dumbish question... how much dc at the output is too much?

[pinkjimiphoton]

been working on an adaptation of the old eti struzz double fuzz thing,
trying to run it on 9v instead of 18.... gotten some amusing results.
but last nite i was doing voltage measurements, and noticed with no signal being put in the input, its got nearly 3 volts output to ground measured off the footswitch!!!

i THINK its cuz i wired the power supply wrong to the actual circuit board, its very weird, and what should be "ground" appears to be hooked up to a half voltage-ish supply.

the whole mess is in the double hit deux thread. the thing SOUNDS great, i'm just worried that its gonna damage something with that much output.
i mean, i'm reading close to the b+ between the audio path and what SHOULD be ground.

the power supply is weird as hell. i seem to remember seeing something like it in an old kay guitar amp, where they used signal from both sides of a cap to drive the speaker with.




where the "ground" goes to is the chassis earth ground. the REST of the "ground".... including the "grounds" for the pots goes to the tap where it says 2.13 v. right now, i gotta 47u cap in there with a 27k resistor going to ground in parallel with the negative side of the cap!! at that junction appears to be the voltage most of this puppy is running on.

every time i try to hook it up different, the damned thing won't fire.

pretty much everything you see as "ground" in this is actually that 2.13 volts... now i got it up to about 3 volts.
the ground rail on the bottom shows zero volts where it connects, but further down the line in the circuit, its showing that 3 volts again.

i realize i know nothing, but.... this seems kinda really weird to me

everywhere circled in this pic is reading that 3 volts instead of ground.. and of course, the ground at the jacks is reading... you guessed it.... zero volts.  i am gonna try reading it from the power jack - to the chassis ground, but i expect to find the same.






so... am i gonna blow something up?

should i change that "ground" to actual, well, ground? and use the lower voltage to drive pin 4 on both of the chips?

i am lost... if someone could shine a light and grace me with some intellegence my stupid arse can handle, i'd be grateful!!

[bluebunny]

You been smoking some dodgy electrolytics again, Jimi?    That's not a voltage divider in the "power supply" section - the 47uF cap isn't a resistor and blocks any DC.  And the 10K resistor beneath it (that you subbed with a 27K) is shorted by the wire immediately to its left.  So it's not really there.  And the "3.19V" is wired directly to ground, so dunno how you got that reading. 

I'd vote for a loose connection.  Can't say whether that's on your board, or somewhere more, um, "organic"... 

[pinkjimiphoton]

dude. yeah, i am insane in the membrane, but this thing is on some serious blotter acid!! it works! and it sounds cool!! but...

yeah, i know. no way the cap and resistor can't be shorted. take either one out, it stops working.

originally, it was a voltage divider with two resistors and the cap coming off their junction to ground. well... it don't work wired any other way.

and i'm getting the same voltage reading at almost every "ground" point... 3.19v. the only place that shows up is freekin at the junction of that cap and resistor to ground that should be shorting each other. the resistor has to go to the - rail on the other side of the cap. that negative rail is supposed to be ground for everything, but its got 3 volts on it. where the hell can that three volts be coming from, and why the hell does this thing work at all?

as near as i can figure, its using the incoming audio to bias itself somehow... that 47 u cap is bleeding audio i think from the postive rail and dumping it on the negative rail. this thing was supposed to be run with two 9volts in series, and "ground" at the center tap between them.

no magick smoke involved this time

this is the original schematic





and the layout. now .... the bottom rail and the half voltage are jumpered together

i made the + red, and the - black on this one to show where the traces go... you can see that minus and half supply are jumpered together




this one i drew in where i added them two components. again, originally, i had a 10k/10k voltage divider there with the cap between the middle and ground. that won't fire at all. this is the only way it works. i also added in the jumpers, you can see that ground is connected to pins 4 of the opamps... they read zero volts.
but everything else reads 3.20 roughly, the only place on the power supply i get that reading is at that node..

take out the resistor? stops working. jumper the resistor? stops working. take out the cap? stops working. disconnect the cap's - pole from "ground"?

stops working. seriously. i can't even make this crap up!!





i do not for the life of me have any idea how the hell this can be happening... i know that u1 can be run "below" the ground rail, but... that concept is alien to my limited knowledge, too.

at this point i'd be willing to ship it to someone else to confirm this weirdness!!!

[pinkjimiphoton]

that 47u cap blocks dc, for sure... but its gonna let the AC thru... and thats where the audio is. a smaller value resistor than 27k makes for a very weird envelope, almost like a gating effect as it dumps i assume the audio to the negative rail somehow. i think its using this signal leakage to ring modulate? smaller r's make the sustain on the skinny e minimal, above 33k the tone starts to get a feeble dirt to it instead of a robust muffy dirt i have right now

way over my paygrade!!

i will draw up the most recent version with voltages. this is really really weird!

[pinkjimiphoton]

this is what i have at the moment. every triangle with a circle is reading that impossible 3 volt supply off the bottom of that freakin 47u cap.
pin 4 of both chips reads 0v. that i expect. but all the other places that should be "ground" read 3 volts. something weird going on i just can't seem to figure out!

[EBK]

Does your multimeter need fresh batteries?

[pinkjimiphoton]

its possible. i'm outta fresh 9v, and in the middle of an ice storm right now. i will check with one of my other meters, but i expect i'm gonna find the same results, eric

[pinkjimiphoton]

i am such a dumbfugginidyot!!

so like, i been messing with this and messing with it, going batshit crazy trying to figure out what's wrong, right?

@#$% @#$% @#$% @#$%ity @#$% me!

i had the pots all on the same ground bus
and had that grounded to the "ground" on the board
which was actually that couple volts offset on the low side of that cap.

once i disconnected it and reconnected it to the star ground on the input jack, the fuzz was back in business, and better than it was cuz no more freekin dc.

i gotta do some more stuff to it, but thats what i did. i had insulation over it and didn't even realize what i did one of them bleary eyed 3am "time to solder" after a nite rabble rousing at the local pub n snout. nice.

so so far, all's well. the power supply is still wack, but it works, sounds good, and no dc. i'll take it as a win.


oh....
and i was thinking... just something to ponder...

if a simple wire link is considered to be one ohm for all intents in some circuits

having that cap hooked up like that... i THImK... may work like this...

two resistances in parallel... 27k, and 1r...
well in math, anything divided by 1 is itself, right?
so 27000 divided by 1 = 27000, right?

so it can't be a short in this case, even tho it SHOULD be, right?

the - side of the cap is usually the outer foil shield.. if ya use electros for coupling caps, and touch the top of them, sometimes they'll buzz or hum cuz the shielding on the outside isn't connected to ground. but the shielding IS connected to the - lead of the cap.

so effectively, we're grounding the shield of the cap, and floating it above the negative rail with the 27000 ohm resistor in parallel with the 1 ohm one.

does that make any sense at all?
i DO remember seeing something like this done at black cat on their vibes to keep hum and noise down.. i was like... weird..

this may be a similar application, and man, i'd love folks thoughts on that one.

[EBK]

two resistances in parallel... 27k, and 1r...
well in math, anything divided by 1 is itself, right?
so 27000 divided by 1 = 27000, right?

so it can't be a short in this case, even tho it SHOULD be, right?

1/((1/27000)+(1/1))≈1

[pinkjimiphoton]

1 goes into 27000 27000 times bro

1x27000 is still 27000.

1/27000th is 1 but i'm not saying 1 27000, i'm saying 1 x 27000 is 27000. 1 goes into 27000 27000 times, right?  either way, the number ends up the same, right?   

https://www.google.com/search?client=firefox-b-1&q=27000+divided+by+1+%3D




its the only way this circuit can be explained i can think of, and the math seems to agree....

i think i'm healed up enough lol 

provided its not too frigid in the dungeon later, i'll supply video proof of it

i can @#$% shit up, but i can't change the physical laws of the universe!!!!.... yet....

[GGBB]

Multiple resistances in parallel look electrically like a single resistance who's value is less than the lowest resistance of the bunch. So 1 ohm in parallel with 27k looks like less than 1 ohm. So here the precise math is unimportant. But for the record, always work out stuff inside brackets before stuff around it.

1/1 = 1
1/27000 = 0.000037
1+0.000037 = 1.000037
1/1.000037 = 0.999963

[pinkjimiphoton]

pinky don' be knowin bouts no ones or twos or kemzars of resistance

still trying to wrap my mind around how the heck this thing is working.

the negative lead of the cap is most definitely going to the ground plane, "between" pos and neg. this plain reads around 3 volts now.
the positive side of the cap <now 100u> goes to the b+
the negative side goes to the "half" voltage ground plane between the two power rails
a resistor comes off that mid point on the cathode/ground side of the cap and goes to the actual -9v ground coming from the star ground.
at the junction of the resistor and cap cathode, is the 3 volt voltage. its not "ground" its something floating between the rails. i think.

so we got -9v, or ground
we got /+9v or vb
and we got "virtual ground" which is offset from "actual ground <-9 v>" by about 3 volts... close to half voltage.

before the star ground connected -9 v/ground to the board.
i had mistakenly thought the mid voltage was ground, not a mid voltage... hey, everything "grounded" to it, right?

but everything SHOULDN't have .... when i moved the ground point for the three volume pots to the input jack star ground instead of where it had been, it worked. when investigating that, i found that a little blob of solder on the second part of the fuzz/oct footswitch was actually shorting to the ground lead for the led's... thats what was letting the dc into the output signal.. again, my bad and a cheap mistake

so in reality, -9v is ground. +9v is vb. the mid point, leaking off between ground and the cathode of that cap is where the 3v is coming from. its gotta be.

sorry, just trying to wrap my head around what was making this thing tick... now it sounds pretty good, and seems to be working. i'll take that as a win, and hopefully get some video of it soon too.

so, does that make sense? i think me mistaking the midpoint for ground is what @#$%ed me up.

of course, the bloody @#$%ing ground IS connected directly to this midpoint as well, by a jumper, still makes no sense... but... hey...

lol
i need a week on a mountaintop with some mycelia-ey goodness methinks

[pinkjimiphoton]

Multiple resistances in parallel look electrically like a single resistance who's value is less than the lowest resistance of the bunch. So 1 ohm in parallel with 27k looks like less than 1 ohm. So here the precise math is unimportant. But for the record, always work out stuff inside brackets before stuff around it.

1/1 = 1
1/27000 = 0.000037
1+0.000037 = 1.000037
1/1.000037 = 0.999963

i dunno how to do that stuff. i went to windham public schools

but i do know how to use this.,...

https://www.diystompboxes.com/analogalchemy/emh/emh.html

thanks gord!!

i was wrong, thought that resistances would work like math does. one more particle of physics less in my universe lol

hey, <in stan marsh from south park voice>
i learned somethings today

maybe

[EBK]

i dunno how to do that stuff. i went to windham public schools

That's ok, Jimi.  You keep doing fuzz, and I'll keep being impressed.  I spent six years formally studying electrical engineering.  Fuzz was never part of the curriculum, so you know way more than I do in that department.   

[anotherjim]

You may have heard about a wire being worth about 1ohm, but probably as an engineering safety rule of thumb in some jobs. When we're working with audio and dealing with most resistors being in K ohms and current at a few mA and often much less, the resistance of wire is effectively zero.

When really high power is concerned, the minute resistance of wire can be enough for it to heat up when large current flows and that can raise its resistance. In practice, it's usually in connections & contacts in a circuit that's weaker than the wires and where you're most likely to get a "hot spot". If the current is high enough, it's sometimes necessary to fill the cable with cooling oil and pump is around to remove the heat.

Up at radio frequency, the resistance of wire gets dwarfed by impedance due to self-inductance and something called skin effect - so that's another case where we can't assume wire is zero ohms. Again, in audio work we don't have any of that to worry about..

You might use a 1ohm or other suitably low-value resistor to solve a ground loop noise problem. The trick is to make one or more ground return paths less favourable by deliberately raising the resistance a little. Then a noisy ground return current will mostly take it's proper path back because the milliohm resistance of the wire is less than the 1 or 10ohm or whatever resistor we put in the other return path it was taking. This often works really well, but must be done wisely. Any electrical safety "protective ground" must not be defeated.

And, an unbalanced balanced supply isn't necessarily wrong. I've a suspicion the usual resistor divider ground might have too much DC resistance and if it's battery powered, we can't lower the resistors too much or that battery won't last long. Probably an active virtual ground with an opamp, supply splitter or voltage regulator would do - but if the circuit is working and finding it's own balance point then ok ... and the sound character might just be depending on that -  I really wouldn't like to bet either way!

[pinkjimiphoton]

thanks for the explanations guys.

i really know next to nothing, but i do have the time and attention span to sit there for days swapping components around playing the same lick over and over and listening for differences.

yeah, i think they call it OCD now lol

fuzz is a whole different world of unlinearities and breaking stuff to make kewl sounds. i can make a fuzz out of almost anything.

try and get something to pass clean signal? or a damn phlanger or something? fuhgeddabouddit.

but gimme something where rules don't matter... and i'll give ya some #fuzzygoodness

lol

[pinkjimiphoton]

okie dokie.
this is actually IT. it works. it actually sounds decent! and yes, its weird!

[GGBB]

okie dokie.
this is actually IT. it works. it actually sounds decent! and yes, its weird!

Makes a lot more sense with the small changes. The 27k is not paralleled/shorted, and the 3.2V is VCC coming from the path around Q1 from +9V then to ground through the 27k. At least that's how I read it.

[pinkjimiphoton]

yeah, i really f'd it up before.. the values were ok, but the connections weren't right. this should be ok now.
now that i got the proper connections sorted out, i could probably get away with a voltage divider there, but i am sick of messing with it.

so my stash wasn't as good as we all thought ..... lol

thanks for helping me, and tolerating me, too lol