DIYstompboxes.com

Whats the difference between a Flangers Clocks VS a chorus Clock?

Are the Flangers and Chorus pedals BBD's the same? or they use different BBD's time stages?

A flangers LFO output is triangle mostly?

A chorus LFO output is triangle mostly?

Quote from: markphaser on December 21, 2005, 11:14:37 PM

Whats the difference between a Flangers Clocks VS a chorus Clock?

Flangers are clocked much faster than chorus, so the delayed signal is about 1ms-15ms, chorus more like 10-30ms

Quote from: markphaser on December 21, 2005, 11:14:37 PM


Are the Flangers and Chorus pedals BBD's the same? or they use different BBD's time stages?

Alot of them use the same BBDs, one with less stages makes sense for a flanger, so you can get a quicker delay.

Quote from: markphaser on December 21, 2005, 11:14:37 PM


A flangers LFO output is triangle mostly?

Usually triangular, some more refined ones have a hypertriangular clock (Paia hyperflange) to make the sweep linger on the lower part more than the upper.  A couple more eccentric ones might have a square wave LFO

Quote from: markphaser on December 21, 2005, 11:14:37 PM


A chorus LFO output is triangle mostly?

Almost always.  I've tried a square wave once, but it sounded bad.  A sinusoidal wave is very nice for vibrato, better than a triangle wave LFO.

Are you building any effects right now or writing a thesis or something?  You should tell us what you're doing with all this information you're asking about... it makes people feel like a encyclopedia or something when somebody asks a ton of questions without offering any background.

A stock medium-class flanger (like BF-2 and Ibanez Flangers) can sound like a pure chorus and many use them that way. The circuits are so alike and the delay range overlaps with chorus delays.
I placed a rare BBD in a BF-2 with half the delay stages (1024->512) and now it sounds pure flanger!

Quote from: markphaser on December 21, 2005, 11:14:37 PM

Are the Flangers and Chorus pedals BBD's the same? or they use different BBD's time stages?

Take a look at some schematics, google the data sheets for the devices you see, then decide for yourself. 

Thanks for the information

The Clock has a time constant? because there is a resistor and cap for the clock timing?

Whats the time constant for the clock for a chorus? the resistor and timing clock cap?

Whats the time constant for the clock for a flanger? the resitor and timing clock cap?

Using a Square wave VS triangle wave VS sine wave LFO does the notches/peaks shift different from using a different LFO output waveform? the notch/peaks arrrangement/order must changes from using a different LFO waveform output?

Look at it as a vibrato mixed with the original signal. The vibrato is achieved by delaying the signal back and forth, thus modulating the pitch. A Flanger is an inaudible vibrato due to small delays, but with large ratio range (0.5ms to 10ms for example) A Chorus is a more audible vibrato hence the longer delays with usually smaller ratio range (10ms to 30ms for example). The smaller the ratio range of a chorus, the less vibrato effect. (in fact a slow rate, short range chorus is the most beautiful IMHO)...

The vibrato signal is emulated by a BBD. This BBD has a delay, determined by a clock signal. 500kHz clock (these are square pulses!) gives around 1ms delay with a 1024 stage BBD, 0.5ms with a 512, ... etc. By bringing the clock rate to 20kHz, you get around 25ms with a 1024-stager. So achieving vibrato is done by fluently changing the clock rate from 500kHz to 400kHz to 300kHz to ... 50 kHz ... 20kHz and back. This done by controlling a current in combination with a cap. The current brings us to the LFO, which in fact gives a plot of how the signal frequency will change at the end.

So there is no "set" clock rate, because it would result in a "set" delay (unless you use "manual" controls = "filter matrix", etc...).

The ideal waveform for an LFO sweeping a clock for a delay/BBD-based effect would depend on the range of LFO speeds to be used. 

Flangers may be used at faster speeds to achieve some of the same kinds of sounds people try to produce with a chorus, but typically flanger LFOs are set for much slower speeds than chorus LFOs are set for.  Naturally, the range of LFO speeds available in typical flangers vs chorus pedals will be different because of that.

When the sweep is slow enough (let's say at least as slow as one complete sweep every two seconds or more), an LFO waveform that changes the HF clock more slowly at some points during the sweep cycle and more quickly in others is generally preferred.  While there is no required waveform, what is frequently called a "hypertriangular" LFO seems to fit the needs nicely.  Here, the LFO is more or less triangular at that point in the sweep cycle where the flanger is sweeping to and from its "highest" point, and becomes more or less sinusoidal as it sweeps down to an back up from its "lowest" point.  This results in the sweep slowing down during the points where you are more likely to notice movement in the location of notches.  If you think of it, it's a bit like picking up speed on the highway when there is very little to see, and slowing down as you get to the scenic parts.

Does this sort of a waveform produce any sort of "improvement" over a triangular waveform when sweeping at the faster speeds (let's say 1hz or faster)?  I happen to think not, though obviously there is plenty of room for other opinions.  Again, let's go back to our driving speed analogy.  If I were driving at 40km/hr (residential speed), I would be easily able to tell the difference between neighbourhoods where there were lots and lots of no parking signs, and those where there were very few such signs.  If I were driving 200km/hr, I would not likely be able to notice or tell you much about differences between those neighbourhoods except at the grossest level.

Just as an aside, hearing is really a specialized version of touch.  Your skin can detect, and humans can differentiate, vibrations applied to the skin over a frequency range that overlaps a bit with the hearing range (up to maybe 100hz or so).  If one varies the manner in which the pressure is applied to the skin (e.g., by having a standard apparatus that used a motor-driven "poker" to apply nonpainful pressure to a specific region), you would very likely be able to tell when the cyclical application of pressure was accelerating or decelerating...if the rate was slow enough.  Once the rate of vibration increases beyond a certain point, you simply stop being able to feel it as vibration.  Double the speed of yuor electric razor, and it stops feeling like a vibration and starts feeling like a steady pressure.  It is simply in the nature of the senses that the qualities of the cyclical variation become imperceptible as the rate of change goes up.

So, the upshot is that changing the waveform from the basic triangle shape coming out of a traditional 2-opamp LFO makes sense when sweeping very slowly, and is likely to produce noticeable changes in the "feel" of the sweep, but changes from the basic triangular waveform (gradual up, gradual down, gradual up, gradual down) are unlikely to produce noticeable or even detectable differences in feel once the sweep speed goes over a certain rate.  Because people rarely or never set their chorus for speeds as slow as they do for a flanger, the LFO waveform used for chorus is typically triangular, while the LFO waveform used for flanging will often be a modified version of a triangle.  Here is an article discussing it: http://ampage.org/hammer/files/hypertriangleclock.gif

Thanks alot Mark hammer and Steben for the information

Whats the frequency ( range) for a Flanger Clock?

Whats the Frequency (range) for a Chorus Clock?

Is the frequency based on a time constants because i see a resisitor and clock capacitor? (in the clock stage or cicuit)

You work out the frequency based on a few things:

1) what the intended delay range is
2) how many stages there are in the BBD device

If you switch from 512 to 1024 stages, and want the same delay time, you have to double the clock frequency to do it.  If it takes 1024 steoes to get from the input to the output, and you want to produce 1msec of delay time, the sample in any given "cell" has to move onto the next cell at the rate of one step every 1/1,024,000th of a second, or a 1.024mhz clock.  Since there are only half as many stages in a 512-stage unit, the same time can be achieved by moving along at only half that speed.

This is why a commonly available 1024-stage device makes a much better chorus than a flanger, and why a 512 or 256-stage device makes a better flanger than a 1024-stage one.  The demands on the clock circuitry are reduced when you don't have to clock quite so fast.  So, you can either aim for longer times, or use fewer stages to get there.

If you switch from 512 to 1024 stages, and want the same delay time, you have to double the clock frequency to do it.  If it takes 1024 steoes to get from the input to the output, and you want to produce 1msec of delay time, the sample in any given "cell" has to move onto the next cell at the rate of one step every 1/1,024,000th of a second, or a 1.024mhz clock.  Since there are only half as many stages in a 512-stage unit, the same time can be achieved by moving along at only half that speed.


A mere technicality/clarification:  a 1024 stage device would require exactly half that clock freq to render a 1 ms delay (hence Steben's numbers). 


Allegorically speaking, half the buckets are empty and half the buckets are full.  A BBD clock typically consists of two pulses per cycle, to empty and fill the buckets, so to speak.


So, for a 1024 stage device, it would be 512/500 kHz = 1.024 ms.   At 1 MHz, it would be 0.5 ms or so.

QuoteA mere technicality/clarification:  a 1024 stage device would require exactly half that clock freq to render a 1 ms delay (hence Steben's numbers). 


Allegorically speaking, half the buckets are empty and half the buckets are full.  A BBD clock typically consists of two pulses per cycle, to empty and fill the buckets, so to speak.


So, for a 1024 stage device, it would be 512/500 kHz = 1.024 ms.   At 1 MHz, it would be 0.5 ms or so.

I think you're correct.

By the way: it is true of course, that it's quite difficult to maintain that 1 MHz without buffering the clock signal. I'm not an expert, but I know  that's what John Hollis did with this Ultra Flanger.

here`s a little season-holiday-experiment:

oftentimes there`s talk about high clocking rate (very short delay) gets close to through-zero-flanging sound...

now what does it sound like,
when you achieve through-zero-flanging at significant lower clocking rates?

Well, judging from a certain recent product, I would say that at longer delay times it sounds like there's a big wave of hiss sweeping in and out!

Hey, you asked...    :icon_biggrin:

check the hiss-reduction mod...

Call me cautious, but I hate the idea of doing SMT surgery on what's basically a brand-new pedal, without knowing how much improvement it will make.

Maybe I'll send it in to EH and ask them to mod it...    :icon_rolleyes:

Thanks Mark Hammer and everyone else for the information

The Clock chip has a Modulation input mostly a triangle waveform LFO hooks up to the modulation input pin. What does the LFO
do to the Clock chip? it converts LFO voltage swing to digital binary inside the Clock chip?

The output of the Clock chip is digital binary? which hooks up to the BBD i see two lines going to a BBD chip from the Clock chip

If the LFO sweeps the Clock chip the output goes to the BBD would be?

I thought the Clock Chip just triggered the BBD chip to shut off and on?

If a LFO is sweeping up and down and is hooked up to the Clock chip's Modulation input then the clock chips output would have to sweep?

Quote from: notchboy on December 22, 2005, 04:55:23 PM
Well, judging from a certain recent product, I would say that at longer delay times it sounds like there's a big wave of hiss sweeping in and out!

Hey, you asked...    :icon_biggrin:

I own the mentioned "recent product" The hiss is barely noticeable, atleast on my unit.
I think you should consider yourself lucky that Ton has been so forthcoming with suggested mods.

How often do the designers do customer service?

Ton is a stand up guy and doesnt deserve to be harassed, take it up with EH instead of complaining here.
It sounds to me as if you have a defective unit.


Back to Markphasers "thesis" project.  :icon_wink:


Has anyone seen "walter" recently?  :icon_wink:

Fair enough.  My jabs should probably be aimed more at EH's so-called quality control than at Ton.

Whats makes the circuit hiss so much? what would cause the circuit to hiss so loud?
maybe once we know what causes circuit hiss we can troubleshoot it

Quote from: nelson on December 22, 2005, 07:38:33 PM
Back to Markphasers "thesis" project.  :icon_wink:

Has anyone seen "walter" recently?  :icon_wink:

Can you imagine a thread with both of them? I think it would be like crossing the beams in Ghostbusters.  :icon_eek:

I wonder which one would be the gate keeper and which one is the key master....  :icon_question:

--john

now what does it sound like,
when you achieve through-zero-flanging at significant lower clocking rates?

Hiss, distortion etc. issues aside, it would sound exactly the same.  Other than the groovy spaciness delay based effects sucker me in with, the other neato thing is that they offer a glimpse into time travel - well, time manipulation, at least.

Anyway, it's the relative difference in time between the two signals that produces the flanging effect, not the overall delay itself.  If a signal is delayed 100 ms and the other signal is modulated between 110 ms and 90 ms, the effect would be the same as if the fixed signal were delayed 20 ms and the other signal were modulated between 30 ms and 10 ms.

If you had a signal passing through one zillion stage device clocked at at a 1 zillion 500 thousand cycles per second, and another signal passing through a zillion stage device modulated between 1 zillion 450 thousand cycles and 1 zillion 550 thousand cycles, and you had a nuclear powered compandor on super-semi-conducting, liquid-oxygen-cooled substrate, your dad hitting a chord in celebration of your birth could be heard, through-zero-flanged, when the signal came out the other side of your inherited TZF flanger* on your 90th birthday.


*The Flanger Hoax 2000 - jointly designed by Ton and Stephen Hawking

Edit:  Hawking emailed and said I had to drop the clock rate by 1 zillion in order to get the overall 90 year delay. 

Lowering the clock rate on a Flanger would do what to the BBD chip?

Does Lowering or Speeding up the clock rate sweep the notches/peaks differently?

Lowering the clock rate on a Flanger would do what to the BBD chip?

Clock the BBD more slowly, increasing the delay time.  Clock it too slowly and the delay would become audibly discrete from the original signal (you would hear the effect as more of an 'echo' rather than just reinforced notches and peaks). 

Does Lowering or Speeding up the clock rate sweep the notches/peaks differently?  

The ratio of delay of one signal to an identical copy of the same signal determines the number and spacing of reinforced notches and peaks (teeth in the comb filter).  Somebody (Transmogrifox?) explained it much better in one of those other threads

Thanks Diyfreaque for the help and information

How can the Clocks frequency change the BBD's delay time?

512 BBD chip = a delay time of ?
1024 BBD chip =a delay time of ?

Quote from: nelson on December 22, 2005, 07:38:33 PM
Back to Markphasers "thesis" project.  :icon_wink:


Has anyone seen "walter" recently?  :icon_wink:

Hmm... well: http://acapella.harmony-central.com/forums/showthread.php?s=&threadid=1116711

Anyway, I'm off on (a very short) christmas vacation - have a good one!

/Andreas

Quote from: johngreene on December 22, 2005, 10:49:26 PM
Can you imagine a thread with both of them? I think it would be like crossing the beams in Ghostbusters.  :icon_eek:

I wonder which one would be the gate keeper and which one is the key master....  :icon_question:

More like The Holy Grail:

"What is the wing clock speed velocity of a BBD chip"
"I don't know....AGHHHHHH!

Then, of course, Tim gets it....

"What is the clock speed velocity of a BBD chip?"

"Panasonic or Reticon?"

"I don't know.....Aaaaggghhhhh"

Markphaser,

The delay time depends entirely on the clock frequency.  You use the LFO to modulate the clock frequency in order to vary the delay time. 

So, if you're designing a device, say a flanger, the delay line of which you want to operate within a certain range (say 15 ms to 0.5 ms), then you look at what devices you have available to perform this task.

By this, I mean you choose the number of stages, how fast (in real life) it can be clocked, and so forth.

You determine some of this stuff by calculating how much delay you can get with a certain number of stages at a certain clock frequency.

Right off the bat, you take the number of stages your device has and divide that in half for your calculations (see above stuff about why you need to do that).

For example, if you start with a 1024 stage device, like an MN3007 or MN3207, or both sections of a SAD1024, you divide the 1024 in half to get 512.

Calculate the required clock frequency to get your minimum delay time of 0.5 ms by dividing the adjusted number of stages by the target delay time:

512/.0005 = 1024000

You've just determined that, if you're going to use a 1024 stage device, your maximum required clock frequency is 1.024 MHz.

Now, do the same with your targeted maximum delay time of 15 ms:

512/.015 = 34133.3

So, now you know your minimum clock frequency needs to be 34.133 kHz. 

Now you know your clock frequency must be modulated by the LFO from 34.133 kHz to 1.024 MHz.  You look at your candidate BBD's and figure out if they can do it.  The MN's might just do it if you take real care in buffering the clock signals and so forth.  The SAD1024 is a natural.

Then you may look at what clock frequencies would be needed if you're using a 512 stage device.  This will result in more easily attainable clock frequencies, but pay attention to the minimum clock frequency you're going to use - try to keep it at least three times higher in frequency than the maximum frequency of input signal you intend to use the flanger with.  You really need to consider the signal *and* its harmonic content.

The clock generator itself provides two pulses per cycle note that BBD's mentioned above have two clock inputs - this is required for the internal bucket brigade system.  It's not really a binary moment there in terms of digital data.  I highly recommend reading the SAD1024 datasheet; it's explained quite well there.

My final recommendation is to actually start building something anything, (but concentrate on that one thing!!!!) and I don't mean on your computer.  If you don't already have one, buy a breadboard.  You'll learn a lot.  Don't start out with any rare or irreplaceable parts, save them for when you have your chops down.

Scott: did I see 2 TDA1022`s on top of your this year`s christmaspresents wishlist?
:icon_wink:

 :)

The usual.  I leave Mouser, Allied, and Digikey catalogs strewn about the house, open to strategically chosen pages.

Last year the same tactics resulted in a watch.  The DIY'ers curse:  no one else in the family understands what joy a simple 16 legged device can bring.

This year, at the company gift exchange, I received six spools of 22 gauge insulated wire for breadboarding.  I was in heaven......

Cheerio,
Scott

Quote from: nelson on December 22, 2005, 07:38:33 PM
Back to Markphasers "thesis" project.  :icon_wink:


Has anyone seen "walter" recently?  :icon_wink:

ROTFLMAO I smelled that instantly too.  walters alert!!!!!  He posted here too:

http://acapella.harmony-central.com/showthread.php?s=&threadid=1116711

Quote from: cd on December 23, 2005, 11:40:50 AM

Quote from: nelson on December 22, 2005, 07:38:33 PM
Back to Markphasers "thesis" project.  :icon_wink:


Has anyone seen "walter" recently?  :icon_wink:

ROTFLMAO I smelled that instantly too.  walters alert!!!!!  He posted here too:

http://acapella.harmony-central.com/showthread.php?s=&threadid=1116711

Oh boy, that thread is simply... SURREAL :icon_eek:

For every post answered he comes up with three or more new questions that don't show a sign of having understood the given answers!

Now to this forum, clicking to markphaser's name you get his profile statistics: (at the time of this posting, of course)

*Joined December 9th 2005
*Currently 144 posts
*"Full Member" position in less than a month!
*An average of 11 posts per day :icon_exclaim:

If you consider each post comes with an average of three very open questions we are talking about more than 30 questions/day :icon_eek:

A little projection: assuming the 11 posts/day rate is held during 365 days --a year-- this gives the terrifying number of 4011 posts  :icon_eek:  So by December 2006 he'll be breaking the 4k post barrier. WOW!

I've refrained as much as possible to make an off-topic comment like this, but I feel the forum is being monopolized for no good reason, and this thread seems to be going nowhere, despite so many nice people have spent plenty of their spare time trying to help.

well, you can also have this:
http://www.diystompboxes.com/smfforum/index.php?action=profile;u=4638

nevertheless,

MERRY Christmas, Sebastian.

Whats the clock rate or frequency for double tracking effect?

Changing the clocks frequencys changes the BBD's delay time?

A BBD is just a sample and hold storage ram chip?

disconnect the LFO with a constant voltage Does what to the BBD's delay?

disconnect the LFO with a constant voltage Does what to the notches/peaks?

I'd like to ask if it's possible to somehow reproduce the 'ray gun' effect of Tommy Bolin on Quadrant 4 (Billy Cobham, Stratus) with PT80 or Rebote delay?

Allegedly, Bolin used echoplex and manipulated the recording head position with his foot...