FLANGER (weird hiss issue)

[anotherjim]

Got any 100nF ceramics on those chip power pins? Especially the 4049. I'm honestly very suspicious of that parallel gate drive idea. I know what it's supposed to do and sure it'll do it, but all those inverters swinging in time means there are large supply currents passing through during the short time the logic levels are in between on & off.

[antonis]

Could we shake hands on an inherently dreadful design originated by a notorious lazy designer..?? 

[Rob Strand]

I'm not 100% sure how you wired the diode but if you wire it back to the Vref used for the audio the diode current can load Vref and cause clicks.    If Vref shifts due loading it could be stuffing up the biasing on the BBD and causing distortion.

Bypass caps on Vref only work at audio frequencies.  At LFO frequencies any Vref loading will cause a change of voltage just like changing the Vref resistors.   A big bypass cap will sag over the LFO period.

The JH design used a zener which is more immune to loading and also it used a separate Vref for the LFO and the Audio.

[DrAlx]

Astro, the following ain't a fix for your problem but a handy trick.
You may not have a scope but you can get the clock frequencies of the flanger sweep another way.
If you make the flanger do the jet plane thing (very slow sweep, strings open with lots of distortion or fuzz before the flanger, single strum) then the "pitch" of the jet sound at each point of the sweep will be the inverse of the BBD delay, and clock frequency is just (512 / delay).
If you can record that jet sweep I can get an estimate of the min/max clock frequencies.
A recording of the noise/hiss you describe in this thread might be handy too. Can't tell if the hiss you describe is regular BBD noise.

[deadastronaut]

hi heres the clean with ''crackle''

https://www.dropbox.com/s/840dxsygeqiv5xd/flncrackle.wav?dl=0

and here is the dirt flange slow sweep..

https://www.dropbox.com/s/19tal72xk05pnur/dirtflange1.wav?dl=0

[anotherjim]

The crackle seems to happen when the audio is strong enough to interact with the LFO and clock. I suspect common ground currents and breadboard stray capacitance.

The dirt flange sounds nice -  I wouldn't kick it out of bed. Nice ghost flange on the tail.

[deadastronaut]

yeah its a really nice flanger, especially on the longer sweep, and the manual filter setting...choose the sweet spot etc....

its not on breadboard btw its on a pcb, so have had to hack it.....

im figuring i'll have to rebread it..

[DrAlx]

The pitch of the jet sweep is going from just over C#4 to G4.
That's like sliding on the B string from 2nd fret to 8th fret and back.
Those notes have frequencies of 277Hz and 392Hz.
Multiplying by 512 to get BBD clock (as measured at one of its clock pins) gives 142kHz to 200kHz.

That is way less than what the ultraflanger is capable of (according to the clock rates on JHs sketch).
You should be able to get several octaves worth of sweep.
e.g. by comparison the widest jet sweep on a original Electric Mistress goes from just below the open bottom E-string
on the guitar to around the 18th fret on the top E-string.

[11-90-an]

im figuring i'll have to rebread it..

Wait wait wait... before you rebread it, we need to see the pcb first! It just might be some bad design or "negligible" mistake...

[antonis]

Ahaaaa...!!

So you dare to insult Rob for sloppy/rushed board design, don't you..??

[DrAlx]

Very odd noise.  My guess is it can be coming from one of two places.

First thing could be as mentioned here is that the BBD bias is getting messed with somehow.
So I would try shift it slightly and see if it makes things better or worse.
Try tacking 100k in parallel to the 10k resistor that sets the bias and see if it makes it better or worse.
If worse, then try putting the 100k in parallel with the 12k resistor instead of the 10k.
This is the sort of thing I do without soldering.  I just try and hold the resistor in place with my fingers.

Other thing could be noise on the clock lines themselves, but this is unlikely 'cos of the hex buffer.
I would audio probe the clock lines themselves (careful not to blow your amp or your ears) and
see if you hear the same sort of fizz on them.  Like I said, unlikely, but I have had AM noise on clock lines affect a BBD before.
If you do hear the same kind of fizz on the clock lines and it is periodic like you have on the audio
then that would explain the problem.  Not sure how you would fix it though.

[deadastronaut]

how dare you....  lol i sound like retard thunberg

the pcb is exactly as the schemo...no different.  all drc checked, erc checked... no issues.

ands works as it should....apart from these f......g noise issues lol...


alex: so im missing a fair bit of range then hmmmm......., like i said, as i had it before it was a much stronger flange .

i'll give that a whirl on the bias...if that doesnt sort it

i think i'll just pull the chips off, (thank fk for sockets) and rebread it...and go from scratch again. 

i'll be pissed if it works perfectly again on bread though...

[DrAlx]

One other thing.  In the sound file showing the crackle, I couldn't hear any any sort of flange effect.
So even in the portions without crackle, I can't tell how much of the sound is from the wet (BBD) path
and how much from the dry path.
Any chance you can audio probe at the BBD output and get just the wet?

[deadastronaut]

yep, i can do that....back i n a bit...

[11-90-an]

Ahaaaa...!!

So you dare to insult Rob for sloppy/rushed board design, don't you..??

Gah! Not exactly, but... um... no.... geh...

i'll be pissed if it works perfectly again on bread though...

But isn't that what breadboards are for...? 

[deadastronaut]

back again to continue the 78 page thread 

well i probed the out on the mn3007, and it was pretty clean....hardly any flange, bit of chorus, and lfo wobble...but not a lot, but not distorted....hmmmmm

[DrAlx]

So I guess it will still be noisy if you pull the BBD out.
So maybe do that, and check.
If it's still noisy, then given that the hex buffer is of no use when there is no BBD in the board,
pull the buffer chip too and check again.
If still noisy, then I would start audio probing what's left.

[deadastronaut]

it only introduces distortion on certain settings on the man/sweep pots...

when its in full flange setting its nice and clear.....

but yep, i'll try that too...

edit: its really the sweep pot that introduces distortion, the manual, doesnt at any setting......hmmm

[DrAlx]

Hollis sketch has 2k7 on 4046 pin 11 rather than 2k2.  Don't know what effect it has.

[anotherjim]

I'm wondering if the LFO Schmitt trigger opamp output isn't clipping - seems counter-intuitive to worry about that in something making square waves, but when the output of an opamp is told to swing further than it should, something gives. R23 might be too large. If it's made smaller, it will change the slope direction sooner and reduce the output swing. You might tack a 1M pot (set at max) across R23 and see if that, or lowering the 1M pot resistance helps.

I'm not sure the Manual can work as I'd expect it to - that is without the LFO, you can tune the BBD delay to act as a comb filter. To do that without LFO, Manual would need to change the pot lug 3 connection to some +supply voltage, so it can inject its own control to the 4046 pin9 on top of any swing from the Sweep pot.