New Phaser Design - “Causality 4” – Built with you in mind!

[mistahead]

Ldr led combo never going dark enough at the highest end of the speed variable, while the effect of one light pulse is still talking place the resistance doesn't change enough before the next pulse, resulting in the later pulse pushing through into the consequence of the one before?

If that made sense... Possibly waaaaaay off the mark even if it did...

[Scruffie]

I suspected I needed to prevent the resistance reaching zero somewhere along the way to prevent a zero R feedback loop somewhere else in the series.

hmm... i had pictured a solution based around what you mean, but i've been up all night so i got it all wrong lol. i'll have a ponder when i wake up. i have something similar (albeit minor) with a compressor's release pot squealing on 0 resistance, but i hadn't really connected the two.  thanks for the suggestion, i'll get on that once i get the whole sweep thing sorted (if i even need to...).

i've tried decoupling the whole LFO, but no end to the thumping. maybe it's to do with the (white) LEDs i'm using, can't remember exactly but i think the current rating was on the higher side. it's only at full speed that this happens, which intuitively just seems like the circuit voiding its bowels to keep up.

Increase regen series resistor/decrease cap value.

Yes making the LFO drive a high bright LED probably is causing thump.

[MrStab]

Increase regen series resistor/decrease cap value.

ah, so i was thinking right then. thats a first. the LED thing does make sense too.
i'm tempted to record a clip of how mine sounds later just so you guys can confirm whether i'm talking crap about the sweep or not. any ideas where i'd set the Shape pot to "default", ie. the resistance there'd be if the pot wasn't there at all?

[Scruffie]

Increase regen series resistor/decrease cap value.

ah, so i was thinking right then. thats a first. the LED thing does make sense too.
i'm tempted to record a clip of how mine sounds later just so you guys can confirm whether i'm talking crap about the sweep or not. any ideas where i'd set the Shape pot to "default", ie. the resistance there'd be if the pot wasn't there at all?

Link to the schematic you used? (this thread's huge!).

[MrStab]

sure thing, scruffie:

http://tagboardeffects.blogspot.co.uk/2012/02/causality-4-mkii-rick-holt.html

[Scruffie]

That's the layout, the schematic it's based on though.

[MrStab]

ah, my bad - i have a bad habit of not reading things properly. to the best of my knowledge, it's this, give or take:



here's a low-quality clip of some rung-out chords with the Shape pot in the middle position (the hiss & distortion are from the recording itself):

does the sweep seem right?

[Scruffie]

Shape would be the middle of the pot for 'normal' sweep but if that doesn't work you could pull the diodes and pot if you don't need it.

NE5532 is a pretty poor OpAmp for an LFO, has quite a high current draw so may well be the cause of your 'thump' you can try putting a large cap (47-100uF) straight across pins 8 & 4 (Positive to 8, negative to 4) making it drive an LED surely wont help much either.

Listening to your clip it sounds like a very poor hyper triangle with a dip at the top, possibly due to current issues. Does it improve with the range/width down?

[MrStab]

i forgot to actually mention what opamp i was using: i've tried a TL072 (currently using that), a LM358 and an OPA2227 (even though the current on that last one isn't on the low side).

with Range and Width at 0, the transition seems a bit smoother but the disproportionate "jump" is still there. apologies for using a lot of layman's terms here.

as the only clues i could find beforehand pointed to current, as you mention, i tried giving the LFO a fixed vbias via. a 78L05 instead of the separate 220k voltage divider, but that killed the effect altogether (despite what seemed like the right voltage readings). think the cap between opamp V+ & Gnd might alleviate the sweep issue as well as the thumping? i should mention that the sweep problem occurred whilst i was skipping the LEDs altogether.

thanks for the help!

Edit: a 1000uF (overkill much!) didn't kill the thumping, although i recall a big improvement in the TL072 over the OPA (as is to be expected, i guess) so im not too concerned about that - i'll switch out the LEDs for something less current-hungry at some point. unless ofc it's directly related to the sweep issue

[jimilee]

Will replacing r12 with a trimmer allow us to bias the phase?

[Scruffie]

i forgot to actually mention what opamp i was using: i've tried a TL072 (currently using that), a LM358 and an OPA2227 (even though the current on that last one isn't on the low side).

with Range and Width at 0, the transition seems a bit smoother but the disproportionate "jump" is still there. apologies for using a lot of layman's terms here.

as the only clues i could find beforehand pointed to current, as you mention, i tried giving the LFO a fixed vbias via. a 78L05 instead of the separate 220k voltage divider, but that killed the effect altogether (despite what seemed like the right voltage readings). think the cap between opamp V+ & Gnd might alleviate the sweep issue as well as the thumping? i should mention that the sweep problem occurred whilst i was skipping the LEDs altogether.

thanks for the help!

Edit: a 1000uF (overkill much!) didn't kill the thumping, although i recall a big improvement in the TL072 over the OPA (as is to be expected, i guess) so im not too concerned about that - i'll switch out the LEDs for something less current-hungry at some point. unless ofc it's directly related to the sweep issue

358 is probably the best of that bunch to use, I favor the 1458 my self.

Could be related to the sweep issue yes, or you may just need to adjust the current to the OTAs Pin1 & 16 via the 4k7, raising it very slightly (don't lower it, too much current could pass and kill the OTA) may reduce the current and stop the OTA 'overshooting' I guess you could say.

I see the regen doesn't use a cap, adding one in series with that resistor may help tame oscillation, letting only higher frequencies through.

[Scruffie]

Will replacing r12 with a trimmer allow us to bias the phase?

From what I can tell with a quick trace is that is to boost/reduce the output volume should it be needed.

[MrStab]

... adjust the current to the OTAs Pin1 & 16 via the 4k7, raising it very slightly

i tried raising it from 5.7k up to about 15k to no avail - it didn't seem to affect much. good call though, i've messed around enough with the LFO section and i didn't think of adjusting the OTA.

i'm at a loss for ideas now. the LFO opamp's V+ pin fluctuates, which is likely
the cause, but a few attempts at decoupling have done nothing. i have no 9V regulators available (even though reading is closer to , and all i have are a bunch of 9V zeners. maybe i should hook up the LFO to another power supply altogether and claim the effect is so badass it needs two adaptors. lol

V+ fluctuates by like 5-10mV, whereas vbias goes absolutely crazy by about 2.5V.

i'd already tried the bias pot you mentioned, jimi. sorry, should've said.

[Scruffie]

... adjust the current to the OTAs Pin1 & 16 via the 4k7, raising it very slightly

i tried raising it from 5.7k up to about 15k to no avail - it didn't seem to affect much. good call though, i've messed around enough with the LFO section and i didn't think of adjusting the OTA.

i'm at a loss for ideas now. the LFO opamp's V+ pin fluctuates, which is likely
the cause, but a few attempts at decoupling have done nothing. i have no 9V regulators available (even though reading is closer to , and all i have are a bunch of 9V zeners. maybe i should hook up the LFO to another power supply altogether and claim the effect is so badass it needs two adaptors. lol

V+ fluctuates by like 5-10mV, whereas vbias goes absolutely crazy by about 2.5V.

i'd already tried the bias pot you mentioned, jimi. sorry, should've said.

The OTAs in the chip should be matched well enough not to need separate resistors so that's good that it didn't work sort of, clears that of being the issue.

So we head back to the LFO, it shouldn't need regulating, get rid of those LEDs if you haven't already and see if there's a change.

Having a think about it, I now wonder if it's an issue with NOT using a NE5532...

[MrStab]

the issue was present before i had LEDs hooked up, unfortunately - i just wired the anode & cathode wires together at first.

i've been holding caps across all 3 IC's (soldered one across the TL074 & am ensuring contact with the rest) but still getting voltage fluctuation. i even separated the LFO V+ from the rest of the circuit at one point.

you may have noticed this from a few pages back, but i copied this from another forum:
"sundgist wrote:Taking vref on the lfo buffer/gain stage from the lfo's voltage divider causes the lfo to swandive just before it gets to each positive and negative peak. I've moved the 10k resistor and wire along a bit and added a link to the main power supplies vref."

pretty sure i tried that, too. while it sounds like the same effect, it doesnt sound like the same cause.

there's a response at the very bottom of the layout page to someone having the same problem, where they question the opamp usage, and the response is: "nope. in mine the TL072 and NE5532 sound nearly identical and operate the exact same way." not exactly scientific, though

[Scruffie]

If i'm reading the layout you used correctly (and i've had a few beers) width 2&3 go through a 47k resistor instead of a 10k to the main V.Ref (Row 3, width 3) that would increase the LFO gain, drop it to 10k and see what happens.

Only the LFO would need a cap across +/- the TL074 & 13600 shouldn't need it.

[MrStab]

no success, but progress! changing that 47k for a 10k made the problem worse, which seems good from a diagnostic standpoint. it flips back to the peak (?) of the sweep much more snappily. should i try going over 47k instead?

Edit: scratch that, 100k just speeds up the effect but still has the problem. interestingly, the LED thumping has now gone. there's a kinda scratchy sound as the sweep descends.

[Scruffie]

no success, but progress! changing that 47k for a 10k made the problem worse, which seems good from a diagnostic standpoint. it flips back to the peak (?) of the sweep much more snappily. should i try going over 47k instead?

Edit: scratch that, 100k just speeds up the effect but still has the problem.

woops, said i'd had a few haha, yup decreasing that increased gain.

No idea why that increased the speed at the moment.

I'll see what I can think of tomorrow.

[MrStab]

i may have accidentally turned up the speed pot, now i think about it...

mucho cheers for the help so far man, you've suggested more after a few beers than i've managed sober lol. it is kinda late round this neck of the woods. i'll sleep on it as well

[MrStab]

i know it's bad practise not to mention parts substitutions, but i've just realised i have 1.5K resistors  in place of the 1.8. could this be to blame?