Relay clicking kinda solved - Just one more question!

[simon111]

I have just finsihed a remote switch unit using RG's schematic.


But when I switch the switch it produces a big thump.

With the amp down low it's fine/acceptable. But up at a gigging level it is very audible. It seems to click worse when going from the NC to NO than going from the NO to NC position (i.e when a voltage is applied to the coil)

Has anyone any useful tips to stop it popping and thumping.

By the way I am using Maplin Relays. Bestar BT12S - data sheet link http://www.maplin.co.uk/Media/PDFs/N17AW%20&%20N18AW.pdf

Is it the relays that are not very good ?


Cheers
Simon111

[R.G.]

But when I switch the switch it produces a big thump.

Big thumps are the sign that there's a big DC level change when something switches. Since the design is set up so the DC is not attached to the signal lines at all, that suggests one of two possibilities:

(a) there is a soldering or wiring error that lets the switched DC get onto the signal pins where it's not supposed to be; for this one, (re) check your relay pinout, wiring, and soldering
(b) there may be a flaw in one of the things connected to the relay setup; use your DMM to check for DC voltages on the relay pins that carry signal (not the coil) both before and after switching.

Let us know what you find.

[simon111]

I have checked for any DC voltages are there don't appear to be any on the signal side.
Below is the full schematic of my switch. Perhaps it's something I have designed incorrectly.

Note..... This is set-up for mono input , stereo output switching.
And I have since connected all the switching ground points together (I kind of went over the top!). So only now the 'hot' signal is switched.



Hope this helps R.G. As it was something that I had asked for a few weeks back an no one had an answer for a mono input stereo output switcher. So this is really my own attempt at doing soemthing (actually my first ever design!) And it would be great to get it to work correctly!!!

[simon111]

Erm....I think there is definitely a fault.

All the grounds are connected together on the signal path as said above.

When I connect this to the Gnd/Ov on the relay circuit it's fine when in the NC positioned (i.e coil not energized) but when I flick the switch to effect B channel. The lights go off and there's a big hum.

Now I am confused!

[Dirk_Hendrik]

Why are you switching the jack's groundconnections?

[simon111]

The reason why I am switching the ground connections is because they are going to two separate amps.  And the effects are stereo output and mains powered (rack units). Therefore by switching the ground and not connecting them together will mean no earth loops.....fully isolated...correct?

I have rewired it up as shown in the diagram now. And it still pops. I have added a 4.7uF capacitor between the input +switch and ground. This does delay the switch over and 'might reduce the pop' one way, but when releasing the switch the capacitor doesn't slow it down and it still pops loudly.
?

[simon111]

OK. folks just done a bit of messing with it.

I have taken out all the relays.

Guess what?

Still clicks!

It clicks when I press the footswitch. Even though none of the signals are now connected. (?) I have literally taken the relays off the pcb and left the wires that then go to the jack sockets in place.

I have also taken out the LED indicators and still clicks.

So it seems like it is stemming from the footswitch.

I think we are closing in!

Any ideas?

[simon111]

Update...............

Soldered the relays back in.

Exactly the same (as you would expect)

So it's something to do with the remote switching bit.

The switch is just a DPDT latching footswitch. Using only one side. It connects the switch + point between 0v and +9v. This somehow seems to be getting into the signal lines.
I'll keep trying different things.


Simon111

[free electron]

So, you have 3 wires from the switching unit to the footswitch?
1. +V
2. Switch+
3. GND

Maybe try to do it, as it is usually done, e.g. in the switching circuits inside the amps. Pull up the Switch+ to +V via 10k in the switching unit and then use just only two wires (Switch+ and GND) to controll it with footswitch.
There is another way to slow down the relay control voltage
http://www.mif.pg.gda.pl/homepages/tom/misc.htm#Pops
just check out the schems.

[simon111]

Cheers for that, I'll give it a try tomorrow.

Yes I was switching between +v and 0v with the switch with the +switch connection as the common pole.

I.e.
              +9v] Switch Position 1
                                               >   +Switch connection
                0v] Switch Position 2

[simon111]

I am about to give up on this idea!

I can't fatham it out.

It clicks when there are not relays connected.
It seems like the switch is somehow sending spikes/intereferring with the signal path. I have tried running it off batteries and it's still the same as running it off it's PSU.

I have connected the switch up so it's got a pull up resistor, then switch the ground. Still clicks. Tried putting the relays back in and using the idea outlined above still clicks.

Whenever I change the voltage applied to the +switch point, it sends a thump/pop.

Back to the drawing board!

I guess designing your own circuit is harder than it looks!

[R.G.]

Any time there is a change a voltage anywhere, that change is radiated to the entire universe. The degree to which this can be picked up depends on (a) the power of the original change (b) the speed of the original change (c) how well coupled the receiver is to the sending voltage.

Very high speed changes radiate away into space quite easily. The lower the speed of the change, the less energy is sent out into space. DC does not propagate into space.

Coupling is quite important. One can couple by antennas, and the size of the signal received will depend on how good the anntennas are. One can couple by capacitance, and the size of the received signal will depend on the amount of capcitance - i.e. the amount of surface area and distance between the sender and receive. One can couple by inductance, and the size of the received signal will depend on the amount of mutual magnetic inductance, and hence on conductor loop area and coupling materials. One can couple by conduction, and the size of the received signal will depend on how conductive the path is.

Why am I blathering about this?

It's becase there is a coupling method you don't yet see.

Relays are used to isolate the coil driving circuit from the contacts conductively - there is no conduction path between the coil and contacts usually. They can still couple clicks and thumps by capacitance, inductance and RF emission.

The frequencies here are too low for RF, we'll ignore that one. There exist RF relays that are designed for switching relay signals where you cannot do that, by the way.

If you pull the relay out of the board and you still get noise, then it's not the relay, and we can discount relay capacitance, inductive coupling and so on in the relay.

What's left?

Your switching circuit is coupled to your signal line somehow. Could be conductive, magnetic, or conductive.

Let's look at conductive. Is your switching ground connected to your signal ground? If you switch a lot of current in a ground wire, the resistance of the wire makes a sudden level change and it can be heard in the audio path. Try isolating signal ground from your switching ground if it's not. Does your DC power for your relays touch the signal path or ground, perhaps unintentionally? Use an ohmmeter to check.

How about capacitive? Are your switch wires tightly twisted so that any voltage changes happen in the space between the twisted wires? Is your audio path shielded? Are the audio wires physically separated from the switch wires by inches of space?

How about inductive? Do your switching wires and audio wires form overlapping loops?

Finally, distinguish between clicks, pops and thumps. A click is a high frequency thing. Pops are more mid frequency, and thumps are low frequency. Clicks are caused by sudden, transient changes. Pops are a long click. Thumps are a sudden level shift, lots of bass energy as the level changes.  You've used all three words. Is the sound one or all of them. If there is any "thump", you have a DC level shift and you should be able to find it with a voltmeter.

[simon111]

Here's a soundclip that I recorded.
You can hear the noises (daren't say if they are pops,clicks or thumps in case I get it wrong!)
I am switching the +switch point between off and the +v

You can clearly hear the noises.
Surely this isn't right?
http://www.indigoc.co.uk/pop.mp3

Hope this can shed more light on the subject.

I'm trying to get to the bottom of it!!!!!!!!!


PS. I have changed the relays to NEC EB2 type (as per RG's stuff!)

[R.G.]

OK, I need you to do a couple of things.

(1) Measure the resistance from your switching circuit ground to signal ground(s).
(2) if that's shorted, open it.
(3) if it's open, connect them with a 10K resistor.
(4) tell me where relay 4 is physically with respect to the relays that switch audio.

[simon111]

First off, Thankyou very much everyone (especially) RG for your ideas thoughts and time with my current project. Much appreciated.

Right to answer your questions.
The resistance between the switching circuit ground to signal ground (all now tied together) is open.
So I connected a 10K resistor between the signal ground and switching ground. Still no luck I'm afraid. No better or worse.

Relays four is the one at the bottom of the photo below. (please n.b. I am using SMD relays - only because I ordered the wrong ones!!!)


Here are two more pics of the project that may of some help.



If I can't this to work then I am already looking at using a couple of CD4053 (as shown on the geofex site for remote switching) (but everyone keeps telling me that relays click less than the CMOS chips!!!)

On a tangent here.
All the projects I have seen for relay's are using them as BYPASS switches. I want to use them as a LOOPER.
I have built up a DPDT switch looper in the past (mono). I got that out last night. That makes exactly the same noises as my relay driven one!

Hopefully I am not clouding the solution by bombarding everyone with as much info as I can think of.

Simon

[R.G.]

I have built up a DPDT switch looper in the past (mono). I got that out last night. That makes exactly the same noises as my relay driven one!

OK, so Mother Nature is trying to tell you something.

The relays issue may just be a distraction. Mother has shown you that a completely passive setup (DPDT switches) also has pops and thumps. That means that your relays are probably working fine, or at least fine enough that you can't see any imperfections. If your passive setup has powered LEDS, then there is still a chance that the switch setup is feeding DC back into the signal to be switched, but if you pull any batteries or power source to the LEDs out and the DPDT setup still does it. then the switches are not the problem.

What is happening is that something about either your switching setup OR THE THINGS YOU'RE SWITCHING is causing a thump/pop/click. That's what I was referring to when I said:

(b) there may be a flaw in one of the things connected to the relay setup; use your DMM to check for DC voltages on the relay pins that carry signal (not the coil) both before and after switching.

It is possible that one or more of the things you're switching has DC on its input, and is feeding a pop back into the switching arrangement. To find out if this is really the case, you need to use a DMM to measure for any DC voltages on the inputs and outputs of the plugs that plug into the switching arrangement, and also between the grounds of the things that plug into the switching arrangement.

It is uncommon, but does happen that amps feed DC back out their inputs, or that there is a DC difference between two pieces of equipment. If the switches don't do it, maybe the amps do it.

[Transmogrifox]

If it pops with the relays completely desoldered and removed, I think you need to redesign the circuitboard.  Your schematic looks fine, and you wiring looks fine.

I noticed you have some rather wide traces on the relay power side.  In addition, the signal traces are wide.  You may be capacitively coupling the the switching noise through the table it's on. 

Make the traces more narrow, except for signal ground traces. 

Ensure that you have a high quality material for your circuitboard so you can ensure that the material on which the copper is clad is not appreciably conductive.  Use your Ohm meter between the signal traces and switching traces to verify that the board material is adequately insulative (should be too great for your meter to read, resulting in an "OL" condition. The wide traces also decrease the resistance to the board, so if the material is cheap and rather conductive then you may be injecting switching signal into the signal path directly. 

Etching your board with a ground plane may help reduce this problem.

Just some more ideas.

[simon111]

Right guys (again thanks for your info)
I have decided to go back to basics.
So built the circuit in the first post (RG's) - excluding the LED.
And I am only switching the signal line. In a DPDT looper type configuration (A/B)

Guess what.......

No massive pops/ clicks or thumps.
It does make a noise but you have to turn everything up full to really hear it.
It works well enough for my needs.

Now all I have to do is have that switching a stereo output.

Eg. the circuit as is is switching the input signal from A to B and the output from A to B. All i need is to add an extra SPDT switch so it will switch stereo. I.e one SPDT does the LEFT one SPDT does the right.

I also want an indicator light on the unit showing both states. So what I'll do is use two DPDT relays in total.
One for the input / led switching.
The other for the stereo output.

As you say I am not sure why my original one doesn't work.

But it seems hard wiring components together works. While on a nice PCB it doesn't!

Anyway. My only worry now is that there are going to be any ground loops on the left and right channels as the ground points are all wired together.
(Effect A and Effect B are both mains powered) - And the two amps (left and right) will not always be the same model. Therefore if my theory is right then I 'might' get ground hum caused by all the signal grounds being connected together.

Would you perceive this to be a problem?

And if so is there anything I can do before I build this unit again to make sure that it doesn't hum. - Ground lifts?? - Isolators??? (my solution was switching the grounds so as to keep Effect A and B completely separate , but this seemed to accentuate the clicking problems in my above photographed design)


I'm learning.....very slowly and from mistakes but progressing!


Simon

[simon111]

One very small question. (last one honest!)

I have built the stereo version.
It works! All the grounds are together but so far so good.
So a big thankyou to all who have helped.

One thing....... 

When I am switching the device off i.e B effect back to A with my footswitch (latching footswitch). It has a slight delay which cause little to no noise. But when I switch from A to B (ie energize the relay coil) it causes at times a bigger noise than usual. Is there any simple way to slow the this down?

I have changed C1 but that effects the turn off time delay. I have changed C2 to a large capacitor that I had lying around (1uF) and this didn't seem to make a perceivable difference. (perhaps I need a larger value).

Any clues?

[grapefruit]

Is the relay control GND seperate to the audio GND?

Stew.