Passive Mic and Preamp switcher-- seeking advice

[tmoneygetpaid]

Hey,

I'm working in an audio school, and one of the teachers when introducing mics and pres regularly positions a number of mics on a sound source, and switches between the mics and pres to give a taste of how they color the sound. We saw the Rad*al products that allow easy switching between mics and preamps, and I thought I could put something together a lot cheaper that got close enough.

Here is my initial draft-- I haven't done any real designing before, just repair, so any general feedback or problems anyone points out would be appreciated. It's somewhat unclear here, but the switches are 3P4T (1 pole for LEDs, 2 for the 2 audio-bearing connections; switching between 4 mics and 4 preamps). The phantom switches are 4 individual SPDT switches. The 48V converter is the Fivefish 12-to-48v adapter PCB they sell under the FABModules banner. Or I'd build a 48V PSU from a PCB kit.

I read through the old posts on similar designs, and the one issue I foresee is popping when switching. Will this happen when both preamp output and mic input are switched? Anything else I can do to minimize/ eliminate popping?

Also, is there a benefit to using relays over these switches? Switches just seem simpler and more cost-effective. I also saw mention of an old Analog Devices IC that does popless switching, but it's been discontinued and seems to only switch between two ins/ outs. Is there anything similar that doesn't need logic (over my head).

Thanks.

 

[PRR]

Welcome.

 

[tmoneygetpaid]

Thanks for the reply. Are those DC grounding resistors to reduce popping? I had some on the other side of the input switch, but they'd work better there?

 

[abbey road d enfer]

Here is my initial draft-- I haven't done any real designing before, just repair, so any general feedback or problems anyone points out would be appreciated. It's somewhat unclear here, but the switches are 3P4T (1 pole for LEDs, 2 for the 2 audio-bearing connections; switching between 4 mics and 4 preamps). The phantom switches are 4 individual SPDT switches. The 48V converter is the Fivefish 12-to-48v adapter PCB they sell under the FABModules banner. Or I'd build a 48V PSU from a PCB kit.

I don't think you need to cater for 48V power; the preamps are supposed to take care of that. Is it because you worry about switching noise? There will be switching noise anyway.
In addition, the capacitors in the signal path would interfere with signal integrity.

I read through the old posts on similar designs, and the one issue I foresee is popping when switching. Will this happen when both preamp output and mic input are switched? Anything else I can do to minimize/ eliminate popping?

I would suggest you implement a simple "kill" switch that you would push before switching mics/preamps and release after about 1 second to let the transient pass.

Also, is there a benefit to using relays over these switches? Switches just seem simpler and more cost-effective.

If you used relays, that would allow you to implement some clever logic that would automatically kill the sound for the duration of the switching exercise. Also the switching time would be faster so the switching transient would be less.

  I also saw mention of an old Analog Devices IC that does popless switching, but it's been discontinued and seems to only switch between two ins/ outs. Is there anything similar that doesn't need logic (over my head). 

These IC's were not capable of of switching the type of signals that are involved here. There are CMOS switches by Maxim that can handle those signals, but they may introduce their own sonic signature, which you don't want.

 

[tmoneygetpaid]

Thanks a lot for your reply.

A kill switch is a really good idea. Could I just put a momentary switch (normally closed, dpst with a pole each for the two audio cables) between the input switch and the output switch?

In order to use relays to auto-mute momentarily, is there a way to accomplish this without using a microcontroller? I would actually like to learn some microcontroller use, and am wondering if this wouldn't be a good simple intro project (work funded, too). What would a good platform be?

Thanks again, all.

 

[PRR]

> resistors to reduce popping? I had some on the other side of the input switch

*Any* capacitor that can *ever* go "nowhere" must have something to drain it to zero in the "nowhere" times. Otherwise when you switch TO it, any leaked charge will POP.

> "kill" switch that you would push before

Right; except it isn't tmoney (who has a clue) working the switches. one of the teachers ... switches between the mics and pres. Professor Pompous is lecturing the class, answering questions, and working switches with his clueless third mind. I used to maintain tech-gear used in teaching. The profs are brilliant in their own way, but can be remarkably stupid around classroom equipment. I see a problem.

My own feeling is that the POPs shouldn't hurt anything (there's worse sounds in a student lab than quick POPs). Let it pop. If a student is annoyed, sit them near the master mute button.

 

[abbey road d enfer]

Thanks a lot for your reply.

A kill switch is a really good idea. Could I just put a momentary switch (normally closed, dpst with a pole each for the two audio cables) between the input switch and the output switch?

I'd rather use a momentary switch that shorts the output.

In order to use relays to auto-mute momentarily, is there a way to accomplish this without using a microcontroller?

You could use good old CMOS or TTL logic or even discrete transistors to delay the command sent to the relays. It's something that can be designed relatively easily by any EE teacher.

I would actually like to learn some microcontroller use, and am wondering if this wouldn't be a good simple intro project (work funded, too). What would a good platform be?

I'm not a specialist but Arduino seems to be perfect for your application.

 

[PRR]

Call me old fashioned, but I think a BASIC Stamp is the fast-track to trivial sequencing.

May depend a lot on whether you have been exposed to BASIC programming, or must start the long path to thinking like a programmer (in which case the more-modern Arudino environment may be best).

 

[tmoneygetpaid]

Going to dive into Arduino-- might still build this as the design above, without relays or a microcontroller. If after learning a bit about the platform, I decide to retrofit it, it shouldn't be too tough to switch it over.

PRR, is the resistor value for those capacitor discharging resistors critical/ how would I determine the resistance?

Thanks again all. Your guidance in these new waters is greatly appreciated.

 

[abbey road d enfer]

PRR, is the resistor value for those capacitor discharging resistors critical/ how would I determine the resistance?

As I wrote earlier, you don't want these caps and resistors for mic switching.
This type of arrangement is regularly used to switch signals that may be subject to DC offset; the caps are ther to eliminate the DC offset, and the res are needed to equalize the voltage of the floating side of the caps (those that go to the switch) to ground (or any reference voltage the design may impose).
In your case, you need DC to flow in order to transmit phantom voltage; that means you can't have coupling caps, and thus you don't need the res.
Putting res in this position would interfere with phantom voltage.

 

[PRR]

> you need DC to flow in order to transmit phantom voltage

It appears that he has 48V and P-resistors *separate* from the preamps.

Kinda has to.... if Phantom flowed in the switching, the mikes would be "dead" until switched-to, and then there would be the mike's start-up thump and stabilization period.

However if Prof is presently switching by cable-plugging, they already have these problems.

tmoney's proposed plan shows some thought. However the dangling caps will charge to undefined voltage then discharge POP when switched-to. Resistors will tend to keep the cap end near zero.

There's no good value for the bleeder resistors. 100uFd is not a bad value for the caps (any less and you will have a rise in subsonic noise); but 100uFd will leak a fair bit. So the resistor should be "small". Without other constraint, I'd pencil 1K or less. However two 1K across the mike is 2K, in parallel with 2K-10K of the preamp input, is significant loading on the mike, and not the same loading as the preamp used direct.

However this is not Hard Learning, no "Facts" deeper than the truth that different mikes and preamps sound different. A particular SM6969 mike may sound different in the switcher than it does naked, but the real lesson is to Use Your Ears.