FLANGER (weird hiss issue)

[deadastronaut]

hi guys ive built this flanger ( based on the john hollis) but very modded , different lfo etc...
and works great...

i had it on breadboard as in the schemo and it behaved perfectly....and now have it built on pcb (unboxed)

and works as it did on breadboard......

however i am having a hiss issue when the sweep and manual pots are CCW...(which i didnt notice whilst breading)

i dont see where this hiss would be coming from. (LFO?) or how to get rid of it....hmmmm

any help much appreciated...cheers guys...rob.

[11-90-an]

 Maybe some parasitic capacitance, cold solder joints, etc.?

Maybe boxing it would solve the issue...?

I'll just say it: pictures! 

[Scruffie]

Have you actually measured the clock frequency? It might just be natural BBD hiss, usually flangers have a clock trim because, well, it varies from build to build and I'm assuming you aren't using the exact same components you breadboarded it with.

[deadastronaut]

hi scruffie, yeah im using the same components, i bag them up of the breadboard while i await the pcbs...
(i always do that now to ensure the same sound)

when both sweep and manual are fully CW no hiss...but turning them CCW introduces hiss big time..

hmmmm

[Scruffie]

Well, I didn't expect you to say that  Did you even use the same 4046? They vary quite a lot from chip to chip. How about the pots? Any change in what was used to breadboard could affect things.

Is CW longest delay or shortest delay? If you're going to attempt to design a flanger, you really should get used to measuring clock frequency.

It could be a whole host of different issues, some voltages pre and post noise and clock frequency measurement would be a start on debugging it.

Do you have a scope?

If I had to guess it sounds like the 4046 is probably latching up and dropping out, I hate that chip...

[11-90-an]

Maybe try increasing R19 value?

And maybe put a like, a 680Ω resistor from CCW of the "sweep" pot to ground...

On second thought, maybe try isolating LFO and cd4046 gnd with a 100Ω resistor?

[deadastronaut]

Well, I didn't expect you to say that  Did you even use the same 4046? They vary quite a lot from chip to chip. How about the pots? Any change in what was used to breadboard could affect things.

Is CW longest delay or shortest delay? If you're going to attempt to design a flanger, you really should get used to measuring clock frequency.

It could be a whole host of different issues, some voltages pre and post noise and clock frequency measurement would be a start on debugging it.

Do you have a scope?

lol... yep i used the same chips. pots etc.....

ccw is the shortest, (least wobbly lol but most hissy) ive never really messed with flangers before so these are a new ball game to me...i went with the hollis topology, but read that the lfo introduced ticking etc, so put in my own lfo...
and a few other mods/tweaks.....

nope , havent got a scope.... 

it seems like i need to narrow the range of the the 2 pots, as when they are in the middle to fully CW they seem fine....hmmmm....

[Rob Strand]

Is there a bypass cap missing off the +9v rail?  Maybe C14 in the wrong place?

[deadastronaut]

hi rob, C14 is there....  you mean add another 100uf cap to the right of R29?.

as i say its only when the 2 pots sweep, and manual are turned CCW....

CW is fine...strange...

[Scruffie]

The LFO did introduce ticking but not because of its fundamental design, just the short cuts the circuit took to keep parts count low.

Well as I said, I can only guess it's the 4046 latching without any data (the breadboard might have bought you some handy resistance/capacitance your PCB lacks) so I'd suggest measuring the voltage to pin 9 and seeing at what point it delves in to hiss (check the datasheet for info on the linearity of the internal VCO) then limiting that somehow.

Edit: Rob does raise a good point though, a bypass cap directly on the clock power line is always a good idea for stability, a ceramic 100-220nF should do it.

[deadastronaut]

cheers guys, i tried a 100nf cap on the 4046 (with my cap sub box) no different ..

but here goes with voltages for these 2 pots...

probe on 4046 PIN 9.

sweep CCW 0.0v ( with manual set at CW)  SWEEP cw 3.7V.....hiss goes around 1.3v when turning

manual CCW 1.43v (with sweep set at CW) 3.78v........no hiss when turning....

[Rob Strand]

I suspect the 100uF C14 should go across +9V not +9V(T).  That way the big cap is bypassing the 9V supply to the IC's.

You can see the difference on the JH schem,

[deadastronaut]

cheers rob so tack a 100uf after my R29 then....(i can still leave that other 100uf cap there though right?.)

[anotherjim]

When you have a series protection diode -  you simply must put plenty of bypass capacitance after it on the circuit power (and the 47r isn't doing much unless to work with the supply capacitors as a filter). The series diode isn't bad itself but it does more harm than good if there isn't bypass capacitance to pass AC currents both ways.

[deadastronaut]

cheers i know what i did, i had a 100uf across my breadboard rails as always, and drew it as i saw it..

....well spotted.


edit: sorted cap....still hiss....

[Rob Strand]

edit: sorted cap....still hiss....

Oh man.

It would be wise to check the voltage on the +9V rail with the manual pot at both the CW and CCW.

The way you have the manual pot is drawn on the schem CCW is when the pot is shorted out.   That should be when the clock is the slowest.   The way JH has set-up the manual control it reduces the amount of sweep when the clock is slowest.

That would make more sense as far as noise goes.    If the clock frequency is too low you will get the "natural" hiss of the BBD, as Scruffie mentioned.     The way around that is to prevent the clock going down to such a low frequency with the manual pot set to the extreme.     One way is to make R39 a bit bigger.  Another is to tweak he timing resistors on the 4046.

You might be able to use a DMM to measure frequency if your meter does frequency.

[deadastronaut]

cheers rob, you mean R19 i take it, the resistor on the manual pot to ground....raise the value of that. ?

( i'll stick a resistor between the pot and wire that goes to the 33k to (add value) test.)


ill check voltage on the rails with pots too....

[Scruffie]

cheers guys, i tried a 100nf cap on the 4046 (with my cap sub box) no different ..

but here goes with voltages for these 2 pots...

probe on 4046 PIN 9.

sweep CCW 0.0v ( with manual set at CW)  SWEEP cw 3.7V.....hiss goes around 1.3v when turning

manual CCW 1.43v (with sweep set at CW) 3.78v........no hiss when turning....

So if I'm reading this right, between 1.43v and 1.3V there is an area where it suddenly enters in to hiss territory, correct? So all you need to do is find that precise point and limit it.

I still just think it's the 4046 having a hissy and putting noise in to the BBD, but I'm biased against that chip.

[deadastronaut]

cheers guys, i tried a 100nf cap on the 4046 (with my cap sub box) no different ..

but here goes with voltages for these 2 pots...

probe on 4046 PIN 9.

sweep CCW 0.0v ( with manual set at CW)  SWEEP cw 3.7V.....hiss goes around 1.3v when turning

manual CCW 1.43v (with sweep set at CW) 3.78v........no hiss when turning....

So if I'm reading this right, between 1.43v and 1.3V there is an area where it suddenly enters in to hiss territory, correct? So all you need to do is find that precise point and limit it.

I still just think it's the 4046 having a hissy and putting noise in to the BBD, but I'm biased against that chip.

stop having a go at that poor little 4046 you bully lol....

yes correct: how would i 'limit' it...?

[anotherjim]

The 4046 VCO cuts off when pin9 is below about 1V. JH lifted the LFO ground with the diodes/Zener and stopped it getting pulled too low with those backward diodes. When the manual pot is minimum, the fixed 47k should stop that pulling pin9 towards 0v as there is still the 3v DC lift from the LFO circuit. With the LFO at 0v ground (no lift), there's nothing to stop the sweep to pin9 going below 1v.