EH EM on sim(LTSpice) hypertriangle

[rx5]

hi guys.

is this the proper way of seeing the hypertriangularity of the LFO? note that I have added R2, or it doesnt oscillate. also im using a different comparator(native to the sim).

green trace = output from ramp gen
blue trace = output from comparator

[rx5]

this image, is still in simulation

ive changed R6 to 10k(from previous 100k)

the bottom part of the hypertriangle is much wider

so I guess I must be on the right track then 

[PRR]

I'm going to need a bigger monitor. (Is your monitor even 4656 pixels wide?)

[DrAlx]

That VCO you are trying to model produces a clock period that increases/decreases linearly with control voltage.
So when you feed it a triangle wave control voltage (from an LFO), the clock period will rise and fall linearly as a triangle wave too.  The end result is a frequency sweep that is hyperbolic, but to be honest I think it is better to simply think of things in terms of time.  I.e. Triangle wave LFO results in triangle wave BBD delay.
See
https://www.diystompboxes.com/smfforum/index.php?topic=116836.msg1083340#msg1083340

[DrAlx]

.. and BTW I never managed to get an LTspice simulation of that VCO to match reality.

[DrAlx]

What you should be looking at is not an output voltage level from that VCO but rather zoom in and look at the period between drops in the comparator output (or look at period between cap discharges).
If you try 3 different fixed CV levels (1V, 3V,5V) and look at the period between cap discharges you should see linear increments in the period .  i.e. the increase in period going from 1V to 3V, should be same increase as from 3V to 5v.

[rx5]

@PRR

Time for upgrade?    sorry still learning to post images. im using phone.


@DrAlx

on the simulation, I found out that the center sweet spot would be around 3v, with 1v swings either way. so 2v min and 4v max... on my actual boxxed unit, with RATE(min) & RANGE(max) , its 'center' is around 2v5 . when its on RANGE(min), center changes,goes down. dc coupled. but I could impact it a bit by changing the 39k value(going to the 3 resistor voltage dividers)

just got lucky I got it to oscillate.

shown below, actual scope shots. the linear triangle is when RANGE is at minimum. I guess we dont need hyperT at this point. 2nd image is with RANGE at full. hyperT. was taken at the LM311 collector output. and is inverted, as opposed to the simulation.

edit: thanks for the link. reading 
edit2: I have tried both positive(metallic sounding) and negative(hollow sounding,like wah) with just a simple transistor circuit. needs bleeder resistors. or better circuit .
use headphones.
https://youtu.be/yqjonyeGMvQ

ive also tried it with a static(variable delay) DRY. TZF/NZF does make the 'negative feedback' sound less of a wah. but tbh tzf/nzf doesnt sound musically pleasing, with the out-of-phasing etc etc..maybe its just me. but it does add more dimension to the flanging sound. I still need to do the 'tracking LFO' just like what you did
edit again:
regarding uploaded video, its NOT negative feedback, just inverted BBD output. duh! 

[DrAlx]

Not sure what you've measured there. Comparator output voltage is not relevant. Remember that it is just used to trigger a flip flop. What you should be plotting is either the period off the comparator output (which should be linear with CV) or frequency of comparator output (which will be hyperbolic with CV).

[rx5]

indeed Ive not made any meaningful measurements,  yet. Just showing visually that there is 'something' to be seen with regards to 'hypertriangle', since this circuit is almost impossible to simulate. scopeshots where taken on 'DC'. on AC coupling, it shows a different set of waveform, a small hint of the hyperbola, but not as 'clearly' as on DC couple. I hope this clears the confusion.

mistress mystery page , tons of info


'seeing is believing' as what most people say. 

[rx5]

Alex

I'd appreciate tips/tricks/hints for a better near zero(or thru zero) sound.

ive done the AP filter,  MEASURED at around 420uS max phase delay + invert. the effect was subtle, but its there, upper sweep. delayed dry & wet have almost same amplitude before mix.

also did the manual 'sweepable' delay, CV control,initially by pot, 1v-5v8 cv, mn3101/mn3007 combo. I find its 'best' setting for tzf/nzf near highest osc(900KHz - 1.1MHz), I didnt measure. so does this mean its better to just 'put it' in the upper freq band?

[DrAlx]

Minimum EM delay is around maybe 700us.  If certainly can't go as low as 400us that you are getting with APs, so you have a delay mismatch of at least 300us.
For low frequency audio signal of 100Hz (i.e. wave of period 10000 us), that mismatch in delay corresponds to a phase mismatch of 3%, so you could almost get cancellation of 100Hz audio. But for 1kHz audio, you would get a phase mismatch of 30% and poor cancellation. So low frequency audio could almost cancel with your APs approach but the higher frequencies wont. You really need the delays to match closely for TZF effect.
If you are using BBD for fixed delay instead of APs and are not duplicating the EM circuitry then you would just have to experiment with different fixed delays and see what works best.

That curve on your scope or simulation has nothing to do with a hypertriangular sweep. You are plotting voltage. Hypertriangular sweep does not refer to how a voltage changes in time.  It refers to how the clock frequency changes in time. In other words, horizontal axis is time, vertical axis is clock frequency.

[rx5]

LFO vs VCO

normal flanging:

matrix mode:

cheers
-Ralph