FLANGER (weird hiss issue)

[deadastronaut]

cheers, yeah man, ive tried those, and tweaking those endlessly....

i just tried running the 4046 on a 5v reg....as i read someone else did, but no joy....

im guessing this design is a no go in the end really, it works yes, but useable no chance unless you like the sound of a rainstorm and ticking......

but hey, ive learned a few things along the way, so some good came out of it i guess.....

i think i'll take a look at some other flanger designs that don't use the 4046....and dont have such a fussy lfo vb too...so the flanger quest continues....

thats me done with this...shame as it is a lovely flange effect, but hey, onwards and upwards....

thanks guys. 


any other 9v flangers to recommend?. to tinker with?...

[Scruffie]

shame as it is a lovely flange effect

Keep the audio path as you have and like it, just stick a new clock on it, it'll still sound lovely I'm sure.

any other 9v flangers to recommend?. to tinker with?...

There's the 9V mistress which is the classic of course, BOSS BF-2 (you needn't use an MN3102, a 4013 can be switched in) and most of the 18V ones will probably run okay at 9V anyway, you just wont get quite as wide a sweep.

[deadastronaut]

hi scruffie, cheers man,

so if i dropped in a 4013 instead of the 4046 like this?   (i still have it on breadboard just in case) 

https://pcbguitarmania.com/wp-content/uploads/2018/05/Electric-Lover-Schematic.pdf

[anotherjim]

I wonder if the 4046 problem isn't because the two clock phases come from different paths. One from the VCO and one inverted copy from the XOR phase comparator. The clock from the XOR will switch later than the one from the VCO. This could lead to a lot of clock noise getting through simply because the clocks are not mirror images and won't cancel out when the two BBD signal output phases are balanced.
Using a 4013 to make the two clock phases ought to give much better matching of the phases and better cancellation of the noise.
Haven't tried it myself, but I suspect using both 4013 flip-flops connected in parallel will improve the clock drive and not need the 4049 buffers. By parallel I mean as simply as every pin to same name pin - CLK to CLK, D to D, Q to Q etc.

[Scruffie]

hi scruffie, cheers man,

so if i dropped in a 4013 instead of the 4046 like this?   (i still have it on breadboard just in case) 

https://pcbguitarmania.com/wp-content/uploads/2018/05/Electric-Lover-Schematic.pdf

Yes and you can parallel the two 4013 halves as Jim mentions and forget about the 4049.

[deadastronaut]

ahhh cool....cheers guys, i'll order up a 4013 then,

then pick your brains a bit more on the actual physical wiring to be sure when it turns up....

so the flanger might go to the ball after all yay.....

[deadastronaut]

hi guys, well i got my 4013's..

so the connections are in parallel as in the picture correct?....

however im not sure how the connections now translate from the lfo to the 3007....

as in pic..

[anotherjim]

Yeh. I don't think S & R inputs need be used so all 4 will be tied to 0v and don't have to route across to each other.
The pair will be clocked from 4046 pin 4 and /Q's connected to D's. 4046 pins 2, 3 and 14 not used. Q's are BBD clock phase1 and /Q's phase2.
This way - using a 4013 - will halve the clock frequency. Might need to tweak 4046 timing to go faster.

[deadastronaut]

oh i thought using this would get rid of the troublesome 4046, and the 4049. 

ive obviously misunderstood...hmmm....

[anotherjim]

Ok, the 4013 is to get 2 equal 50% duty cycle opposite phase BBD clocks with no overlaps = smarter handovers by the bucket brigade = fewer spills and arguments. The 4013 itself can't easily make a VCO, it has to be clocked by something else and can only make the 2 BBD clocks wired as a frequency divider.
The 4046 VCO pin 4 does already run at 50% duty, but making the other clock by inverting pin4 with the 4046 XOR gate delays it so its pin2 output is switching a little later than the pin4 output. Adding 4049 buffers can't make up for that.

Does this mean that Rick Astley won't sing?

[deadastronaut]

Does this mean that Rick Astley won't sing?

lets hope so....

so i am keep the 4046 and the 4049 then?.

[Scruffie]

You can just use the 4013 which is what I'm suggesting you do but you either a) need the LM311 portion from the mistress circuit or b) you can use halve of the 4013 for a VCO with a handful of transistors and half as the divider but that will limit you to about 800kHz-ish for a top end, but that's not too bad.

But you probably can't just stick your LFO & Manual control on to it and have it work.

See how the 4013 is connected here?

[deadastronaut]

aha ok,  im going to go with the lm311 then...

as in brians (madbean) current lover type setup....but with my mix etc...

should be a lot smaller foot print in the end too without the poxy 4046/4049 ...

rick astley just turned up....i told him where to go...

[Scruffie]

Good choice 

Back to Jim's 4046 comments though quickly for prosterity about using it to provide both clock phases causing issues, you could just use a single output and use the 4049 to buffer and invert the phase for the other half rather than involving a 4013.

[anotherjim]

I was going to suggest try an inverter to make the opposite phase. No 4046 XOR and no 4013 involved. I think we've been around this one elsewhere in this parish sometime ago. You can't make precise perfect opposite phases with an inverter -  it will add some edge delay to its output, but it has to be much less than the XOR in the 4046.
Temporarily, use 2 inverters. One fed from 4046 pin4 and outputting BBD phase1 and one fed from phase1 and feeding phase2. Feeding both of the BBD clocks from inverter outputs will give them the same drive.

[deadastronaut]

hi guys, back again....been busy.

got my 4013, and 311......and got it breaded up, and whooohooooo no hiss. or ticking etc....fantastic.

and sounds excellent.....thanks guys..

here is what i have on bread at the mo....sounds great to my ears....

does frank marino, gilmour, halen, very nicely.....been having a blast with this...

thoughts?.

the only thing that has little effect is the 'range' pot.....but its cool anyway..

i am using the 'clock/time as a pot too...great , thick warbly flange to thin, nice....

and the lfo isnt fussy either...

[DrAlx]

Range pot wiring is wrong.  It should be wired to scale the control voltage applied to the VCO, not scale the current into it.

For this VCO, you get a BBD delay that varies linearly with the control voltage applied to pin 2 of the LM311.
The bigger the CV, the bigger the BBD delay.

IIRC, the CV must be bigger than about 0.8V or else the VCO it will freeze.  (The VCO can only go so fast).

So you should make your range control circuitry scale the size of the triangle wave coming out of your LFO, but make sure the resulting triangle wave is always above 0.8V.
Take a look at the EM3207 or current lover schematic for one way of doing that.
It takes the triangle wave out of the LFO (centred somewhere near 4.5V).
Then has a voltage divider to scale that triangle wave down (so it is smaller wave, and centred some way below 4.5 V).
Then uses op-amp gain stage to scale up the triangle wave according to range pot setting.

A simpler option would be to pick LFO resistors (R22, R25) so you get a big triangle wave  out of the LFO (lets say triangle wave out of the LFO sweeps from 1V to 8V).  Then wire a RANGE pot so that this maximum size triangle wave is at one end pin of the range pot.   A fixed 0.9V is at the other end pin of the range pot.  The pot wiper then gives you a triangle wave that can go anywhere from maximum amplitude, down to zero amplitude and it will always be bigger than 0.9V.


Also, the 22pF clock cap works but gives a non-linearity in the VCO characteristic for low control voltages due to diode reverse current through D1 (explained at length on the EM3207 thread).  47pF is a better choice.

[deadastronaut]

whoops..the range pot was meant to be ''sweep''  like the sweep control in the original john hollis.

but i just couldnt get it to work the same... sorry for the confusion. 

the 'time/clock' gives me all the range i need, thin to wide flange....

ive actually put a 68k across the time/clock pot to limit it, as it got way to warbly..and also re-introduced hiss again...so...

i have swapped the 22pf for a 47pf though.... 



EDIT. i used the 1MA pot a LFO depth control..... sorted. 

[deadastronaut]

hi guys, back again....well ive built 2 of these and they were fine, no issues. its a cool flanger
and does what i need it too.

however, i just built another on pcb, and it has breakup distortion under, and along with  the flanger sound.

ive double checked all my values, and am using the same chips etc, but im still getting this distortion along with the flanger.

it pissed me off so much i pulled the socketed chips off the pcb and re breadboarded the whole thing again to try and nail the issue....and sure enogh it is the same as it was when on the pcb.

here is the final schemo which i have built twice successfully.

where should i be looking for the cause and cure? seems strange as its been fine before.....

i bought all different chips again too to try, to no avail.

[anotherjim]

I can't help thinking that Vb needs to be adjustable to find the best spot to get the most headroom through the BBD.
A 10k trimmer in between to fixed 10k should let you try anywhere between 3v & 6v.