FLANGER (weird hiss issue)

[deadastronaut]

Sod off lol...

[duck_arse]

Sod off lol...

be nice - he got you to page 5.

[11-90-an]

Nah, don't worry, astro, the "incompentence" was a joke...

However, i do like looking at pcbs as to see how far i am from even making one...

[deadastronaut]

hi guys, got my pcb designed.....

just to be sure, this latest schemo all looks good yes?.

https://www.diystompboxes.com/smfforum/index.php?topic=125122.msg1191286#msg1191286

[deadastronaut]

well, more pages coming.....

this fckng flanger, still hisses like a fkr......

built the 2nd pcb...after breading, re-breading, double checking....and all was fine on bread...or so i thought..

then get to the build..and voila..(with the same exact parts) and still hisses..(and has a weird sound like that of the aliens in the old war of the worlds film)....too.....arghhhhhhhh....

so....off to the breadboard i go again, and it works as planned, as per schemo....and still hisses like a mutha...the same as on new pcb....

wtf...

its such a swine as the flanger sound is awesome....especially with the mix level, nice and subtle to full on...

but with the hiss and weird noise, utterly unusable....

i found a modded schemo by dimebugg on here, and he added a 10n cap across pins 2/1 of the lfo with a 56k resistor that joins the lfo vb which he said cured any 'ticking' so tried that...no joy.

anyway im back on breadboard again now..so any suggestions (other than pee off) welcome...

circuit all the same as this schemo on page 4...

im using 4046be......is that correct or should it be the unbuffered version?.....or make no difference?.

https://www.diystompboxes.com/smfforum/index.php?topic=125122.msg1191286#msg1191286

[anotherjim]

There is no unbuffered 4046. But different brands are known to behave differently. Nearly everybody is selling the TI part. Can you find any other brand?

[deadastronaut]

cheers jim, so 40406/4046be are the same then, i thought one was buffered and the other wasnt, ok i'll find another one online then....i have my suspicion that its the 4046 thats causing it..

as suggested by scruffie on page 1.

is there an alternative before i order one up?.

[Scruffie]

as suggested by scruffie on page 1.

Well I don't want to say I told you so... alright, I do actually 

Seriously, buy one of these - https://www.ebay.co.uk/itm/VC97-Auto-Digital-Multimeter-Voltmeter-AC-DC-Current-Capacitance-Data-Hold-w-Bag/224162717328?hash=item3431239e90:g:deUAAOSwZM9fTamN

You can measure the clock frequency and actually get an idea of what is happening rather than trying to guess your way around. A flanger isn't a dirt pedal or even a chorus... the 4046 can work, it's fine for choruses, I just don't like any IC that variable for very high frequency tasks and its VCO can be awkward as I think you're discovering.

[deadastronaut]

i have a PG012 DMM which has HZ(trig) hi/lo on it...never had to use it. (bought from maplin years back)

im guessing thats what its for then lol.....

[bluebunny]

Not all 4046s are the same, as your namesake documents here.  Not sure if it's relevant in this application, but it might be worth trying some other brands, if you can lay your hands on them?  I've got a TI and an NXP over here if you want them.

[deadastronaut]

hi marc cheers man, i just bought another 2 so i'll try those....(will prob arrive tuesday/wednesday

if they still mess about i'll take you up on that...nice one. 

thanks for the article too...so there are differences between brands it seems....hmmmm.....

[anotherjim]

NXP I think are the former Philips parts. I think they get a bit wider linear range out of the VCO than TI, so it might be relevant. I think Thomas Henry's X4046 synth VCO has trouble with different brands and guess what, the best ones for it are no longer made. I think the Motorola 14046 is still made by On-semiconductor, but only SMD?

[ElectricDruid]

I also had trouble with the 4046. I think it's the most variable chip I've ever used. Two "identical" chips from he same manufacturer vary widely. Two "same number" chips from different manufacturers can have completely different specs. I needed one that could run to 8MHz. Some can go as high as 10. Some only manage 4. Some do less. It's a bloody minefield.
I found I had to fine-tune the resistor values in my circuit (https://electricdruid.net/adventures-in-top-octave-generation/) to suit the individual device - possible, but not exactly just a case of "solder the bits in and it works".

[DrAlx]

I used Audacity to inspect the flncrackle.wav that Rob uploaded.  The crackles look like short bursts of 2345 Hz ripple superimposed onto the signal (which is much lower frequency than that).  So it's not regular BBD noise (hiss), or distortion/clipping since the noise bears no relation to the input signal.

I don't know if the noise being talked about with the latest build is the same noise as in the wav file.

Rob, before you did your rebuild I thought you said you could not here the noise when you audio probed the BBD outputs.
I am wondering if that is still the case, and if you pull the BBD out of the board, do you still get the noise?

[deadastronaut]

hi alex, (on breadboard)  if i pull the bbd out, i get nice and clean signal....no hiss or crackle at all.


i just got my new 4046 swapped it over, and no difference still hiss n crackle....

[DrAlx]

Just looked at the 4046 datasheet.  One of the two clock signals for the BBD (let's call it CLOCK1) is taken from pins 3&4 of the 4046.  The other clock signal for the BBD (lets' call it CLOCK2) is produced by XORing CLOCK1 with the logic level set at pin 14 of the 4046 (which is currently held high in the schematic).  The XORing would also work if pin 14 was held low instead of high.  So just for the hell of it, I would try grounding pin14 instead of connecting it to the supply and see if that makes any difference.

[deadastronaut]

cheers man, nope tried that....no flange at all....but no hiss...just subtle ticking.

[anotherjim]

2345Hz? Isn't that big bang background noise?

[DrAlx]

cheers man, nope tried that....no flange at all....but no hiss...just subtle ticking.

Sorry man, my mistake.  I mentally got the truth table for XOR totally wrong.  Pin 14 has to be high. My brain ain't working today 

EDIT:  After some further hunting on this forum (https://www.diystompboxes.com/smfforum/index.php?topic=67967.msg542587#msg542587) I found that I am not the only one who made that mistake about XOR.  The image from the start of that thread shows pin 14 incorrectly wired to ground instead of being held high.

https://www.flickr.com/photos/loscha/2503124435/

[DrAlx]

Look at the TI CD4046B datasheet in the section that says Design Information on page 2.

The selected components must be within the following ranges:
5k <= R1,R2,Rs <= 1M
C1 > 100pF at VDD = 5V
C1 > 50pF as VDD = 10V


R1 is the resistor at pin 11
R2 is the resistor at pin 12
Rs is not relevant in the ultraflanger since pin 10 is not used
C1 is the capacitor between pins 6 and 7

So 22pF is out of the range mentioned on the datasheet (too low),
and so is the 2k7 at pin 11 (too low), and so is the 3M3 at pin 12 (too high).

So if the crackles are originating from the 4046 (and I strongly suspect they are) then maybe it's because those part values are outside the recommended range.

The design is using phase comparator 1 on the 4046.  I haven't the time to go through the whole sheet but there should be some formulas on there that describe how R1,R2,C1 give the centre frequency and range of the VCO.  I don't know if its possible to change R1,R2,C1 to lie within the allowed ranges yet give the same sweep range.

EDIT: No formulas on datasheet but there are graphs.  Based on that, I'd say there is no way to tweak the R1,R2,C1 values to be within the "allowed" range and still get the same VCO behaviour as Hollis' values.  e.g. the ratio of R2/R1 maps directly to fMAX/fMIN for the VCO.  Hollis' values deliberately push the chip beyond what is spec'd on the sheet.

If I was going to simply tweak values to keep them in spec and see if the problem disappears I would change 2k7 to 5k1, 3M3 to 1M, 22pF to 68pF.


EDIT 2:  You seem to have already done those sort of changes on the latest schematic you posted. Doh!