4 button selective switching..

[danelectro]

Does anyone have a link to more info on a switching scheme such as this?

I basically want to put 4 EQ pedals in one box. each EQ will have its own footswitch. When any footswitch is clicked, that EQ will engage and all others will disengage.

Whats the best way to tackle this?

thanks

[toneman]

Sounds like a job 4---

ASMOP

see GEOFX pages....

up up and, away........................

tb

[danelectro]

cool ill read up. thanks

[danelectro]

ah, looks familiar. i actually tried to design some logic to control this using some cadence and synopsys tools im familiar with. I just didnt know if that was overdoing it or not. i guess not.

[niftydog]

You can buy mechanical switch gangs that will do this for you... but I know of no attempt to use them in this way, pop could be an issue...  and they're not suitable for stomping on!

I'd say you're looking at momentary footswitches and some kind of logic or a micro... like ASMOP.

[The Tone God]

Use a 4052 (which has a pair 4:1 switches) to do your switching and a pair of flip-flops to maintain the logic levels received from a couple of footswitches. Have the data and clock inputs of the flip-flops trigger by each footswitch. Easy enough.

Andrew

[danelectro]

cool thanks. ill look into that. thanks guys

[danelectro]

tone god.. what would the 4502 be doing? deselecting the non selected inputs?

i dont quite understand..

[R.G.]

I basically want to put 4 EQ pedals in one box. each EQ will have its own footswitch. When any footswitch is clicked, that EQ will engage and all others will disengage.  

See GEO: http://www.geofex.com and look at "A Programmable Footswitch System" for illustation 7 - footswitching logic. This shows a single CMOS chip that does up to eight switches so when you press one switch it latches on and turns any others off. You combine that with any electronic signal switching method, such as a CD 4066 quad switch to gate the signals through and you have the system you're looking for. Two CD4066's give you eight parallel paths to switch from.

To get click free operation from the CD 4066's, you'll need to bias the 4066 inputs and outputs at half the chip supply voltage. It also helps if the bias resistors are under 220K on the output pin sides.

You can also drive relays from the 373 latch chip's outputs with a transistor driver. See "A latching relay driver" at GEO for how. The slowed-down transistor driver prevents the relay from clicking.

And yes, the ASMOP approach works. You could use a single PIC to read four switches, debounce them and then run either a 4066 switch array or some of the 405x switches. I don't like the 405x switches because they're encoded, so you have to either patch around the encoding or use more chips. Matter of taste.

[The Tone God]

tone god.. what would the 4502 be doing? deselecting the non selected inputs?

i dont quite understand..

The 4502 would serve as the switching element. The 4502 is a pair of SP4T switches. With only two pins to decode I think its simple enough to do the decoding manually. I would use something like a pair of latches or flip-flops to store the decoded information and clock it with the same switch pulse. Once again this is just me thinking off the top of my head.

If anymore switches were needed I would agree with R.G. The article he suggested is a good read and I would recommend looking at it.

Andrew

[danelectro]

^thanks

yes i read  R.G.'s recommended article and it was great. infact i changed plans and itll now be 6 EQ channels w/ an FX loop each with global reverb and a global FX loop. like whoa. i think itll work perfectly

im sacrificing the preamp section of an old powered mixer for this it should be awesome

[donald stringer]

Would this chip work on the asmop thing. 74hct373 octal d-type transparent latch.

[danelectro]

well i drew up schematics and PCB layouts using expresspcb today for 6 channels of button logic and 8 channels of relay drivers. its basically the logic shown here: http://www.geofex.com/Article_Folders/fxswitchr/fxswitchr.htm

its $120 for 3 of each of these to be made.. obviously i only need one each. If anyone wants to split this and take a PCB of each lemme know Unfortunately 3 is the fewest you can get

these boards will also work fine with fewer buttons and/or fewer relays

[R.G.]

Would this chip work on the asmop thing. 74hct373 octal d-type transparent latch.

The 74C373 differs from the 74HCT373 in a couple of ways that are important.

First, the "C" version can work from 9V directly, the HCT is only 5V, so with the HCT you'll have to put on a 78L05 regulator or equivalent to make 5V, then buffer the outputs so that only 5V is needed.

Second, the Tdh (data hold time after clock) on the "C" version is negative by a few nanoseconds, and on the "HCT" version it's a few positive. What that means is that the data that's going to be latched in can actually change either coincident with the clock falling or slightly before. That slight amount of slop allows the "C" version to work with no diddling of time constants between clock and data signals. The Schmitt trigger inputs of the cmos chip allows this to work and act like there are clean signals there.

The "HCT" version requires data to remain valid on the data pins until a few nanoseconds *after* the clock changes, so you have to figure out some way to keep the data high on the newly-changed data pin for longer than the clock - and the clock IS the data signal.

I did this in some experimental work by using a series diode to the data pin with a cap and a 1M to ground. The diode lets the data signal and clock go high together, and the cap holds the data signal up for a few nanoseconds longer than the clock, which is not slowed by a capacitor. The 1M sets the rundown time for the data signal and assures you're ready for another different input within a few milliseconds. Again, the Schmitt trigger inputs clean up what would otherwise be a sloppy signal to the latch.

So no, the HCT will not directly sub for the C version. Close, but you have to tweak.

well i drew up schematics and PCB layouts using expresspcb today for 6 channels of button logic and 8 channels of relay drivers. its basically the logic shown here:

A couple of caveats on that. If I were committing it to PCB (especially a $40 PCB), I personally would first breadboard exactly the circuit I was going to get the boards for; Mother Nature has an irritating way of pointing out every tiny detail that we mortals didn't understand along the way.

As a particular caveat, if I were building that for myself, I'd add a resistor in series to the 1K on the gate of the relay driving MOSFETs and a slowdown cap from drain to the junction of the series gate resistors. That would let me guarantee that I kept the voltage transition on the relay coil slow enough to not give capacitive clicks in the audio path.

I really like the NEC EA-2 low signal relays for switching audio. They're about $2.50 and work very nicely indeed.

Yell with any technical questions you might have on reducing that to practice - people who build my designs get free consulting. keen at geofex point com

[R.G.]

... um... I'm assuming you did not build the any-order selector switch matrix, right? That didn't seem to be the focus of your notes.

[danelectro]

R.G. what values would you recommend for the extra resistor and cap you mentioned adding?

ill update the schematics and post them here.

thanks again

edit:

well heres teh schematics. please let me know how i can improve them

http://www.duruskyamplifiers.com/relaydriver.bmp
http://www.duruskyamplifiers.com/footswitchlogic.bmp

[danelectro]

bump

[R.G.]

Try 10K for the series input resistor, and 0.1uF for the feedback cap. These work well for bipolars, and the MOSFET should be a bit slower I think.

I recommend you breadboard one and try it - I had to diddle with the values on the bipolar version.

[danelectro]

will do. thanks a lot.

the schematic is updated