from the "ground" to the relay switching section with a resistor. Try about 100 ohms, and then 1K.

The fact that you get different pops when energizing and de-energizing means that there are two different popping mechanisms; it appears that what is making one better makes the other worse. (At this point, you're thinking "well, duuuh..."     )

But that means that there may be several answers that may have to be combined. Interference always happens one of two ways, conducted or radiated. Conducted can be the sneaky way of moving the ground around, or it can be the sneakier way of the through-air capacitance between metal parts. The slow-down stuff attacks the through-air path by slowing down the edge of the voltage on the coil. However, putting a BFC across the coil can make ground transients worse because they let a big pulse of current run through the power and ground wires, bigger than the coil would have done on its own. This makes coil-capacitance popping better, but conducted noise pickup worse.

[stalefish]

Ooh, bugger. But that's the story of life, isn't it? Haha! I used to work on military aircraft and I remember the manual once saying "If you see symptom 'A' do this, if you see 'B', do that. If you see both, it's a complex situation that requires troubleshooting." and that was it. Naturally, we had A & B. 

Anyway, I tried 100 ohms and 1k isolating the grounds, but I'm still popping on energise. Was this intended to stop or reduce the current pulse? The way I'm reading your reply, I'm guessing that it might end up being a balance of cap values to get the mid-point between conductive and radiated popping? Also, would I be able to use an electrolytic cap in C2? I tried one or two and I got some weird results. The indicator LED ended up being half lit when the "+on/off" was low.

What's "BFC" stand for, by the way? (I'm thinking "Big F'ing Cap", but hey, that's me..)

Thanks again! Really, really appreciate the help!

[PRR]

> What's "BFC" stand for, by the way?

Technical term not found in military manuals.

100uFd across a small relay is a relatively BFC.

[stalefish]

PRR: So I was right. Haha!

So I went and plug a couple of effects into the thing, just to see what would happen and the popping lessened considerably. If I'm playing and I switch, there's no discernible noise to speak of. When not playing, the popping was also minimal. Even with a fairly high gain pedal with the gain and volumes cranked up all the way (an old SansAmp TRI-OD on the California setting), the pop was less than what I would get on, let's say, a true-bypass distortion pedal without a pull-down resistor. Any idea why? (Other than a physical perception thing.)

I did add 3.3M pull-down resistors on my BFC and from C2's base connection to ground. Don't know if that helped at all.

(By the way, I was wrong about the 100uF across the relay. It was 1000uF...    It's now back to 100uF, which seems to work best.)

Interesting discovery: If you use 1000uF as C1, it takes almost a second for the relay to switch. Don't know why, but I found that amusing. Too much solder smoke inhalation, I think..

Thanks again!

[PRR]

Are you _sure_ there is no stray DC on the signal contacts? (Switched DC on audio is click/pop every time.)

Disregarding R.G.'s next sage debugging advice; I'm thinking further on his power supply transients. Can you temporarily remove the +12V feed from the relay and connect a fresh 9V battery as so?



(Yes, most "12V" relays will trip with a fresh 9V battery. Not as a reliable long-term thing, just to see if localizing the relay current mutes the pop.)



> 1000uF as C1, it takes almost a second for the relay to switch. Don't know why

What does a cap do? Store energy. What does a resistor do? Impede the flow of energy. So your CMOS lifts the left end of R1, energy flows right, but C1 sucks it up. Much like filling a barrel through a straw. Bigger barrel, slower filling.

There is a formula R*C. In this case R is say 4,700 and C is 1,000uFd, which is 0.001Fd. 4700*0.001= 4.7 seconds. It takes 4.7 seconds for C1 to "fill" to 63% of the voltage from the CMOS. (The relay clacks sooner because C1 does not actually rise to full CMOS voltage and Q1 will turn-on with much less than full CMOS voltage at C1.)

But slow-switching isn't the whole answer. The relay pull may change slowly, but the relay contacts "snap". Imagine it is dusk. Sunlight getting dim. You stand by the light-switch to turn on the lights when you think it is dark enough. 5pm, still light. 5:15pm, tension mounts. 5:30, finger starts to lean on the switch. 5:37, lever moves SNAP, and the lights come-on instantly.

> If I'm playing and I switch, there's no discernible noise to speak of. When not playing, the popping was also minimal.

Every switch clicks. Truly ZERO-click/pop switching is truly hard. Mix-down consoles which may switch during program and the recording will be heard a million times go to great lengths. In performance, pops that don't make the bartender reach for his gun are often perfectly acceptable to all. (And I can cite a recording with a reverb-switch which does sound like an air-gun.)

[stalefish]

PRR: Tried the battery, but it didn't make a discernible difference. I'm assuming that the +5V supply is still hooked up? All I did was remove the 12V connection to the relay as per the diagram.

Every switch clicks. Truly ZERO-click/pop switching is truly hard. Mix-down consoles which may switch during program and the recording will be heard a million times go to great lengths. In performance, pops that don't make the bartender reach for his gun are often perfectly acceptable to all. (And I can cite a recording with a reverb-switch which does sound like an air-gun.)

Yeah, I'm going to A/B the pops with my current loop pedal's switching noise. I'd hate to be wasting everyone's time here. Of course, if the switching noise is the same, then yay for it working, but then wasting peoples' time and......

Thanks!

[stalefish]

Ok, I tested the switch out with a few of my pedals and I seem to be getting pop on some pedals, but not on others. This might seem a little obvious, but those that have some sort of signal amplification (ie booster, overdrive) show more noise than those without. Switching for my compressor, for example, is pretty much silent.

Again, I know this may sound a bit obvious, but I'm wondering if there's something I don't know. Is there anything further I could do?

Thanks!