Signal Routing Pedal

[duffrecords]

The CMOS switching approach seems like the smart way to do it.  If you put ICs on a PCB and route the traces efficiently, you can accomplish a lot of switching on a tiny board, and it's silent.  I'm not familiar with the SN74HC4xxx series, though.  I got my 4053 and 4013 from Small Bear Electronics because they specialize in pedal parts so I don't have to filter out the hundreds of thousands of results on Mouser's web site, plus they're very helpful by email.

I wired up all the components but it still needs some troubleshooting.  One of the LEDs doesn't light up and I'm not detecting continuity on the tips of the phone jacks.  Here are some photos in the meantime.






These photos don't show the vias I forgot to etch.  The board is actually looking a little uglier now that I'm retrofitting it with extra jumpers.

[duffrecords]

Troubleshooting is finished.  It turns out the LED in question failed.  I probably brushed the anode leg and accidentally shorted it to 9V.  Checking the phone jacks for continuity doesn't work, probably because of the biasing, but I connected it to my pedalboard and it works as intended.  Now I just need to drill holes in the enclosure and close it up.

LED Legend

| BLUE	| RED	| combined	| guitar -> looper -> amp1	|
| BLUE	| GREEN	| combined	| guitar -> looper -> amp2	|
| YELLOW	| GREEN	| separate	| guitar -> amp1, looper -> amp2	|
| YELLOW	| RED	| separate	| guitar -> amp2, looper -> amp1	|

This is going to be tons of fun, creating on-the-fly accompaniment through the bass amp while I play through the guitar amp.

And now for the hum issue.  The electrician finally came and connected the open ground for me, so no more lethal shock hazard.  There is some noise when both amps are connected but it's not the 60 Hz hum I expected.  This is more of a higher pitched, sizzling noise floor.  I bought a Radial Engineering StageBug SB-6 and inserted that in line with the cable going to the bass amp.  With the default settings, it doesn't make any difference, but there's a ground lift button that removes that noise completely.  Even with the bass amp disconnected, there is still a loud 60 Hz hum, which I've isolated to the switch box.  My guess is that this is because the circuit is exposed and I have a bunch of LED lights, computer monitors, and other appliances in the room.  I'm hoping that once I close up the chassis and tie it to ground the hum will at least be attenuated.  If not, my trusty Decimator II will take care of it.  I'm also considering shielding the inside of my guitars.  Most of them have single-coil pickups and pick up obnoxious amounts of noise depending on where I'm standing, even when the guitar is plugged directly into the amp.

[macfilipe]

That is really cool, thanks for sharing your progress. 

On my side of things, I'm looking for a couple of separate things:

1 - AB box before the pedalboard to allow me to switch between the Variax and the "analog" guitars. Easy enough to do with a simple ABY box, which is what I'm using right now, but here are the twists on my wishlist: the analog guitar input should have a built-in switchable buffer (a simple JFET jobbie will do - I've successfully breadboarded it and it worked), which I don't need for the Variax, then go into the AB selector, followed by a mute / tuner out, and *eventually* a "true bypass" loop for the CryBaby. I'm looking at the CD40xx specs to see if I'll need multiple ICs (one for each switch), or if there's one that'll allow more than one of these functions, etc.

2 - "amp" switcher coming out of the Digitech Jamman looper pedal. It's a stereo looper, so I'm running the guitar through the L channel and the bass through the R (with hum isolator, otherwise we know what happens...), but I want to be able to switch between the bass amp and the aux in of the guitar amp for late night practice through headphones. Bonus points if I can integrate a passive volume control on one of the outputs to make it easier to match levels, as the guitar amps' aux IN doesn't have level control.

I'm still wondering how to do all of this using the minimum amount of parts, keeping it as clean as possible, and I'd love to have these boxes "hidden" at the back of the pedalboard, remotely actuated by switches at the front. I've been investigating R.G. Keen's "remote pancakes", and my head is already scheming to integrate the switches in a row at the front of the board itself. I suspect this should be easy enough for someone with moderate experience, but to me it will be a welcome challenge.

Bonus points: built in booster on the AB box for the single coiled guitars; true bypass loops for the Jamman in both the bass and guitar paths. 

I guess I'll be spending some time at the drawing board. 

[duffrecords]

In order to save space and fit all phone and power jacks on one side, I consolidated the looper in/out jacks into a single stereo TRS jack and made a "Y" insert cable. Amp switching still works the same as before, but now the looper starts to feed back when the input level knob is turned beyond unity. The feedback frequency decreases as the input level increases, which makes me think it has something to do with the capacitors and resonance. Here's an example:

https://goo.gl/photos/HN8bzXmNQXowdb8m6

Sent from my XT1254 using Tapatalk

[duffrecords]

Pretty simple. I have a jack that looks deceivingly like a stereo jack (3 contacts, and not the kind that separate when a plug is inserted) but the tip and ring are connected somehow. I'll have to go pick up a real stereo jack.

[macfilipe]

Pretty simple. I have a jack that looks deceivingly like a stereo jack (3 contacts, and not the kind that separate when a plug is inserted) but the tip and ring are connected somehow. I'll have to go pick up a real stereo jack.

Right - not a very helpful jack then. 

Quick question: why the 560K resistors on both the Vr voltage divider and the input / outputs? I understand the biasing concept to get the best performance out of the CMOS, so it's the values I'd like to know more about.

R.G.'s page has both 510K (close enough to 560K), and 1 Meg on two different diagrams, but no mention of the values used for the Vr voltage divider. I've also read the article about calculating biasing resistors and got nowhere really quick  I'm actually tempted to just use the suggested LM386 chip...

[duffrecords]

I just happened to have a lot of 560k resistors on hand. In the notes on the first diagram, R.G. says that the value of the biasing resistors is arbitrary as long as they don't load your signals down too much. Obviously the two resistors in the voltage divider must be of equal value in order to tap 4.5 V from the center. I imagine they need to be reasonably large or else there would be excessive current flowing through the divider.

[macfilipe]

I just happened to have a lot of 560k resistors on hand. In the notes on the first diagram, R.G. says that the value of the biasing resistors is arbitrary as long as they don't load your signals down too much. Obviously the two resistors in the voltage divider must be of equal value in order to tap 4.5 V from the center. I imagine they need to be reasonably large or else there would be excessive current flowing through the divider.

Cool, thanks. My first thoughts about the current flow through the divider went along the same lines as yours, but then I read the article I mentioned before and became a bit lost: http://www.geofex.com/circuits/biasnet.htm. I mean, 3.3K is two orders of magnitude smaller than I'd be expecting, and even though I understand that's a specific example for a given current flow demand / tolerance, it's still quite low - hence me thinking I might just go for the LM386 chip and be done with it...

[duffrecords]

Yeah, it's a balance between efficient power consumption and reference voltage accuracy.  I had read that article last year but forgot all about it.  Now I'm thinking I should use smaller resistors in the voltage divider.  My PCB is already wired, so it's too late to add an LM386.  It would be nice if someone could point us in the right direction to figure out the optimal resistor values.

[duffrecords]

I got some more stereo jacks from Radio Shack and realized this is the same product.  The ring plate in the old jack (on the right) has shifted and is slightly touching the tip lug, causing an intermittent short.  I don't know if I accidentally melted the plastic and it slid or if it was a manufacturing defect but note to self: stick with the Switchcraft hardware.



Other than that, I put some higher value LED resistors in there because they were blindingly bright and then I closed up the enclosure.  I didn't bother experimenting with different voltage divider resistors because it was late and I was tired.  Maybe I will in the future if I hear any of the clicking associated with the CD4053.  For now, at least, the pedal works great!  The ground lift on the Radial SB-6 is essential for decoupling the second amp but the hum isn't otherwise any worse than my previous signal chain was.  Thanks to everyone who helped answer my questions.  I'll try to post some videos of this thing in action once I reorganize the cables on my pedalboard.

[macfilipe]

Excellent, that's really encouraging news. I'd love to see video of it at some point, if you do get to shoot it. 

I've got my schematic sorted out with a couple of 4013's controlling one 4053 and one 4052, already ordered the electronic parts, now I need to come up with the enclosures / switches, then breadboard / prototype and put it all together. Long term project, me thinks...