first layout attempt: Guitar Synth

[12afael]

can you post some soundsamples?
look interesting.

[choklitlove]

i haven't made it.  i'm just in the process of the layout.

[RickL]

I built this last night using the most recent layout (with the ground trace added back to pin 8 of the 4046 that went missing after the first revision) and it's not working.

I think the problem lies in the 4007 which is listed as a complimentary pair plus inverter. First, the layout shows no connection to ground for Vss (pin 7). Second, I don't see how connecting a resistor across pins 3 (gates of complimentary pair) and 5 (source/drain of same pair) with no other connections to the other pins of the pair will make an amplifier. Same for pins 10 and 12 (same configuration). I think pins 2, 11 and 14 have to go to V+ and pins 4, 7 and 9 (plus pin 6) have to go to ground.

I'll give that a shot and see what happens.

Edit: okay I tried that and at least I'm getting some sound out of it now. I'm getting guitar signal out of pin 12 of the 4007, although I can't hear anything at pin 5 or pin 10. Output at pin 3/4 of the 4046 sounds kind of synthy but most positions of the frequency and VCO pots give just a sustained pitch out, not related to the input pitch of the guitar, pitch controlled by the frequency pot. VCO and frequency pots have to be all the way down to have any control of the pitch by the guitar, maybe a little bit of control with the VCO turned up somewhat. Do yourself a favour and wire up a volume pot (I used a 100k) at the output. This puppy is LOUD.

[choklitlove]

i knew it was a bad idea to post this before i was done...

i was hoping no one would try it before i did.  anyway, thanks for coming back with suggestions, and i'll work on getting the thing working.  if you find anything else out, please let me (and the rest of us) know.  thanks!

[choklitlove]

have you tried it without the pots?  1) put a jumper connecting pads 1 and 2 of pot 2.  2) replace R4 with a 100K.  3) put a jumper connecting pads 1 and 2 of pot 1.  that should get it back to the schematic in those areas.  maybe you'll have more success going more toward the schematic than what i added.  after we get the original circuit going, then we can add.  let me know!

either way, here's an update with an added level pot and some flaws of mine fixed.  it also has rick's suggestions:

http://i49.photobucket.com/albums/f295/choklitlove/guitarssynth6.gif

[gez]

As Rick points out, the 4007 needs power connections: pin 14 connected to V+ and pin 7 to ground.

The schematic assumes you know how an inverter is wired up.  Pin 1 needs to connect to pin 5...won't work without that connection

The unused inverter could be used for a millenium bypass.  Need a pulldown on the input.  Can't see the layout, but might need one on the output too.

With the 4046 they've chosen to use the first phase detector, which is a bit of an odd choice if you ask me...who knows!

[gez]

Ah, now I can see the output.  You need a large cap there plus a pull down resistor on the output.  I'd have to check the data sheet to see what the VCO output can source but it might be an idea to use the spare inverter as a buffer (not biased as an amp, logic type buffer).

Edit.  It would be better to divide down the output...or set up the spare inverter as an amp with gain a lot less than one (output is going to be a tad huge) and use it as a buffer.

PS  you can run a trace under C1 to connect pin 1 to pin 5

[RickL]

Was I right about the extra V+ and ground connections or should V+ only be connected to pin 14 and ground to pin 7?

[gez]

Sorry Rick, didn't see that bit.

I think pins 2, 11 and 14 have to go to V+ and pins 4, 7 and 9 (plus pin 6) have to go to ground.

Yes, absolutely right.

[choklitlove]

gez,

thanks for all of that info.  i don't know what half of it means, but i'll learn.  so, after i connect pins 1 and 5, what exactly do i have to do to make this work?

you're saying either a large cap and a pulldown resistor on the output, or divide down the output.  both of these to decrease the huge output of the circuit?  if this is correct, how exactly would dividing down the output be accomplished?

thanks again for the help, i'm just not as experienced as you.

[RickL]

Well that made some difference. I can get a reliable output from it now. I don't think the PLL is being triggered very well, I'm mostly getting squeaks and squawks but if the pots are set to give the original settings it sounds fairly synth-like, in a kind of massively overdriven, heavily compressed sort of way. I wonder if what I'm hearing is just the output of the 4007 with the PLL not actually being triggered?

Putting a volume pot on the output is the same as dividing the output down. Wire the output of the circuit to one end of a 100k pot, the other end of the pot to ground and the wiper becomes the new output.

[choklitlove]

did i do that right on the last layout?  the volume pot.

[toneman]

on the pot, pin2 usually is the wiper.
input would be at pin1.
gnd would be pin3.
output would be pin2.
as pin 2 goes "towards pin 1", volume is increased, until pin2 = pin 1 = max volume.
Normally U would have a resistor divider to keep the level in the 1-2V range.
afn
T

[choklitlove]

thanks!

okay, here's an update.  i connected pins 1 and 5 of the 4007, the level pot should work now, i changed the cap values to uFs instead of nFs, and i also made a different layout without the 2 pots i added.  this should help us get this thing working, then we'll work with those.  maybe this layout will work a little.  thanks to everyone helping!


http://i49.photobucket.com/albums/f295/choklitlove/guitarssynth7.gif


http://i49.photobucket.com/albums/f295/choklitlove/guitarssynth7nomods.gif

i also deleted the past layouts so there's no confusion.

[gez]

Well that made some difference. I can get a reliable output from it now. I don't think the PLL is being triggered very well, I'm mostly getting squeaks and squawks but if the pots are set to give the original settings it sounds fairly synth-like, in a kind of massively overdriven, heavily compressed sort of way. I wonder if what I'm hearing is just the output of the 4007 with the PLL not actually being triggered?

No, you're hearing the output of the 4046.  All this is really, is a circuit to square up your guitar signal, so it should sound very similar to a lot of comparator fuzz circuits.  The difference here is that the output of the PLL will have exact 50:50 duty cycle so that might give it the edge in sounding a little more synth like. 

I'm surprised they chose the first phase detector as it likes to see nice straight square waves, not manky approximations from the output of a 4007.  You might get better triggering if it were wired up to use the other detector.

[gez]

you're saying either a large cap and a pulldown resistor on the output, or divide down the output.  both of these to decrease the huge output of the circuit?  if this is correct

You already have a pot on the output so, as Rick pointed out, that'll divide down the output.  I don't follow layouts very well (only my own) so I got confused by the pin numbers on the pot (I use PCB mount, so pin order would have to be different from yours).  Anyway, you need a cap on the output to block DC, followed by a pull-down resistor to avoid switching pops.  You also need a pull-down on the input for the same reason.  If you're not sure what I mean check out RG's switching article etc over at Geofex.

[choklitlove]

okay, thanks for the info.  i read the article, and there's a lot of good info.  one of the only questions i have is: how do you determine the values needed for the resistors and caps?  any pointers on how to do this?

also, does this mean i need both a cap and a pulldown on both the input and output?  or just both on the output and just a resistor on the input?  sorry, but thanks for your help and patience with my learning.

[gez]

how do you determine the values needed for the resistors and caps?  any pointers on how to do this?

The pulldown on the input is in parallel with the impedance of the input stage.  To prevent loading it's best to make the resistor as large as possible.  I'm pushed for time so don't want to go into the detail of it all, but there are many posts on all this by RG in the archives.  Between 1M - 4M7 should do (latter would be a good choice seeing as the circuit already uses them).  No need for another cap at the input as it already has one (DC is being blocked).

For the output, you don't want the pulldown resistance to be too large as a voltage large enough to cause pops might develop across the resistor (Electrolytics tend to get used and their leakage is higher than the smaller caps used for input stages).  How large you make it all depends on the output impedance of the circuit.  In this case there's a pot, so there's no need for a further pulldown resistor.  You do need a large cap though, to block DC.  between 1u - 10u should be fine.  Again, best search for RG's posts on all this as he explains the whys in detail (don't have the time, sorry).

Edit:

Examples (didn't read thru but should help - search for input/output impedance if you still have questions):

http://www.diystompboxes.com/smfforum/index.php?topic=38132.msg269538#msg269538

http://www.diystompboxes.com/smfforum/index.php?topic=42049.msg303686#msg303686

http://www.diystompboxes.com/smfforum/index.php?topic=26755.msg178604#msg178604

http://www.geofex.com/Article_Folders/impednc.htm

[choklitlove]

thanks a lot gez!  shit, that's a lot of information.  "Having thought through this once or twice, most people sigh, and just use 1M."  and that's exactly what i'll do.  at least that's what i will try first.

[choklitlove]

also, "For mechanical switches, put a 100K to 4.7M (exact value does not matter) resistor from the "outboard" end of both the input and output capacitors to ground."

so i think 1M will work fine for the input.  now, for the output, you said i have a pot, so that will pull that down, but i still need a cap.  between 1u and 10u.  i'll try 1uf at first.  thanks again!