supposed one-shot generator does two shots

[11-90-an]

So I'm making this switching system, for a (secret) project, but when I sim it, it seems to work, only outputs 1 shot(or you could say pulse), but then on certain occasions, it seems to output 2 shots... (note that I breadboarded this, and it seems to have the same problem. sometimes 1 shot, sometimes 2 shots)

A more in depth explanation, basically this *supposed* to be a rising edge detector (1 shot) made out of discrete components (transistors, no IC's..). What it's *supposed* to do is that when the momentary switch is clicked, no matter how long, it would output a "high"(around 9v here), for a specified amount of time.

The problem here is that on random occasions, it would suddenly become a rising AND falling edge detector. it does the "click switch for a timed pulse", but the problem is it also does a "if you hold the switch long enough then release it would also output another pulse" which would be a really big problem

Now I also took an output from the opposite side of the circuit (the positive leg of the capacitor). This basically is an inversion of the the output above ^^^
(without switch clicked, outputs high, and when switch clicked, outputs low for same amount of time above ^^^)

I wonder if this is a problem with the circuit itself?



here's the falstad code...

Code Select Expand

$ 1 0.000005 18.278915558614752 1 5 43
r -192 464 -240 464 0 10000
O -176 304 -176 240 1 0
w -176 304 -144 304 0
s -192 464 -192 208 0 1 true
g -240 464 -240 480 0 0
g -256 272 -256 288 0 0
w -256 208 -192 208 0
v -256 272 -256 208 0 0 40 9 0 0 0.5
w -144 208 -96 208 0
w -192 208 -144 208 0
r 16 208 16 272 0 1000
r -144 272 -144 208 0 1000
w -32 336 -16 336 0
w -112 336 -96 336 0
w -32 304 -96 336 0
w -96 304 -32 336 0
w -144 304 -144 320 0
w 16 304 16 320 0
w 16 272 16 304 0
w -144 272 -144 304 0
r -32 304 16 304 0 10000
g 16 352 16 416 0 0
g -144 352 -144 416 0 0
t -16 336 16 336 0 1 0.5984195123385558 0.6579909244253883 100
t -112 336 -144 336 0 1 -8.938714354874618 0.05957140990569687 100
w -96 208 16 208 0
r -96 208 -96 304 0 6800
209 -144 304 -96 304 0 0.00001 8.340294840354927 0.001 1
O 16 304 64 304 1 0
c -96 336 -96 464 0 1e-10 0.05957140681430993 0.001
r -192 464 -96 464 0 1
o 1 512 0 4102 10 0.1 0 1
o 28 512 0 4102 10 0.1 0 1


[iainpunk]

now i'm curious about the secret project.

i copied your code and can't for the life to me get it to double shot.

these sort of circuits are good for really odd and cool fuzz pedals, kinda glitch-filter-bit-rate-reduction-ish sound, and doing some octave downs at higher input frequency's, but it gets really thin sounding really fast, when you are more than two octaves higher or lower than the 50% duty cycle frequency.
i prefer its 555-timer based cousin, that only sets a lower limit to the pulse time, not a higher limit.

cheers, Iain

[anotherjim]

I ran it through my grey matter simulator and it should work as intended.

[11-90-an]

i copied your code and can't for the life to me get it to double shot.

I ran it through my grey matter simulator and it should work as intended.

Aaand that is the problem...it seems to work, but after leaving that falstad tab open after around 8 hours, watching some youtube videos and online school, and I go back to it, it starts to 2 shot!   

What makes matters worse was that it was breadboarded... and it had the same problem of sometimes working reliably and double-shotting... this circuit is working against me! 

these sort of circuits are good for really odd and cool fuzz pedals, kinda glitch-filter-bit-rate-reduction-ish sound, and doing some octave downs at higher input frequency's, but it gets really thin sounding really fast, when you are more than two octaves higher or lower than the 50% duty cycle frequency.

Unfortunately this isn't really part of the audio part of the effect, but for JFET switching...

[anotherjim]

I think this is the classic Eccles-Jordan one-shot...

Give it a shot?

[Rob Strand]

The problem here is that on random occasions, it would suddenly become a rising AND falling edge detector. it does the "click switch for a timed pulse", but the problem is it also does a "if you hold the switch long enough then release it would also output another pulse" which would be a really big problem

It seems like a switch bounce problem more than a circuit problem.

The Boss footswitch circuit  (1M + 1K + 10n) is an example how to fix it but you might need to increase the cap to say 47n or 100n.

http://www.freeinfosociety.com/media/images/1936.jpg

Oh, you will need to flip the ground and +V connection because you want a positive (trigger) pulse (the Boss trigger is negative).

[11-90-an]

Thanks for all the replies, the switch debouncing proposed by Rob works...

Code Select Expand

$ 1 0.000005 27.727228452313398 1 5 43
O -240 176 -240 112 1 0
w -240 176 -208 176 0
s -272 160 -272 80 0 1 true
g -432 160 -432 176 0 0
w -432 80 -336 80 0
v -496 160 -496 80 0 0 40 9 0 0 0.5
w -208 80 -160 80 0
w -272 80 -208 80 0
r -48 80 -48 144 0 1500
r -208 144 -208 80 0 1500
w -96 208 -80 208 0
w -176 208 -160 208 0
w -96 176 -160 208 0
w -160 176 -96 208 0
w -208 176 -208 192 0
w -48 176 -48 192 0
w -48 144 -48 176 0
w -208 144 -208 176 0
r -96 176 -48 176 0 10000
g -48 224 -48 288 0 0
g -208 224 -208 288 0 0
t -80 208 -48 208 0 1 0.6027791372918947 0.6510864021956028 100
t -176 208 -208 208 0 1 -8.58115138690655 0.04830687874982175 100
w -160 80 -48 80 0
r -160 80 -160 176 0 6800
209 -208 176 -160 176 0 0.00001 7.978371863460769 0.001 1
O -48 176 0 176 1 0
c -160 208 -160 320 0 1e-10 0.04467377481985505 0.001
c -336 80 -336 160 0 1e-8 8.91343470665719 0.001
r -336 160 -272 160 0 390
r -336 160 -432 160 0 1000000
w -336 80 -272 80 0
209 -432 80 -432 160 0 0.00009999999999999999 8.917067802070976 1 1
w -432 160 -496 160 0
r -496 80 -432 80 0 10
O 400 192 464 192 1 0
c 208 208 288 208 0 4.7e-10 -7.565147011586387 0.001
c 48 208 128 208 0 4.7e-10 0.04796463319086848 0.001
w 48 176 48 208 0
w 128 176 128 208 0
w 288 176 288 208 0
w 208 176 208 208 0
g 208 112 192 112 0 0
g 128 112 144 112 0 0
r 208 112 208 176 0 56000
r 128 176 128 112 0 56000
w 128 384 208 384 0
c 208 240 208 384 0 1e-10 0.023225869748254482 0.001
c 128 240 128 384 0 1e-10 0.5917724140593125 0.001
g 288 256 288 288 0 0
g 48 256 48 288 0 0
w 48 80 288 80 0
r 48 176 128 176 0 100000
r 208 176 288 176 0 100000
w 48 208 48 224 0
t 128 240 48 240 0 1 0.5205819111201895 0.5954055099144056 100
w 208 176 128 240 0
w 128 176 208 240 0
t 208 240 288 240 0 1 -8.133693555897445 0.02685896560334759 100
w 288 208 288 224 0
r 288 80 288 176 0 10000
r 48 80 48 176 0 10000
w 48 80 -48 80 0
w -336 320 -336 384 0
w -336 160 -336 320 0
r -336 384 128 384 0 10
r -160 320 -336 320 0 220
w 288 176 320 176 0
d 400 192 400 256 2 default
t 480 256 512 256 0 1 -8.913240646786845 0.00363312633963831 100
g 512 352 512 384 0 0
r 512 272 512 352 0 5600
r 400 256 480 256 0 10000
162 512 80 512 240 2 default-led 1 0 0 0.005
S 400 192 320 192 0 0 false 0 2
w 320 208 320 384 0
w 288 80 512 80 0
r 320 384 208 384 0 22
x 323 170 367 173 4 13 latching
x 323 225 388 228 4 13 momentary
o 0 512 0 4102 10 0.1 0 1
o 26 512 0 4102 10 0.1 0 1
o 35 512 0 4098 10 0.1 1 1


Here's the circuit so far... I added a slightly changed Boss bistable flip-flop, LED indicator, and a momentary/latching option... the voltage readings there will all control individual JFETS, (with the 1M resistor, 47n cap, and reversed diode, of course)... do you guys see any impedance problems, etc.?

P.S. no guesses as to what this secret project would be? 

[anotherjim]

So the simulator simulated contact bounce? That's sophisticated!

One effect type that I suppose has to have soft bypass is a delay with tails.

[iainpunk]

i was gonna guess a reverb, because of trails, but i didn't expect him wanting to be asked what it is, since its a secret.
or maybe a ringmod, using the switch to switch between an oscillator and DC, or maybe a loop switcher with 17 loops.

cheers, Iain