Designing a channel footswitch for the Mesa Boogie Mark V (logic switching)

[ianmgull]

The Mesa Boogie Mark V is a very complex amp that has the ability to switch multiple settings via a large 8-button foot switch that connects to the amp with an 8 pin DIN cable. There are also several 1/4" TS jacks to switch functions using latching switches. I would like to design a two or three button foot switch to control only the three amp channels.

The three amp channels can be switched using two 1/4" TS jacks. I think I've determined the switching logic. I'll using the notation V₂ to represent the voltage between the sleeve and tip of switching input 2, and V₃ to represent the voltage between the sleeve and tip of switching input 3.

The switching logic is:

Ch 1:
V₂ = 12v (open)
V₃ = 12v (open)

Ch 2:
V₂ = 0v (closed)
V₃ = 12v (open)

Ch 3:
V₂ = 12v (open)
V₃ = 0v (closed)

So essentially, when both jacks are 'open', the amp is on channel 1, when jack 2 is closed the amp is on channel 2, and when jack 3 is closed the amp is on channel 3. The complication comes when switching from channel 2 to channel 3 or from channel 3 to channel 2:

There is also a 4th state available not shown above (V₂ = 0v & V₃ = 0v), meaning both switching inputs are closed. This sets the amp to channel 3.

So if I wanted to design a foot switch, I realize there are a few options. I'm hoping to get opinions. Since I'm hoping for the most compact option, I'd ideally go for a two button. I'm wondering if it would be possible to build this with relays and LEDs, or if you'd recommend some kind of transistor switching network?

I'd like the switches to behave logically, where for example, switching from channel 2 to channel 3 doesn't JUST set V₃ = 0 (closed), but also sets V₂ = 12v (open). I think for the network to behave logically, I'm required to avoid the 4th state (where both voltages are 0v).

Very open to suggestions, thanks for reading!

[antonis]

I'd like the switches to behave logically, where for example, switching from channel 2 to channel 3 doesn't JUST set V₃ = 0 (closed), but also sets V₂ = 12v (open).

Don't you think you ask too much from a poor and hamble switch..??

[ianmgull]

I'd like the switches to behave logically, where for example, switching from channel 2 to channel 3 doesn't JUST set V₃ = 0 (closed), but also sets V₂ = 12v (open).

Don't you think you ask too much from a poor and hamble switch..??

I've realized I absolutely am!   

I'm currently reading up on flip-flops and basic digital logic (which is uncharted territory for me).

[MikeA]

If it helps in finding a logic device, you have a two-bit binary counting sequence, using V2 as the least significant bit.  0 means open, 1 means closed/shorted.
V3 + V2 =  Channel
0 + 0 =Ch1 (binary 0)
0 + 1 = Ch2 (binary 1)
1 + 0 = Ch3 (binary 2)
1 + 1 = Ch3 (unused, binary 3)

[ianmgull]

If it helps in finding a logic device, you have a two-bit binary counting sequence, using V2 as the least significant bit.  0 means open, 1 means closed/shorted.
V3 + V2 =  Channel
0 + 0 =Ch1 (binary 0)
0 + 1 = Ch2 (binary 1)
1 + 0 = Ch3 (binary 2)
1 + 1 = Ch3 (unused, binary 3)

Thanks for that. I actually just wrote down a similar truth table so it's reassuring to have the sanity check.

One point of confusion I'm dealing with in reading about digital logic:

All of the truth tables I've seen for different logic gates have one or more inputs, but only one output. It's clear from above that to have three possible outputs (binary 3), I need at least two 'bits' to form: 11,10,01, (00 not used).

EDIT: Reading up on Priority Encoders which seem like a promising route:

https://en.wikipedia.org/wiki/Priority_encoder

[PRR]

> gates have ....... only one output.

Multiple gates?

[DIY Bass]

Sounds like a job for a small microcontroller - or Arduino if you want.  3 momentary switch inputs, some debouncing and 2 outputs that drive relays through a transistor.

[PRR]

...Sounds like a job for a small microcontroller....

Only because we have all forgot how to do it, and coding a 10,000 transistor brain is "quicker cheaper" than relearning how to do it in a dozen devices. (Betcha the cost of a 7404 that Mesa did it without a CPU.)

The general problem is "decoders". Specific decoders binary to decimal were insanely popular. Often a little combinatorial logic after would do odd logic like a stomp-amp.


Hey, here's the guts of a 74154 decoder. Yes, it is "multiple gates", one per output, inputs tied per the state we wish to detect.

[ElectricDruid]

...coding a 10,000 transistor brain is "quicker cheaper" than relearning how to do it in a dozen devices.

Yes, that's right. It is.

[PRR]

Damsite easier than figuring out Mesa's partial schematic.

[ianmgull]

The general problem is "decoders".

Hey Paul, when you say "decoders", are you referring to Binary Decoders (https://www.electronics-tutorials.ws/combination/comb_5.html)?

[duck_arse]

I'll prolly get this wrong, but -

a D type flip flop [4013] with barQ tied to D, when CLKed [one switch], gives 1/0 or 0/1 outputs. when both R and S are pulled high [second switch], outputs are 1/1. that could be inverted to give 0/0. is this not a solution? 4013 and 4093, for eg.

[MetalGuy]

For reference the original FSW schematic:

[ianmgull]

For reference the original FSW schematic:

Hot Damn that helps!! I've been looking for this with no luck.

All three of those ICs are pretty cheap. I think I might just look into building something like this while removing all of the switches  and LEDs that aren't relevant to channel switching.

Thanks!!

[ianmgull]

So I've ordered parts and will post updates once they arrive.

Just a note for anyone interested:

This method will use the same 8 pin DIN connection that the factory foot switch uses as opposed to the two 1/4" TS jacks. My plan is to omit the 5 extra switches shown at the top of the schematic and build a box with 3 channel switches.

Fingers crossed.

[PRR]

That's over half the missing bits. Obviously you can build the part you think you need for a couple bucks.

But how do they select CH1 CH2 CH3? Looks to me like Pin 3 V/CH does all *three* channels with 1.5k 3.3k 5.6k and other resistors. Like if you ran a pot across 5V and put the wiper to pin 3, turn it, it would jump 1-2-3. Re-working that to stomp-switching pretty much comes to the same solution they used.

I'm not going to back-analyze the logic. I did hilite the rails and paste the pinouts, mirrored to match MB's lefty draftsperson. Three momentary switches load the '564, the '20 gates the state into the '564, the '06 seems to buffer the D/A resistor network.


The stuff in dashed blue is NOT the 1-2-3 switching. It is simple mechanical switches and lots of wires in the cable (they even use the shield and not for common). Leave that out, or put it in, there's not much to it. Note that the switches are not all wired the same!

[ianmgull]

The 3-resistance values used to switch channels is interesting. I wonder what's going on inside the amp to decode that. Perhaps a circuit that measures the three different current draws depending on which of the three resistors is high?

Also, I'm surprised to see that the shield is used to switch the EQ while having a dedicated ground pin. So many questions.

[PRR]

The 3-resistance values used to switch channels is interesting. ....

My 1996 and 2002 Hondas use exactly the same scheme for the cruise control. Except not even logic chips. So as the switch ages, if you tap gently, the few hundred ohms of contact resistance signals SLOW instead of FAST.

[PRR]

> what's going on inside the amp

Look what is there. A 10K and 1.5k/3.3k/5.6k. Just like that we have:
10K:1.5K  0.65V
10K:3.3K  1.240V
10K:5.6k  1.795V
Thresholds of 0.92V and 1.5V would sort the three values. A resistor-string, half a LM324, and a couple diodes.

[ianmgull]

Well I think I understand the logic switching a little better. I mocked up the NAND and G Flip Flop sections and one button illuminates one LED:



I kind of wish I could figure out how to implement a 2-button version (defaulting to ch1), but this is progress in the meantime. Now I wait for Mouser to send me the goods.