Tube Amp Forum: The Ultimate Tone

Full Version: Channel switching
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Hi Guys

Channel switching is the most common requirement in guitar amps as far as having any kind of "switching" system. It is truly quite simple, but can be made "too simple" sometimes impairing the ergonomics for the player.

As we know from "The Ultimate Tone" (TUT) series, ANYTHING can be put on a switch and once you've done that, all the switchable things can be synchronised and/or be made remotely accessible, all at the whim of the builder or player.

London Power offers a range of switching kits and has a PDF to make selection easier.

Switching can be done using relays, transistors (BJTs), jfets, mosfets and even tubes, depending on what needs to be switched. Sometimes the simplest circuit can turn out to be tricky to design or to get working reliably, but we will show step-by-step methods of design and get you to a satisfying finish.

Have fun
Great to have a section of the forum dedicated to switching! I have yet designed or built a two-channel amp, but I want to do so in the near future. Something like a LPSP with the two channels sharing the first gain stage (and maybe staying separate from there; but who knows). My plan is to start as simple as possible.

I like the form factor of the ERK, but are there disadvantages to using jfets (say, instead of relays)?
Hi Guys

There are pros and cons for every switching element.

Jfets are inexpensive, require no power to be on or off, and come in a wide variety of capability. For switching, you want low on-resistance and suitable voltage rating for the circuit point. Using them as shunt switches is simple but as series switches they require a bipolar supply, which you may find easy or difficult to employ. The gate control voltage is not 100% isolated from the channel, so a small amount of care must be taken in the placement of parts.

Relays require a bit of power to energise and this current has to come from somewhere. Supplying the coil is usually the biggest hurdle in using relays. otherwise, relays provide negligible on-resistance and near-infinite isolation between poles and to the coil, so the coil control circuit can be at any voltage, as can be the contacts. Relays cost a lot more than jfets regardless of what deals you might come across. The contacts can be configured for series or shunt switching with no effort at all.

Have fun
I've reviewed the example 2-channel wiring of ERK on page 3 of the instructions, and I've read on the LP website that the LPSP uses (or can use) the ERK for shunt switching. And you have told me before that the Jfet gate control voltage should be high enough to accommodate the peak signal that is present at the mute point. So here come some basic questions:

- Does this mean that the preamp signal must be <12V for 12V Jfets? (just checking my understanding; not assuming that I understand correctly)
- Wouldn't the signal at the second mute point of a high-gain channel (e.g., in the gain channel of the LPSP) be greater than this?
- How can ERK be used to mute the signal toward the end of a high-gain channel?
Hi jmcd

If you look at TUT Fig.9-90 which shows A,B,A+B switching in the LPSP, the signal levels are shown throughout the preamp and specifically at the muting points. Everything is well within the capabilities of any jfet (at the mute points).

The jfets supplied in London Power's Electronic Relay Kit (ERK) are rated for 25V and the control BJTs for 40V. Provided the filter cap is 25V or more, the zener could be changed to 20-24V and signals of twice the standard amplitude can be handled. As my post above stated, the control voltage is the restriction to signal size in most cases.

If you have a preamp that you wish to add muting or switching to, measure the signals throughout the preamp prior to adding the switching. These measurements will guide you towards the correct switch element type and/or the voltage rating it must exceed.

Usually at the output of multiple preamps it is desirable to use series switching, which ERK is not designed for. You could add series resistors to allow shunt elements to control the selection, but this incurs a signal loss and increases the required impedance for the input of the following circuitry. The latter is not usually a problem in a tube amp.
Thanks Kevin. I look forward to learning this stuff!

I recently sent you a couple of series switch variants to switch in/out the input gain stage of a 2204-type amp. Now I imaging that I could restructure this switch as a shunt switch. Time to roll up my sleep and get to work!
Hi jmcd

One thing I forgot to mention above is that in the TUT Fig.9-90 example, attenuation has been added so that the signals at the mute points would be within the capability of the BJTs. A BJT can only handle an AC signal of about 6V peak before it clips the signal. The signal level needs to be quite low where the the channels come together prior to the final tube stage, which makes it convenient as a switching point and allows proper balancing of the channel loudness.

In the LP-PRE kit, based on the London Power Standard preamp, switching is more likely to be by jfet or relay and the signal levels are still quite low at the various muting points.

The plate voltage to the preamp will influence the signal sizes at the various points throughout the circuit. Obviously, with a high B+ the plate signal of a second or later stage triode in a high-gain preamp can easily be 120V peak or more - often asymmetric, say +80V and -120V. Needless to say, there is no place in the circuit other than driving the output tube grid or the OT itself where we NEED such a high signal, which is why there are interstage attenuators everywhere. Even at the power amp input we only need 2V maximum to clip the output, so there is a lot of opportunity within the cascaded stage design to add muting and switching elements.

For a clean path, a single muting point is sufficient to kill the clean sound, where a high-gain channel should have at least two mutes.