FLANGER (weird hiss issue)

[DrAlx]

Another difference to the Hollis LFO is that after the triangle wave, Hollis has a voltage divider made up of a resistor and 2 diodes just before the sweep pot.  That serves to warp the triangle wave (rounds off the triangle bottoms I think).  The reason for doing this is that the 4046 maps the control voltage linearly to frequency, but the sweep will sound better if frequencies are swept exponentially.  Without that warping of the CV, the sweep will seem to spend too long at one end.

[deadastronaut]

Cheers alex/guys.....

im going to re-breadboard it today, and i'll go from there....then i can endlessly tweak, modify....

and get any issues ironed out....hopefully. 

[deadastronaut]

UPDATE:

have it on breadboard and working  ( as per schemo with all chips powered after the 47r with 100uf...)

and still distorts when man/sweep are around 12 oclock....(i read about another guy having the same distortion with the original lfo) so.....

but at least now i can sub parts...

[DrAlx]

Astro, your output sallen key config is a bit different to Hollis's.
Hollis uses an odd scribble to represent bias voltage, but it look a little like a ground symbol so it's easy to make a mistake.

1) The Hollis design assume that the BBD output bias level is the same as the input bias level.
    That's an approximation BTW.
    So when you look at that final opamp in his schematic, the + input has bias level coming from the BBD output,
    but the - input has bias level coming from a 100k connected to Vbias (i.e. BBD input bias).
    All well so long as BBD output bias is close to BBD input bias.

    Your simpler Sallen key doesn't have that problem.  Bias is set simply by BBD output bias level.
    So I'm assuming simpler filter is not a problem.

2) The bottom of the Regen pot in Hollis figure goes to Vbias, not Ground as in yours.  You should fix that in yours.

[deadastronaut]

hi guys, well i replaced the lfo with the original, except i used 2x 10k's for the voltage divider

and voil'a it works, and no distortion .   ( i tried the 10k>led>led>1n914.) and it worked out the same with the 2x 10k's...excellent.

however this lfo is a bit more 'lumpy' (when trying both dividers)  than the one i used in my schemo.....

ive tried changing various values in the lfo, but still not as smooth as my original...

anyway, distortion has gone so thats good...    (and yes i tied the regen to vb.)

now to try and get it smoother...rather than lumpy jerky on the bottom of the lfo...(i added an led indicator. )

[11-90-an]

Cap to ground on output of lfo? 

You out it on pcb already?

[deadastronaut]

lol..no....

got it kinda sorted....now with 3 way on/off/on for normal/manual/ultra slow lfo....

getting there...

[bluebunny]

So what kind of shape is the LFO right now, Rob?  There's some interesting stuff in ROG's Tri-vibe project page about how they went about adjusting the shape of their LFO.

[deadastronaut]

cool, i'll have a read..


well its a lot smoother triangle...

the led is pulsing a lot better/smoother etc....

i used 10k x 2 for the divider,  used a 250kb for the speed, with a 2.2k pot 1....much better...

off lug 2 of the rate i have a switch for 470k/0/4.7m  for norm, man, ultra slow as fk....

dont worry, we still have another 72 pages to fill as per tradition......

[11-90-an]

If you still have some ICs left, to ensure 72+ pages, you could try these ideas...

2 or more flangers in parallel with adjustable speed and depth for both...
2 or more flangers in series with adjustable speed and depth for both...
Mixing these ideas up...
Fitting it in a 1590a...

[deadastronaut]

or i could just dress up as a frog and move to tunisia, but i wont just yet 

[DrAlx]

Regarding lumpy sweep, check out "hypertriangular clock" info from Mark Hammer's site.
Second gif on this page...

https://hammer.ampage.org/?sort=new&page=10

I mention that because it's an example using 4046 as the VCO, and shows one way of warping the triangular CV so it gives more exponential sweep.   I don't know if its better or worse than what Hollis is trying to do with the diode before the sweep pot.

[PRR]

or i could just dress up as a frog and move to tunisia, but i wont just yet 

Do you have a GoFundMe for that?

link

[deadastronaut]

well i dressed up as frog and moved to tunisia, and it seems to have helped....thanks for the tune paul, i think that helped too, though it annoyed the locals

anyway, here is where i am at the moment, hiss, and dirt has gone now, and its behaving itself...thank f...

i cut the input feedback down to 82k... then boosted the output volly on the recovery stage...and voila' distortion gone..i now have unity around 12oclock..and a little boost if needed. great..

i added the 1n914 and led before the sweep as the lfo was kinda strange, and this has rounded it off a bit...without those the lfo was way lumpy, and quite unusable...now its smoothed out  ( i found my el cheapo DSO150 scope that i had forgotten about so that was handy to see the waveform.....great fun.

i'm still tweaking the manual control a little, and was trying to suss another way of rigging it to be more useable/responsive.....anyway onwards and upwards....

[bluebunny]

Is it too early for popcorn? 

[deadastronaut]

yes..

[bluebunny]

Bu99er.    Leftover pizza for lunch, then... 

[anotherjim]

[deadastronaut]

Aha, my manual pot was dodgy....sorted now.    

[11-90-an]

Now that you're done...

GIBB US PIKTURES OF THE PCBBB...

Please..?

(Also the problematic one.... we need proof of at least some incompetence....)