0
\$\begingroup\$

Problem statement:

  • I have 2 guitar amplifiers and I have 2 speaker cabinets that normally connect via speaker cables (1/4" connectors). I would like to create a passive device that lets me switch which guitar amplifier to connect with which speaker cabinet.

What I think want/need:

  • I think I could do this by wiring together 2 identical switches; one for the amplifier selection and one for the cabinet selection (the cabinet selection wired in reverse of the amplifier selection since the signal would be coming IN from the select amplifier yet going OUT to the selected cabinet)
  • I think each switch would need 6 connections (4x2)
  • 2 sets of 2 connections connected via switch to a single pair of connections

For context:

  • I've done some soldering for car audio applications and an old synthesizer chip replacement, and done some LED projects with breadboards and microprocessors but do not have an incredible amount of experience in this space.

After some research:

  • Within each of the 2 sections, the components must be completely isolated (the amplifiers speaker outputs do not share terminals) - therefore, I think I'm looking for DPDT (Dual Pole Dual Throw) switches
  • In an effort to super-simplify the project, I looked up DPDT "light switches" and found these.

Would that work for what I'd like to do? I have a few concerns:

  • that switch is very expensive for a light switch, and I'd need 2!
  • what if I wanted to grow this to handle 3 or 4 amplifiers, and maybe another cabinet or 2? I'd need to handle more than 2 sets of inputs/outputs.

So after some additional digging, I came across this video. My main question is around how he selected the right switch at this point in the video. The switch he selects is rated at 500 mA and I think that is too small. According to this question, even an 8 Ω speaker rated at 25 W could peak at just below 2 A. My concern is slightly reinforced here where it is mentioned:

Let’s say you’re listening at loud levels and hitting 100W peaks, and you have 8-ohm and 4-ohm speakers. Using the equations above, you can calculate that the 8-ohm speaker will demand 3.5A of peak current, while the 4-ohm speaker will demand 5A of current.

And yeah, guitars can get loud so I think I'd like this device to suffice for a max. 100 W speaker.

All that to say, if I go the Mouser route, I think I'd be safest with a slightly beefier switch, so I tried to tweak the Mouser search and much to my surprise, the results of that search are listing up to almost $100.

So - thought I'd ask for input from anyone that might be able to temper my logic and help make sure that indeed that is the type of switch I'd need to consider - or, if you'd offer a different suggestion. At this rate, I'm starting to cost more than buying something off the shelf.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ Relays are easy to find with contacts rated at 230VAC at 10 Amps .You could get 24VDC coils ,Consider the relay approach which should be cheap. \$\endgroup\$
    – Autistic
    Jan 28 at 6:03

3 Answers 3

1
\$\begingroup\$

Since you want to be able to expand the system you might as well start with a system that can grow with your needs. Whatever solution you choose, you need to ensure that you can't short the outputs of two amplifiers together so switches would need to be break-before-make type.

I suggest that you use a patch panel in conjunction with a 2-input / 2-output selector box similar to the off-the-shelf unit linked in your question. Using the patch panel removes any limits on the number of devices that can be connected.

schematic

simulate this circuit – Schematic created using CircuitLab

Figure 1. A combination of two patch panels and a 2-in / 2-out selection box gives a good low-cost solution. As shown here, either AMP1 or AMP3 can be connected to SPKR2 or SPKR4.

Notes:

  • The patch panels could be made with standard 1/4" jacks or Speakon(?) connectors. Make sure that the jacks can't short sleeve to tip during insertion.
  • You now have the ability to do A-B testing between any two amplifiers and any two speakers.
  • You need to be sure that the amplifiers are OK running open-circuit.
  • Watch out for amplifiers with DC blocking capacitors on the output. These may result in a thump from the speakers when switched in.
  • You need to be sure that you can't switch, say, a 4 Ω speaker into an 8 Ω amplifier.
  • Check your local car parts stockist for heavy duty toggle switches.
\$\endgroup\$
1
  • \$\begingroup\$ Thank you very much for taking the time supply a drawing and include concerns/caveats. The points you make have come up in passing on different posts but not been consolidated and contextualize specifically to what I'm trying to create. TY for that. I've got a few line level patchbays laying around but your pointer to Speakon connectors led me to some options that would support speaker levels google.com/search?q=speakon+patch+panel - I think the only thing remaining is to pick up those switches! \$\endgroup\$ Jan 28 at 19:16
1
\$\begingroup\$

I use rotary multi-position contact switches aka "wafer switches" for these kinds of applications. You can harvest them from obsolete test equipment like old television test equipment, etc. This gear is beyond useless and headed for the electronic recycling pile or worse, especially the old NTSC TV test equipment. It is fairly easy to find at junk stores and on eBay. Many of these units contain very high quality rotary switches that will easily handle the power levels you're considering.

And to the folks who might say wafer contacts are "insufficient" remember that the 1/4 jack on the output of the typical guitar amplifier has a contact area to the jack's tip contactor that is smaller than the contact area in a typical rotary switch. Yet 1/4 jacks have been working on very high power amplifiers for almost half a century. A Marshall Major 200W uses one with no problem. More to the point, I've used wafer switches in this exact application with no problems for many decades. I have a setup for this on my bench right now.

Of course you don't hot switch it, but why would you ever hot switch any arrangement like this (hot switch in the sense of changing positions while in the middle of a massive power chord!)

As far as expandability, many rotary wafer switches can have a large number of elements added to them by simply stacking on more wafers on a single shaft. The sky is not the limit, but a four wafer switch is not beyond reality.

The downside of wafer style rotary switch design is that working the logic of a wafer switch arrangement is somewhat of a lost art. It can be tricky to get right, but it can be done.

In the scheme I propose, you'll end up with a two layer rotary wafer switch that will give you multiple choices, thus;

Position 1

Amp A - cabinet A Amp B - Cabinet B

Position 2 (SWAP)

Amp A - Cabinet B Amp B - Cabinet A

Position 3

Amp A - null (resistive load only) Amp B - Both Cabinets

Position 4

Amp A - both cabinets Amp B - null (resistive load only)

These four positions would seem to provide for all the obvious combinations, although there are some series vs parallel options for the dual cabinet mode which could be selected by a second switch, that you could call "series/parallel" selector. It would only affect positions 3 and 4, of course.

Remember too that you need to properly deal with DISCONNECTING your amplifier if you have a "null" position for a particular amp's output. You must provide a load at all times for the output of any amplifier. It is not enough to disconnect it from the driven loudspeaker. It must also CONNECT to something that will function as at least a nominal load. There are many options. Look at the output jack wiring of a Fender 400PS amplifier for clues for how this might work without requiring massive heat dissipating resistors.

IF you wanted remote control footswitch capability, you could dump the wafer solution and use gangs of multi-contact relays to accomplish the switching. I'm not a fan, because a failed single relay might result in a surprise and unpleasant combination, but it can be done. YMMV. Such relay gangs are available pre-built on Amazon and elsewhere, and the price point is attractive.

Of course, if you use wafer switches, it is purely passive - no power required except your wrist and hand - and as a bonus you get to use chicken-head knobs to their full effect.

Good luck!

\$\endgroup\$
2
  • \$\begingroup\$ Thank you for suggesting an alternative approach! With this approach, would something like this work? thompsonguitarandthrift.com/products/… ... that seems like just a standard guitar 5-way (which I assumed would be built for low levels). Or something like ebay.com/itm/265206395626 or ebay.com/itm/133818954341 ? Those aren't that expensive ... I was just concerned about their power ratings. \$\endgroup\$ Jan 28 at 19:27
  • \$\begingroup\$ In addition, I really appreciate you emphasizing the DISCONNECTING paragraph. I'd read that here and there but wondered if it really mattered. In my current setup, I only have one cabinet and so one of my amp heads is always disconnected. I leave it in standby and swap the speaker cables (when both amps are in standby) but was doing that out of precaution only. Does standby on a guitar amp obviate the necessity to keep a load connected? In any case, it is an accident waiting to happen. What sort of electronic circuit would I need to add to provide load across the unused amp. A resistor? \$\endgroup\$ Jan 28 at 19:30
1
\$\begingroup\$

As mentioned above you have to provide load at amplifier output (it's just resistor but it should be rated for higher power) if amp you "unplug" is turned on but I don't think it's good idea to switch cabinet with amp turned on especially if it's tube amp. Whether you press switch or use relay there is short period of time when cabinet is disconnected and load isn't connected yet. I don't have much experience in that matter but it may happen that even this short period is enough to fry your output transformer. With solid state amps it's probably safer but there's still a risk.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.