Speaker Cabinet A/B Switching

[bob p]

I'm interested in doing some A/B comparisons between speakers and cabinets attached to the same amp head. I'd like to be able to A/B between the speaker loads while playing, so I guess its time to build an A/B speaker switching stompbox.

What sort of practical considerations do I need to look out for? Is a simple SPDT type of stomp switch good enough to do the job? I had thought about leaving the ground leads to both cabs intact and only switching the "hot" lead. Or am I overlooking something? In doing this, am I likely to run into nasty pops and clicks?

I won't be A/B switching different heads with the same cab, so I'm not worried about the problems associated with leaving an amp running without a load on it -- I'll just be switching one head back and forth between two cabs.

Thanks!

[Ray Ivers]

Bob,

I had thought about leaving the ground leads to both cabs intact and only switching the "hot" lead.

I've been using a switch like this for years, with no problems (I switched both hot and ground and isolated the jacks, for no ground-loop issues if/when using it for two-heads-one-cab) - I always stop playing while switching, though. You could connect a 220 ohm 5W resistor or similar across the input jack, which should protect you pretty well in case switching takes place during a note and a momentary open-contact condition results.

Ray

[bob p]

You could connect a 220 ohm 5W resistor or similar across the input jack, which should protect you pretty well in case switching takes place during a note and a momentary open-contact condition results.

Ray,

I wanted to ask about your recommendation for placing a 220 ohm 5W resistor across the "input jack." I guess that in saying that you mean across the input jack to the switchbox / across the output from the amp.

I've wired up a DPDT switch, where the left SPDT half switches the "hot" lead and the right SPDT half switches ground. the output from the amp goes to the "center" pair of contacts on the DPDT switch, and each set of "end" contacts go to one of the speaker cabinets.

the switch works well enough, though i do run into the occasional circumstance where the flip of the switch isn't rapid enough, and there's a momentary lapse between the time that one set of contacts is broken and the other one is made. i think that I need to do something more to protect the amp during the occasional "open contact" condition. optimally, i guess that i should be using a make before break (shorting) switch.

Do I understand your recommendation correctly -- that you'd put a 220 ohm resistor across the output of the amp and leave it there at all times? I can see how having a 220 ohm load would be better than an open circuit, but won't that have an adverse sonic effect when the speaker is switched-on in parallel with the 220 ohm load? Granted, the lower impedance would make the largest contribution to the net impedance when they're wired in parallel, but there's something about leaving a fixed resistance in parallel with the speaker at all times that just seems aesthetically wrong. :/

thanks.

bob

[Matt T.]

...but there's something about leaving a fixed resistance in parallel with the speaker at all times that just seems aesthetically wrong.

Sounds like a great 'preliminary test' to do with your new A/B switch ...try it with no 220R and with a 220R and see if you can hear a difference.

FWIW, I've had amps sound better when I ran them through an Airbrake attenuator but still left the Airbrake 'wide open'...I assumed it was because, for whatever reason, the amp 'preferred' a more resistive and less reactive load. I have no idea what that resistive load is with an Airbrake set to 'no attenuation'.

Actually, my Airbrake is my resistive load when using my oscilloscope. I just dial it all the way down and lay the cabinet flat...you can hear it but not loudly...I'm not even sure I needed the speaker connected but I did it anyway.

Geez, why the heck do I type the word 'resistor' or 'resistive' incorrectly like...75% of the time...it's not like I haven't typed them a bazillion times.

[Ray Ivers]

Bob,

optimally, i guess that i should be using a make before break (shorting) switch.

Yes - I've seen toggle switches like that, but not as yet in push-on/push-off stomp-switch format. There may be some industrial footswitches for machinery that have this feature, but they're large and expensive.

Do I understand your recommendation correctly -- that you'd put a 220 ohm resistor across the output of the amp and leave it there at all times?

Uh, well... I wouldn't do that myself - I would just stop playing while switching. When I make recommendations on a public BBS, though, I always try to err on the side of caution.

A 220 ohm resistor will draw about 4W of power from a 100W amp at full output, in addition to damping the speaker slightly at all times - to many these effects are negligible, but to me they are not. Here's the circuit I would use:




This circuit has zero effect on an amp's output up to @ 120W @ 8 ohms with a normal speaker load. The component values can be juggled for best operation at other impedances and/or power levels.

Ray

[bob p]

FWIW, I've had amps sound better when I ran them through an Airbrake attenuator but still left the Airbrake 'wide open'...I assumed it was because, for whatever reason, the amp 'preferred' a more resistive and less reactive load. I have no idea what that resistive load is with an Airbrake set to 'no attenuation'.

I just finished building an Airbrake and I'm still in the testing mode. For reference, here's the schematic from a real TW Airbrake that was drawn-up by another ampager:

Airbrake Schematic

when the control is turned to position 6, both of the vitreous enamel resistors are totally OUT of the circuit. i'm thinking that there shouldn't be an audible effect when its wide open, unless you can make a case for the nichrome wire in the resistor acting as an antenna. my guess is that maybe the amp doesn't really sound better with the airbrake wide open. maybe we just like to think it does? i guess i'd have to install a bypass stomp switch and do some a/b testing to know for sure.

Actually, my Airbrake is my resistive load when using my oscilloscope. I just dial it all the way down and lay the cabinet flat...you can hear it but not loudly...I'm not even sure I needed the speaker connected but I did it anyway.

I haven't ever used my Airbrake as a sole dummy load. I have other good dummy loads around, so I just use them. One of hte things that I have thought about doing is to hook a dummy load up to the speaker output of the Airbrake, and do some insertion loss measurements to calculated the -dB down points for each position of the switch dial and rotary dial. that would be a good way to double-check my attenuator math.

while we're on the subject of the Airbrake, do you use the 150 ohm rheostat very much? when i first built mine i left the rheostat out, but i found that there wasn't enough attenuation for some circumstances. i rebuilt the Airbrake putting it in. some people say that you only really need the fixed resistor and the tapped resistor, and that you can leave out the rheostat because the attenuator sounds best at more moderate attenuation levels. even though it does sound best at moderate attenuation levels, i find that having the extra level of attenuation available is helpful for playing in the wee hours and makes the added expense worthwhile. the one thing that's hard to get used to with the Airbrake is the loss of bass response when speakers are fed milliwatt-level signals. things tend to sound kind of thin and fizzy.

[bob p]

Uh, well... I wouldn't do that myself - I would just stop playing while switching. When I make recommendations on a public BBS, though, I always try to err on the side of caution.

well, somebody else recommended a similar value resistor in a similar situation and i didn't really understand it there either. that thread made me think of this one, and since i really didn't understand the rationale behind the 250 ohm resistor in either thread, i thought it to be to ocoincidental. i thought that i had to be missing something, so i thought i'd ask.

Here's the circuit I would use:




This circuit has zero effect on an amp's output up to @ 120W @ 8 ohms with a normal speaker load. The component values can be juggled for best operation at other impedances and/or power levels.

ok Ray, I'll bite. can you explain the theory?

[Ray Ivers]

Bob,

i really didn't understand the rationale behind the 250 ohm resistor in either thread, i thought it to be to coincidental. i thought that i had to be missing something, so i thought i'd ask.

Just FWIW - I came up with the 220R value during simulation as the lowest standard value I could get away with and still have acceptable OT-primary spike attenuation along with minimal loading. I found that above maybe 300R or so the spiking was intolerable (in simulation) and I just wasn't comfortable going lower than 220R - but I wanted better attenuation and zero loading, hence the Zener circuit.

ok Ray, I'll bite. can you explain the theory?

Well... when the waveform swings too far positive, one Zener conducts like a conventional diode and the other avalanches ("Zeners"), so you basically have the resistor in series with both diodes as a load on the OT secondary - in the opposite direction, the same thing happens in reverse. It's what I've seen referred to as a 'Zener bounding' circuit. During normal operation with a correct secondary load, it does nothing.

I figured out what the maximum peak-peak voltage was that I wanted to pass without Zener conduction, along with my ideal load during the conduction period... then chose Zener values equal to the positive and negative peak waveform values, and began simulation. It's kind of a complex calculation since the resistor voltage drop adds to the total voltage swing during Zener conduction - plus, I wanted to use 5W Zeners and a reasonably-sized resistor - so I ended up with the values shown for a 120W/8 ohm output stage.

Ray

[bob p]

Ray, I've run into another example of the "Zener bounding" circuit -- its in a HiFi crossover system, where the Zeners are essentially in parallel with the tweeter. I'm thinking that it must be some sort of tweeter protection circuit, so that when the breakdown voltage is exceeded, the HF signal gets passed to ground. Am I on the right track? Of course, I have absolutely NO real knowledge about Zener bounding, so my take on this isn't much more than an educated guess.

Link to schematic

[PRNDL]

Ampeg System Selector

The Ampeg System Selector can do what you want.
Also check out the Morley Tripler switch, which switches 3 amps into one speaker.
And the Radial Tonebone Headbone Amp Head Switcher

I'd love to see the schematic for them.

[bob p]

Thanks. I'm not familair with those products, though I have found plenty of Morley schematics on their site. Their tech support staff has always been very generous, both with schematics and parts. I've received free parts from them on many an occasion for their vintage gear just because I asked about it.

For this project, I'm not particularly interested in purchasing a commercial product -- I actually built the switching unit several months ago, when this thread was first started. As a matter of timing, right now I'm more interested in the theory behind the practical application than the practical applicaiton itself.

Just for an update, a well-informed source tells me that the tweeter protection circuit I linked to earlier works in exactly the way that I had interpreted it to work. (just in case anyone is following the tech side of things)