DIY Audio Source Switch Box: Avoiding a loud pop through speakers

Well that's not a schematic, and is pretty unreadable because of it.

But my immediate thought would be simply different DC levels at the preamps or power amps, nothing to do with the relays at all.

I should have mentioned that I can measure the spikes on my scope without any preamp attached.

As Nigel says make sure that all the incoming signals and the input are at the same dc level. Have them all capacitor coupled and the switch side of the capacitor connected via a high value resistor TO THE SAME ground reference.

Les.

Thank you both for your advice. I'm sure that my inexperience is already on display, so it's futile to try to hide my ignorance.

Les, regarding the difference in DC between the preamps, I am not measuring any DC at all from the preamps under test as they are already internally capacitor coupled. So, with or without any connections to my box there is no difference in DC between any of the inputs and yet I see a spike on my scope whenever the relay come on or off.

I do have DC coupled preamps that I may want to A/B so I'm definitely going to take your suggestion and capacitor couple the output from the preamps. Les, may I ask for confirmation about the details of your advice? It sounds like I might:

1) add a 10-47uF bipolar capacitor (I have 47uF available) from the (+) side of the preamp common output on the relay (i.e. right channel would be Relay 5 - Pin 12 in my diagram), to (+) of each RCA common output (to amps) and then

2) connect a 20k resistor from the relay side of the cap (in parallel) to the star ground point (near the DC side of the rectifier board)?

Is that what you meant and are those values close?

Finally, when choosing the resistor value do I have to be concerned with the input impedance of the amplifier downstream from the common output of the preamps? In other words, if the downstream amps have input impedances of around 20k, does the added resistor leading locally to ground have an effect that needs consideration? Thanks!

It is the impedance of the signal source that effects how much loss of signal you will get. If the signal source impedance is 20K then loading it with a load of 20K will reduce the signal voltage level by half. In effect you have potential divider.

Les.

Hi Les,

That makes perfect sense. So if I use a 47uF cap and a 1K resistor the high-pass frequency is about 3 Hz which I would think is fine, right? Per your suggestion for fixing the DC mismatch, is there any reason that I should use something higher than a 1K resistor? Thanks again --Jonathan

Adding the 47uF cap and a 4.7k resistor did not make the pop any better but did introduce some RF noise into the speakers. Maybe I connected things wrong? As described above, I added the cap in series on the positive signal wires between the relay and the output RCA terminal (leaving the grounds alone) and then attached the resistor in parallel from the relay side of the cap to the star ground on the chassis. Does that sound correct?

There is no positive signal. Audio is an AC signal of varying amplitude and frequency. You also need to add capacitor and the resistor to the input of the the amplifier being driven. (This will be done at the output of the switching device.) The idea of doing this is so there is no change in the DC component of the signal when the relay contacts open or close. The voltage at the bottom end of the resistors does not have to be ground. The important thing is that it must be the same value at the bottom end of all the resistors. I would try to avoid using electrolytic capacitors if possible as the can have some leakage current. Only you know what the source impedance is of all the audio sources and the input impedance of the amplifier that the switching device is driving. After looking at the picture of your switching unit another possible cause of your problem could be coupling between the relay dive signal and the audio signals as you have not used screened cable for the audio signal wires. As requested early in this thread you have still not posted the schematic. We don't know if the relays are being driven from the same power rail as any of the audio circuits.

Les.

It's not clear to me exactly what you switching with all those relays.
Can you clarify that for me?

Thanks Les! Until I connected the newly added resistor to the star ground the entire voltage path was separate from the audio signal. Once I connected the resistor the RF noise started, so that explains the noise issue. Before I tried implementing your first suggestion of adding the capacitor and resistor I shorted each of the preamp inputs with phono cables and the popping disappeared. Does that perhaps suggest that the popping isn’t leakage from the power rails to the audio circuit but rather a DC mismatch like you and Nigel have always suspected?

Les Jones said:

After looking at the picture of your switching unit another possible cause of your problem could be coupling between the relay dive signal and the audio signals as you have not used screened cable for the audio signal wires. As requested early in this thread you have still not posted the schematic. We don't know if the relays are being driven from the same power rail as any of the audio circuits.

Click to expand...

Other than where the DC enters the relay blocks I’ve kept all of the voltage quite far away from the audio signal. I am attaching a hand drawn voltage diagram (I hesitate to claim that it is a schematic) to illustrate the wiring scheme. Apologies for my lack of skill creating a formal schematic. I admit to being somewhat of a novice, so I do appreciate people’s forebearance.

I had tried using shielded cable as you suggest, but the lengths were so short that the flexible cable proved too stiff to work with. But if it’s essential, I can measure carefully and try again.

Les Jones said:

The voltage at the bottom end of the resistors does not have to be ground. The important thing is that it must be the same value at the bottom end of all the resistors. I would try to avoid using electrolytic capacitors if possible as the can have some leakage current.

Click to expand...

I might need a bit more clarification about how to wire up the caps and resistors. You explained well the AC nature of the signal, but may I refer to the ground connection as (-) for simplicity? I tried 2 caps connected to relay terminals (output L+, R+) on one end and to the corresponding RCA plugs on the other end. I connected the matched resistors to the L- and R- terminals of the relay (no ground connection). Please see photo.That didn’t make the popping better and the circuit seems noisier. Is that wiring scheme correct? Did I connect the resistors properly?

Many thanks! --Jonathan

crutschow said:

t's not clear to me exactly what you switching with all those relays.
Can you clarify that for me?

Click to expand...

Hi Crutschow! Relays 4 and 5 are for switching between two preamp inputs and sending to a common output. Relays 1-3a are for switching between two amplifiers sending to a common set of speaker out cables. Working on the preamp before even attempting the amplifier section.

--Jonathan

Are the relay contacts make-before-break or break-before-make?

Are the relays switching both the signal wire and the ground wire from the RCA input connectors or just the signal wire ? I am thinking that having a separate star ground for the signal wires might help, Are the voltage regulators linear or switch mode?

Les.

alec_t said:

Are the relay contacts make-before-break or break-before-make?

Click to expand...

They seem to behaving as break before-make-switches. I'm getting some feedback on another forum suggesting that I should consider adding optocouplers and flip-flops to minimize the amount of time that there is no signal to the amplifier. Is that why you are asking? Thanks for your interest in helping me with this Alec.

JKDenver said:

Is that why you are asking?

Click to expand...

No. I was just thinking it might influence whatever solutions are proposed to prevent the pops.

Les Jones said:

Are the relays switching both the signal wire and the ground wire from the RCA input connectors or just the signal wire ? I am thinking that having a separate star ground for the signal wires might help, Are the voltage regulators linear or switch mode?

Click to expand...

I have both the signal and ground wires connected to the relays. Are you thinking it may be better to tie all of the ground wires from the RCA connectors and relays directly to a separate star ground on the chassis? If so, how would you recommend that I configure the coupling caps and resistor? I

The voltage regulators are linear. Except for the use of the 433Mhz remote control relays to turn the relays on/off this is all pretty old-school.

Up until now I've had everything connected to a 55w TDA7293 solid state amp (with no input capacitor coupling), but I just switched to a 35w tube amp (similar to a Dynaco ST 70) the noise is completely gone and the popping is reduced to the point of a non-issue. The difference in the power output of the two amps could certainly be one factor, as the popping noise has always stayed the same regardless of the preamp volume. But I'm not sure what to make of the disappearance of all the noise. It makes me think that your suggestions for properly coupling and grounding the output of my switch box will prove to be correct.

Thanks for all of your help Les!

I've now tested my preamp switching box on four different amplifiers with the capacitors/resistors that Les recommended and then tested again without them. Things work much better with the with the additional components. Without the caps/resistors the pop is always an annoyance with every amp. The higher the amp's power rating, the louder the pop.

With the caps/resistors: 20W tube amp the switching noise is virtually non-existent. Only slightly more noticeable on the 35W tube amp and not a problem. 120W Onkyo 504 and 55W TDA7293 solid-state amps benefit from the caps but the pops are definitely an issue.

So, that's definitely progress. Thank you Les! The question is whether the pops can be reduced further through additional measures.

From what I can discern by listening I don't think the grounding or shielding is a problem as the noise floor seems to be pretty low.

As you can see in the photo above, the 47uF MKP caps are huge and unwieldy so I'd appreciate it if people could recommend a different approach to the coupling. Should I look at smaller value MKP caps and higher value resister (e.g. 22uF cap and 10K resistor) or is there an even more sensible approach that won't degrade the signal path?

I dont see what looks like a "stiff" ground in the system for starters. Are your amps
also dumping their power into that star ground ? "Normally" one creates, in an
analog system with small signals and power switching, a ground for the low level
signal circuits and one for the power, and join them where they come into the
chassis as a single point ground. Eg. connect the grounds together as close as
possible to the source, in your case the entry to chassis.

Your scope, where are you monitoring noise, for the 12V supply to the star ground ?
Curious, have you done that with scope set to single shot, edge trigger, say 12.5V,
sweep start at 1 uS, acquisition say 100 MSPS, and see what you get. Might be
significantly larger that you are aware.

You could always consider a time delay relay that keeps a ground on inputs
until relays and power settles, then lift the input grounds. When switching small
signals one usually uses signal type relays, like the ones found in scopes.

Relay noise is predominantly off cycle where large V transients occur. So sounds like
inputs are picking up control signals either from ground drops or parasitic C
coupling....., using scope, look at each circuit, one at a time, the other disabled, to
get clues about noise origin....just a thought.

What is the model # of your scope ?

Google this "mixed signal grounding pdf" for some guidelines.


Regards, Dana.

I like danadaks's suggestion of adding a delay circuit that grounds the inputs while the switching takes place.
The choice of capacitor size depends in the input impedance of the circuit on the output side of the relays and the lowest frequency you want to pass through without significant attenuation. (This is normally given as the -3db point.)
You have NOT answered the quesions in post #14

Les.