[BradJudy]

I'm not a really good DIY person, I'm mainly a PCB-stuff (aka paint-by-number DIY), but I just finished this project this weekend.  If any of the good EE's see any glaring problems, let me know, but it seems to be working well. 

(copy and paste from another forum post...)

I've mentioned a few times here that I've been building an amp switcher and I finished it today. 

First, some basics of what I've built.  This article covers some basics: http://sound.westhost.com/absw.htm To support switching of T-amps and other class D amps that have a potential between negative and ground ("balanced output" amps, bridged amps, etc), I built one to switch both positive and negative.  To avoid abuse to amps and support items like tube amps, I also added 8ohm 100W dummy loads that the non-playing amp is switched to.  That way both amps always see a load except for the fraction of a second during the switch. 

Here's a basic schematic (color coded to help follow the overlapping lines):



Some parts info:

8ohm 100W dummy loads from PartsExpress (non-inductive loads designed specifically for bench testing amps)

10A 120VAC DPDT relays with 12VDC trigger from Mouser (also available from Digikey, etc)

Hammond 1441-24BK3 12"x8"x3" black steel enclosure from Mouser

Binding posts from DIYCable

Serpac C-28 enclosure for remote

Parts image:



Notice the connector template in the foreground.  Even if you're drilling your own template, I recommend downloading the free panel design software from Frontpanel Express.  You can lay things out exactly, then print out your perfect template and tape it to the enclosure to get all of your spacing right.  It was great for things like the binding posts that need correct spacing.

Closeup of relays and dummy loads:



Binding posts installed:



To enable remote switching, I wired the 12VDC line to an output jack that runs to a remote control sized plastic enclosure with a simple toggle switch:



Dummy loads and relays installed:



Relays wired for power and tested, plus labels:



All wired and ready to go:



It works great.  You can hear a click when you switch, but once I have a lid and sit 10+ft away, I expect it won't be noticeable.  There is NO interruption of the music when switching - it's very impressive. 

It isn't a cheap project the way I built it, but it could save you a lot of money (or cost you a lot - depends how your comparisons go).  I'd be happy to give advice on building your own.

Now, this thread is NOT for discussing amps, but the amp switcher - I won't be goaded into discussing amps. 

[mgalusha]

Nicely done Brad.  So, do you have plans to add a level control either before or after the amps so you can match the levels?

Mike

[BradJudy]

Nicely done Brad.  So, do you have plans to add a level control either before or after the amps so you can match the levels?

Thanks Mike. 

No level control plans at the moment - I'm only using it with integrated amps right now, so there's no need to add another level control.  Although it would be nice for adjusting the volume between tracks of different levels - then you wouldn't have to re-match them.  I'd rather do that upstream from the amps, so that would be a different box.  Something like a basic passive pre-amp with one input and two outputs. 

I will say that using an amp switcher versus any sort of manual swapping is a world of difference in comparing amps.  There's no "I think this is more X than the other", it's just flip a switch and flip it back.  Since I didn't label the remote, it's also very easy to forget which is attached to which. 

[Russell Dawkins]

with a level control I imagine every equipment reviewer on the planet would need one of these!

[BradJudy]

with a level control I imagine every equipment reviewer on the planet would need one of these!

You'd think, but this is a simple box to build and there have been commercial ABX boxes (QSC made some for a few years), but reviewers rarely use them.  Naturally, as with all things in this hobby comes controversy - there is a good amount of debate on the proper way to use AB or ABX boxes.  Happily, since I'm not a reviewer or anyone of consequence in the industry, I can use mine to my own purposes without getting pulled down into debate. 

[BradJudy]

Mike, so thinking about your volume control point, would the best solution be a pair of stepped attenuators between the pre-amp and the amp to adjust for relative gain levels?  In theory, only a single amp would need it (the one with the higher gain), but the "even playing field" theory would push towards both having an attenuator in-line, set to different levels.  Of course, pots would be cheaper, but you'd probably get more argument about impeding sound quality. 

I'm envisioning a box with one stereo input, split to two attenuation devices, each with their own output.  For pro use, you'd want it to be able to handle balanced or SE, but I probably wouldn't bother. 

[mgalusha]

Brad,

I agree, you would want to have one for each amp so the signal path would be the same. I'd use some type of buffer as well, no gain but something to prevent nasty impedance mismatches with the attenuators. Perhaps calculate the attenuator resistors to .1 db steps to allow for precise matching or use a two part attenuator with one having 6 or 12 1dB steps and a second with 10 .1dB steps or 20 .05dB steps.

Mike

[BradJudy]

Ah yes, impedance mismatches.  While it's probably overkill, I happen to just stumble across this project: http://www.diamondstar.de/jisbos/jisbos_overview.html for a discrete unity gain buffer.  Glass Jar audio is selling complete kits for $41 for a stereo pair.  One interesting side-application would be having a buffer box to play with just for impedance mismatch testing. 

I hadn't thought about dual attenuators with different scales.  I'd need an inexpensive source for attenuator switches/frames, otherwise this could get expensive real fast.

BTW: I was reading another article on ABX testing that was set up for level matching with a mute setting and probe points for matching with a volt meter rather than an SPL meter.  A cool feature, but not something I'm going to put into this right now.  Although there's room on/behind the front panel for a panel mount volt meter. 

[mgalusha]

BTW: I was reading another article on ABX testing that was set up for level matching with a mute setting and probe points for matching with a volt meter rather than an SPL meter.  A cool feature, but not something I'm going to put into this right now.  Although there's room on/behind the front panel for a panel mount volt meter. 

I have done it both ways and the interesting thing about matching voltage into a resistive load is that depending on the real world impedance of the speakers and the Zout of the amp/it's ability to act as a voltage source the SPL's don't always match. Of course this is an interesting way to see how well an amplifier deals with real loads.

Perhaps a set of pin jacks for voltmeter probes, a lot less costly than adding a panel meter but of course not as cool. 

[Davey]

Attenuators between the preamp and amplifiers is the only way to do it.  A voltmeter should be all that's needed to match the gains from attenuator inputs to power outputs perfectly.

However, I don't understand the need for dummy loads in the switch box.  Those could be dispensed with and some money saved.

Dave.

[BradJudy]

However, I don't understand the need for dummy loads in the switch box.  Those could be dispensed with and some money saved.

So I don't kill a tube amp or other device that doesn't like running without a load attached. 

[BradJudy]

I have done it both ways and the interesting thing about matching voltage into a resistive load is that depending on the real world impedance of the speakers and the Zout of the amp/it's ability to act as a voltage source the SPL's don't always match. Of course this is an interesting way to see how well an amplifier deals with real loads.

Good to know. 

You don't happen to have a good source of inexpensive attenuator switches (unpopulated), do you? 

[Occam]

Audio reviewers aren't interested in doing blind comparisons (for reasons that should be obvious).

EDIT - just realized this is the lab circle.  Mod - please censor this post if it's inappropriate.

Opie,

You had ample opportunity to censor yourself. 
If you've nothing constructive to add in the way of technical suggestions or questions as to Brad's switcher, please take your trolls elsewhere where you can mount your own particular hobby horse.
Why you'd want to hijack this excellent thread into an inevitably pointless squabble is beyond me.....
What I'd like to see happen is that BradJudy's project become something that would be acceptable as an evaluation tool for all interested parties, even those dreaded Audio Reviewers. I would think you'd want the same. You seem like a very intelligent fella, and would hope that your goals would be similar.
You post was intentionally divisive, casting aspersions upon others. Other than providing some level of gratification to you personally, your comment served no purpose in advancing towards that goal of providing an acceptable tool for all parties concerned. . If you don't want to participate in the 'hands on' technical hardware development, perhaps you can, at the appropriate later juncture, provide some help in the more cerebral aspects of proper test construction and statistical analysis.
Taking potshots is easy, providing constructive input is substantially more daunting. If you're not up to it, for whatever reasons, I'd appreciate your not posting in this thread.

TIA,
Paul, Lab Circle Facilitator

[Occam]

Davey,

I assume the need for the dummy loads stems from the fact that some tube (as well as the DartZeel) amps will go into oscillation and possibly self destruct if powered with no load attached. ( yeah, I know... I wouldn't consider owning such an amp either)

Regards,
Paul

[mgalusha]

You don't happen to have a good source of inexpensive attenuator switches (unpopulated), do you? 

Pretty much any electronics supplier. Jameco has 12 position rotary switches for $2.89 each. I have an article in audioxpress on a nifty cascaded attenuator so you don't need 24 position switches. I'm sure DigiKey or Mouser would have even more options for pretty cheap.

I know there are calculators on the web to determine all the resistor values for a given attenuation (so you don't have to deal with the tedium). If you can't find one, email me and I can calculate them with a little program I have (but is not freeware, so I can't give you a copy).

mike

[Occam]

BradJudy,

That's one spiffy project!

You might consider adding a dummy switch, that does absolutely nothing when switched other than makings its 'switch' noise. In a single blind situation, with a someone else switching the amps, this would avoid giving the evaluator a clue that nothing has actually changed.

Regarding attenuators (as Davey said, the place to do this is at the amp input, iff'n you're evaluating poweramps) you'd want both amps to see both the same attenuator output impedances, as well as the source/preamp seeing the same attenuator load impedance, regardless of relative attenuation. This would indicate? that a 'L' or 'T' type attenuator might be most appropriate. I'll try and dig up some reference URLs. Then again, if the gains of both amps are close, the relative mismatch of impedances with a normally configured potentiometer might not be material.

In matching levels between poweramps, if you're going to the expense of switched attenuators, you might save costs by having different attenuation ranges for the 2 stereo attenuators. For example one stereo attenuator having an attenuation range of -18 to -12db and the other -12 to -6db, with whatever granularity you determine necessary. By switching them around as appropriate this would allow you to compare amps with up to a 12db difference in gain. This would allow you to have half the number of switch positions in the attenuators.

FWIW,
Paul

[ratso]

hey bradjudy, seeing as how we almost inadvertently got into it in a different forum (just a simple misunderstanding folks nothing serious) i thought i'd drop a line here to say WOW i'm jealous! i actually went online to buy an amp switcher and found this ugly thing

http://www.onecall.com/ProductDetails.aspx?id=862

i sent it to the good guys at ampbox and they turned it into this:
http://www.flickr.com/photos/38854667@N00/2213271739/in/photostream/
http://www.flickr.com/photos/38854667@N00/2214064828/in/photostream/


very nice transformation, but i sure wish i could do these things on my own 

nice job!

[ratso]

oh one thing i wanted to ask you. the big problem with finding a switch was they were all rated for only about 100w per channel and my SS amps put out 300w. this switch can handle up to 350w per channel. any idea what yours is rated for or is it only for your SET low powered amps?

[richidoo]

I have been told by an experienced audio designer that most well designed tube amps will be fine with no load attached to the output posts, as long as there is no input signal driving them. Since this project is not switching off the input when it switches away the load, the output must remain loaded while input signal remains on. I have been told that this is a good idea for all tube amps, but I don't know why - I just do as I'm told and that system has worked well for me, so far... haha

Brad your project is timely for me. I was just talking with some friends over the weekend about building such a beast. We were comparing a 30W ss integrated with 100W tube monoblocks. In the time it took to shut down, change over all the wires and repower, the mental reference evaporates. Not to mention cooling down the tubes, and adding various amounts of emotional distraction during the swap. You get out of the mood when the music stops, and sometimes differences are so subtle that it's the "being in the mood" part that you are evaluating...

I was gonna build Rod Elliot's version, it is versimilar to yours. But you have some great ideas here and thanks for the part list too. Once the basic switcher is made, lots of options can be added, like processor to control relays for blind testing, etc, input opamps, switchable inputs, ad nauseum... like Rod Elliot's crazy fancy version.

I guess it could be used to switch low level signals between sources/preamps too, right?
Thanks Brad!!
Rich

[BradJudy]

hey bradjudy, seeing as how we almost inadvertently got into it in a different forum (just a simple misunderstanding folks nothing serious) i thought i'd drop a line here to say WOW i'm jealous! i actually went online to buy an amp switcher and found this ugly thing...

Wow, that's a nice looking box.  It looks like it has a lot of inputs, how many does it have?