Seamless Speaker/Amp Switcher....Neil?

[nick29498141]

Hello All:

I need to switch between 3 pairs of speakers for testing purposes, think of it as a fancy A/B/X. Who has a good design for a switch? I was thinking Neil Davis for some reason... Thanks.

[neildavis]

Hello All:

I need to switch between 3 pairs of speakers for testing purposes, think of it as a fancy A/B/X. Who has a good design for a switch? I was thinking Neil Davis for some reason... Thanks.

I've made quite a few A/B switchers for both speakers and individual drivers. However, they aren't very satifying unless you have volume control at the same time. Almost always, the driver or speaker that is louder seems more satisfying initially, and that skews the results. Dennis Murphy has done a lot of testing with speakers and said that even a half db of difference will influence the results.

So the last two A/B switchers that I made use a preamp that has an integrated control bus that can change relays. The preamp adjusts the volume at the same time that it changes the relays, so you can make sure the volume levels are identical to within 1/2db.

I used a preamp design that I already had for my switcher, but it is an expensive design that has a lot more features than absolutely required for a simple A/B switcher. Aaron and I talked about ways to reduce the cost but we aren't going to tackle this project until it is obvious that there is an adequate market for it.

You are welcome to use the front panel designs, board designs and software that I have, but I really don't recommend that you attempt this particular design, because it is overkill and it will require a LOT of work. I know that answer doesn't help you much, but until I come up with a 3rd-generation version of this A/B switcher I don't have much to offer.

http://www.audiodevelopers.com/temp/Relaybox2.jpg

[nick29498141]

I knew I thought of you or a reason.. :)

But I'm not worried about true A/B testing, I'll be reviewing these speakers, and just need a very transparent switch. The one that my editor gave me/uses is pretty noticeable in the chain, and I'd like something better. 3 speakers and 2 amps would be ideal.

[nick29498141]

True Dat.

[envisionelec]

True Dat.

You thought the thread would get closed?? That only happens to me. :D

[andrewkim]

I've made quite a few A/B switchers for both speakers and individual drivers. However, they aren't very satifying unless you have volume control at the same time. Almost always, the driver or speaker that is louder seems more satisfying initially, and that skews the results. Dennis Murphy has done a lot of testing with speakers and said that even a half db of difference will influence the results.


How do you match two speaker level to within 1/2 db? What reference frequency(s) did you use? Or do you have to do averaging on multiple frequencies or averaging on overall power? Did you do a spatial averaging to get average level or did you use single microphone position? Or did you level match by "ear"?

There are people who say they can hear 1/2db difference in sound level (I suppose that depends on the frequency). Therefore, 1/2 db seems reasonable goal, but to attain that level of accuracy would be very difficult.

[neildavis]

You thought the thread would get closed?? That only happens to me. :D

OK--let's see what else I can do to discourage you so we can close this thread :D

A really simple switch box is shown in the picture. You can get lamp remote controls at Costco really cheap, but these boxes used the Heath/Zenith remotes that Home Depot sells for about $30. You can set up identical addresses for the boxes so that one remote control will switch both left and right boxes simultaneously. The boxes have 110V relays that you can buy at Radio Shack, and I crammed the lamp dimmer in the box to turn on and off the relays. Instead of nice 3mm aluminum mounting plates for the banana jacks, these boxes are plastic, and there is no volume compensation. No blue LED's, either. But they work.

Unfortunately, a lamp dimmer doesn't work well with an inductive load (relay), so it took some circuit mods (a "Zobel" resistor-capacitor network) to make it switch reliably. But once that was done the relays stopped buzzing and chattering:o. The lamp remotes work for at least 30 feet and go through walls rather well.

In order to get around having to modify the triac dimming circuit, you could use the X10 module that has a built-in relay, along with their remote. I made another switch box that way many years ago. For that switcher I just used one receiver and ran AC line between the left and right relay boxes. It's a lot "cleaner" to use two receivers like the switchers in the picture.

43

[JRT]

Hello All:

I need to switch between 3 pairs of speakers for testing purposes, think of it as a fancy A/B/X. Who has a good design for a switch? I was thinking Neil Davis for some reason... Thanks.

If you have 6 channels of amplification, you could use 6 VCAs to balance levels and provide mute function.

Here is an application note that shows how to improve the performance of VCA muting/isolation:
http://www.thatcorp.com/datashts/dn103.pdf

[neildavis]

How do you match two speaker level to within 1/2 db? What reference frequency(s) did you use? Or do you have to do averaging on multiple frequencies or averaging on overall power? Did you do a spatial averaging to get average level or did you use single microphone position? Or did you level match by "ear"?


Good questions, but I really haven't used these boxes all that much for comparing speakers. My interest has always been with driver comparisons and building fixtures for DIY events. Dennis Murphy has one of these switchers, and he has done quite a bit of speaker comparisons. He could answer your questions.

[neildavis]

If you have 6 channels of amplification, you could use 6 VCAs to balance levels and provide mute function.

Yep--having 6 channels of amplication allows you to do everything at preamp level, so you don't need relays to switch the amp output. And there are a lot of affordable multi-channel receivers with amps, so this is probably a cost-effective solution for A/B/C switching. I used the PGA4311 volume control chip rather than a VCA circuit, but the basic idea is the same.

Actually, since the 6-channel receiver will already allow adjusting the volume for each channel, all you would really need is some simple on/off input switching.

[JRT]

Yep--having 6 channels of amplication allows you to do everything at preamp level, so you don't need relays to switch the amp output. And there are a lot of affordable multi-channel receivers with amps, so this is probably a cost-effective solution for A/B/C switching. I used the PGA4311 volume control chip rather than a VCA circuit, but the basic idea is the same.

Actually, since the 6-channel receiver will already allow adjusting the volume for each channel, all you would really need is some simple on/off input switching.

I am a little interested in playing with the PDIP flavor of the sibling PGA2311A.

Its back burnered behind several nontrivial loudspeaker projects, a headphone amp, (and numerous "honey-do" projects that have been put off too long), etc.

[neildavis]

... a headphone amp,

I should have an extra TPA6120 headphone amp board, although I looked and couldn't find it in my stash of excess boards.

It has some extra power supply circuitry to power another board but you don't have to populate all the regulators. The headphone amp portion works fine. Just add a transformer...

Schematic:
http://www.audiodevelopers.com/temp/HeadphonePS.gif

PWB (actual size is 2.5 x 3.8"):
47

[JRT]

I should have an extra TPA6120 headphone amp board, although I looked and couldn't find it in my stash of excess boards.

It has some extra power supply circuitry to power another board but you don't have to populate all the regulators. The headphone amp portion works fine. Just add a transformer...

Schematic:
http://www.audiodevelopers.com/temp/HeadphonePS.gif

PWB (actual size is 2.5 x 3.8"):
47

I'm interested in buying one (or preferably two if you have two available for sale). Let me know when you find the board(s).

What are you using for the input buffer U9?

I was thinking about using a few meters of Canare microphone cable (a pair of cables) as a relatively rugged headphone cable, but am a little concerned about the CFA's sensitivity to capacitive loads. Do you think it will remain stable with the 150_pf per meter parallel capacitance of that cable?

[neildavis]

I'm interested in buying one (or preferably two if you have two available for sale). Let me know when you find the board(s).

What are you using for the input buffer U9?

I was thinking about using a few meters of Canare microphone cable (a pair of cables) as a relatively rugged headphone cable, but am a little concerned about the CFA's sensitivity to capacitive loads. Do you think it will remain stable with the 150_pf per meter parallel capacitance of that cable?

I put the ExpressPCB files and the parts list at the link below. The ExpressPCB software is free--just download it and make any changes you like. ExpressPCB has a 3-for $51 rate for that sized board and that's what I used previously. I built up two of the boards and should have one extra laying around but I couldn't find it right away. So if you want two of them you will need to order some new boards.

I used the OPA2134 for the buffer. But the board has series regulators that will provide plenty of current for any op amp you want.

The circuit is pretty much a direct copy of what is in the application notes for the TPA6120 Evaluation Module. The 10ohm resistors on the output should provide adequate isolation from the capacitive load.

http://www.audiodevelopers.com/temp/HeadphonePS.zip

[JRT]

I put the ExpressPCB files and the parts list at the link below. The ExpressPCB software is free--just download it and make any changes you like. ExpressPCB has a 3-for $51 rate for that sized board and that's what I used previously. I built up two of the boards and should have one extra laying around but I couldn't find it right away. So if you want two of them you will need to order some new boards.

I used the OPA2134 for the buffer. But the board has series regulators that will provide plenty of current for any op amp you want.

The circuit is pretty much a direct copy of what is in the application notes for the TPA6120 Evaluation Module. The 10ohm resistors on the output should provide adequate isolation from the capacitive load.

http://www.audiodevelopers.com/temp/HeadphonePS.zip

Thanks for the board data and PL.

I'm still interested in the single board if you don't mind parting with it.

Was that PowerPad difficult to hand solder?