Combining TSL100 switches

[MrStab]

Hi guys,

Someone's asked me to make them a 4-switch channel switcher for their Marshall TSL100 head. The stock switcher has 5, but it seems some builders combine the Clean and Crunch channels onto one switch.

Everything usually uses an SPST momentary (except the Reverb and Effects switches), so to combine Clean and Crunch, would i use an SPDT ON-(on) with the common in the middle and throw that between each channel? Or would i need to do something else?



cheers!

[MrStab]

actually, now i'm thinking maybe i should use a latched SPDT? what i suggested before would have it flick between channels.

[Ice-9]

actually, now i'm thinking maybe i should use a latched SPDT? what i suggested before would have it flick between channels.

Exactly that SPDT, should do the job of having one footswitch to switch between Clean/Crunch

[MrStab]

thanks for the confirmation, Mick. i'll do that then. it eventually dawned on me how this setup actually works in use, and that latching one of the switches low was probably okay.

[Ice-9]

On second thoughts, these are momentary switches which are switching a flipflop, although the latching SPDT should work. Try it and see.

[MrStab]

going by how it works on the surface, i would assume that the state of this flip-flop only changes when another switch is pressed, as opposed to subsequent presses of the same switch putting it into the next state. i've only ever built that type of flip-flop.

anyways, i'll get parts ordered, build it and report back in case it helps someone else. or goes terribly wrong. cheers!

[MrStab]

just to recap: i'm trying to make a pedal for the Marshall TSL100 which has one switch for "Clean/Crunch" and one for "Lead", as opposed to the stock pedal, where each channel has its own switch (Clean/Crunch/Lead).

so my plan to latch between "Clean" and "Crunch" doesn't work! At least, that switch itself does, but it seems to make the "Lead" switch useless. i think i've figured it out, though. i realised that Crunch and Lead, ie. OD1 and OD2 on the front panel, are "sub-modes", and should be treated as such instead of 3 channels in isolation. This is quite common in Marshall amps, so maybe i should've actually thought about the routing sooner.

Anyways, in order to make one "Lead" switch activate OD2, it should be momentary and activate both OD1 and OD2 at the same time (to put it into OD mode in the first place, then to choose OD2). The "Clean/Crunch" switch, on the other hand, only activates OD1 like it always did. Does that sound right to you guys? i have limited access to the TSL series, at a weekly band practice, so any nods of approval or pointers in the right direction would be helpful.

cheers!

[MrStab]

basically this. both channels on the "(ON)" side, ofc, just on different sets of poles.

[PRR]

[MrStab]

thanks, Paul. my slot is booked in an hour, but better late than never! lol. there's no rush either way, just delays in-between.

so what i had originally, only non-latching? wouldn't that still hard-wire it into the "Clean" state with the standard ON-(ON) momentary footswitch, though? it didn't seem to like being locked into either. i might bring a bunch of wire, wire stippers and crocodile clips with me to try that out if my own alteration doesn't work ^

[PRR]

> crocodile clips

+1.

[MrStab]

no luck, unfortunately. at one point i pretty much tried sharing every pole across the 2 momentary switches to see if i could get the desired response.
on the upside, i came across this post, which makes up in useful info what it lacks in nomenclature!

http://www.marshallforum.com/threads/tsl-footswitch-wiring-information-for-repair-or-remote-switching.20943/

so it seems it's do-able with plain ol' latching switches, albeit with the rigid behaviour that disables the "Lead" switch when "Clean" is enabled. whenever i try to think of some exotic wiring scheme to make sure both switches are "armed" in any state, i come up with an illogical loop, so i'm guessing it's inevitable.

4013 flip-flop? to be honest i think i've kept the to-be owner waiting long enough by trying the active approach. unless i could knock together a dummy TSL100 switching section without too much hassle.

[slacker]

Do you really want to combine clean and crunch? Using the rest of the existing switching circuit, if you do that you'll lose the ability to switch from lead to crunch, which seems like something you'd want to do. If you're in lead and you press clean/crunch you'll get clean, you'd have to press it again to get back to crunch. Or you could wire it so if you're in lead pressing clean/crunch would take you back to crunch, but then you can't go from lead to clean in one press.
You'd have the same issue if you combined crunch and lead, you'd only be able to go from clean to crunch or lead not both.

[MrStab]

The way i see it, with astigmatism, is that the 2-switch method winds up emulating the front-panel control, which was never intended to be used mid-tune. what you describe is basically my own issue with that, Ian, but it seems to be what people are after in the aftermarket controllers out there (and what this particular person is after).

i'm thinking this might work: either half of a 4013 controlling either switch, each operating poles on a 4066 or 4053. Engaging the Lead switch disconnects Clean/Crunch's ground connection, but every press of Clean/Crunch activates the Reset pin on the Lead switch's flip-flop. or did i just start the weekend too early? lol

i'm planning on building both the plain-SPDT version and the CMOS version just so both bases are covered. i use 4053's a fair bit so hopefully wouldn't be fumbling around in the dark too much. may need to adjust the 560R series resistor going into the 15V zener on the stock circuit. i'm a bit unsure about any need for biasing of the inputs in that approach, too.

[MrStab]

hmm, maybe a 4066 would be bad at some of the currents i'm seeing on the 2nd page here:

http://drtube.com/schematics/marshall/TL60-60-02%20(2001)%20iss5.pdf

http://www.ti.com/lit/ds/symlink/cd4066b.pdf

[slacker]

I think this will work, the simulator says it does anyway https://drive.google.com/open?id=0B45V-V1ycMU9ZURhdURNMXAxdVk. When you press switch A so that it connects the end of the cap to ground it briefly discharges the cap so you get a short low pulse to the flip flop and it switches, the cap then charges up and the input to the flip flop goes high.
Switch A should be latching, you might need to experiment with cap values, I used 4k7 resistors and 1uF caps. Switch A switches between clean and crunch, lead works the same as in the original. This does what we talked about earlier, you can't go from lead straight back to crunch.

If you don't mind completely replacing the logic part of the circuit there's probably a way to do switching that lets you go between any of the "channels" with two foot switches. Eg: if you're on clean, switch 1 selects crunch and switch 2 selects lead. If you're on crunch or lead switch 2 changes between them, switch 1 goes back to clean. Be easy to do with a micro controller but there must be a way to do it with logic chips.

[MrStab]

cheers Ian, that might save me a complete rewire.

how about this for the other approach?



Half of the Clean/Crunch momentary DPDT switch, S1a, triggers its corresponding flip-flop (U1a) via. Q1, switching on that channel via. Q3, but is deactivated by a PNP (Q5) when the Lead channel's flip-flop output (U1b) is High. The other half of the Clean/Crunch switch pulls the Lead channel's RESET pin high. S2 is more straightforward, just triggering the Lead channel's flip-flop on without affecting the other channel (which the amp disables with the Clean/Crunch channel anyway).

i haven't considered values yet. shame there's too much current for IC switches, it might've brought the part count down quite a lot.

[MrStab]

fixed a few things and used the other reset pin to lock the amp into "OD mode" when in Lead. latching Reset high prevents the state from changing at all, right...?

i think this might work:

[MrStab]

okay, so i've borrowed a TSL60 until tomorrow!

i built this pretty much exactly* and it worked great with LEDs at 9V (without the zener):



in the studio, however, it kinda worked as-intended, but the amp seemed to make it bounce like crazy. i had absolutely no tools or components with me and i was eating into band time, so i asked to take it back to the Stab Lab! i have no hard debug data yet, but gimme a few hours and i'll have it. i need to recover from carrying the f****** thing. i even find it difficult to explain exactly how it didn't work.

what it's meant to do:
-Two 4013 flip-flops where each channel switches on and off as usual, via. the transistors with "Clean/Crunch" and "Lead" next to em
-The Lead switch "locks" the Clean/Crunch switch off (which is required for Lead to be on) via. Clean/Crunch's reset pin
-Make Clean/Crunch disable that lock by using extra poles on a momentary DPDT to send a pulse to Lead's reset pin

i've been trying to have a look at what's going on amp-side, i think that schem is here:

http://drtube.com/schematics/marshall/TL60-60-02%20(2003)%20iss7.pdf

...but to be honest, i don't have the knowledge to figure out how the relays in the amp could be affecting the bounce. i mean, it sometimes does latch, so AFAIK no transistors are being overloaded with current. surely it couldn't be as simple as even bigger debounce caps? the one thing i did try was clipping one of those caps so there was just 10n for the "Lead" switch (see footnote) and that just made it less responsive. one thing confusing me here is how much of this problem is the regular bounce, and how much of it is my use of the 4013 Reset pins? it's part of the reason i'm asking for rough ideas before i debug, i just have zero clue about any possible interactions.

i went with 270R on the PSU because i estimated the worst-case current to be 25mA. nothing overheating there. will get readings in different switch states, maybe current limitation could be to blame? the 1n4007 is just there because it's there in the original. orginal Marshall footswitch is confirmed-working on this head, btw. that uses a Darlington at one point (schem), whereas i only used singular NPNs. weird combos because i had to resort to a brick & mortar shop with minimal stock.

one strange anomaly was that the "Effects" switch, which is as simple as a latching switch & LED pedal-side, seemed to disable the "Lead" switch in certain configurations, making its status LED just flash when pressed. this could just be something dumb on my part, no shorts on the DIN inputs or cable, though.

any ideas?


*okay, i lied slightly - i really need to get rid of those 11k resistors somehow. also, the 100n debounce caps are in parallel with 10n caps (which i put in initially as per the Geofex 4013-4053 switch).

[MrStab]

update:

there was a dumb error related to the "Effect" switch, there was some sag and 1uF debouncing caps didn't help. those first 2 things have been fixed, but it's still not latching. so it's not bouncing per se, it's just not holding. i think i'm making some progress, though.

i've disconnected both Reset pins for now. upon writing this i've just realised i haven't held the Clean/Crunch half of the 4013's Reset pin low, but that's the half that does work! switching Clean to Crunch (OD1) works more or less fine now.

for now the problem seems to be the Q pin for the Lead half of the 4013 is low upon power-up, but goes high and never comes back down after the first press of the switch (save for a split-second when you do press it).  i think that's a common problem with 4013's, something not being pulled to any logic level, but i can't spot anything yet. both sides are pretty much symmetrical at this point.

if i get that working, what should i be doing re. my strategy for the reset pins? delaying them with an RC or directly connecting em (with pulldowns where necessary)? i didnt put one where the output of Lead's Q connects to Clean/Crunch's Reset pin, i figured that was a fixed logic level from the Q output


update 2:


Now it's an annoying case of one of the flip-flops not working, even in total isolation. With the reset pin grounded and the connection to the amp severed on the Lead input, that half of the 4013 just won't go low after it first goes high. no shorts measured, no obvious errors, the Collector of the NPN driving the clock (and the clock pin) receives the pulse as far as i can tell (or part of it?), the 4013 replaced... damnit! this sucks cos it's nearly 5am and i'm running outta time with the amp. bummer.