Pulsar tremolo: want to reduce its noise? here's how

[bioroids]

Hi!

This may be obvious to some of you, but I'll post it anyway

This is what I did to get rid of most ticking and pulsing noises:

Use a separate ground and power path to the oscillator. Decouple it with a 1k resistor and a 470uf cap in parallel with a 10nf ceramic.

I added a 100nf ceramic before the 1k resistor but maybe its not necessary.

Use a 100ohm resistor with a 470uf cap to decouple the power supply that goes to the rest of the circuit. Add a 10nf ceramic in parallel with the 470uf

Have the oscillator section (with its pot and switch cables) separated from the rest of the circuit (in a far corner of the board for example)

It really improves the noise performance of this tremolo (wich I like very much)

Luck!

Miguel

EDIT: changed the title to clarify it

[puretube]

this thread should be read by more people!!!

:!:

"a circuit is only as good as its supply"

[bioroids]

He he maybe the title sounds more like a question about how to reduce noise (asked 1000 times)  instead of a proposed solution to do that?? who knows... :?:  

As a side note, I tried running the oscilator with a 7805 regulator but it didn't worked as well.

Luck!

Miguel

PS: I like that quote! you made it up??

[audioguy]

Use a separate ground and power path to the oscillator. Decouple it with a 1k resistor and a 470uf cap in parallel with a 10nf ceramic.
I added a 100nf ceramic before the 1k resistor but maybe its not necessary.
Use a 100ohm resistor with a 470uf cap to decouple the power supply that goes to the rest of the circuit. Add a 10nf ceramic in parallel with the 470uf
Have the oscillator section (with its pot and switch cables) separated from the rest of the circuit (in a far corner of the board for example)
It really improves the noise performance of this tremolo (wich I like very much)

Let me start off by saying that I REALLY love this pedal...
Im doing some research on killing the tick- I found some info stating to raie C2 & C5 (On TonePads layout) to 100uf and 33 or 47uf respectivly, but that didnt really solve my problem.
I would love to give these recomendations a try, but Im afraid Im not advanced enough to understand exactly what you're saying here.
You suggest we use a separate ground and power path to the oscillator... but what part of the circuit is the oscillator and how would I hook up the decoupling resistor and cap?

Thanks for your time and assistance!
Audioguy

[bioroids]

Hi!

I'll post a schem in the next few days

Luck

Miguel

[audioguy]

That would be awesome!


Thanks!

[R.G.]

Good work, bio!

It's sometimes hard to remember that even if power and ground conductors are copper, they still have resistance and that current variations on them will get reflected into the circuit, but you got it!

Again, good work.

[loki]

Hi!

I'll post a schem in the next few days

Luck

Miguel

Do you also know how to add a LED that pulses in synch with the trem?

[aaronkessman]

the pulsing LED is a simple addition to the circuit and is found in at least the latest schem for the trem lune on the commonsound site.

[bioroids]

Hi!

I posted a schematic with some descriptions on my page:

http://www.dedalofx.com/bioroids

Hope it is clear, if not, let me know!

By the way it has some other mods besides the deticking, but you can take the parts you like and "paste" them in your favorite version .

Also the component labels are different from other layouts, so be carefull!

Luck!

Miguel

[audioguy]

AWESOME! I'll give it a shot sometime this week.

[NaBo]

THANKS!  
I was so excited to make a pulsar cuz it had both sine and square wave trem, but then I got un-excited after finding out about the ticking problem.  I am now excited again!  So using this separated-power-supply-trick for any ckt with a ticky LFO should solve the problem, right?

One quick question though:
You say your part substitutions give a different range for the LFO... is the range slower, faster, or just wider than stock?  Nothing a comparison between schems wouldn't answer, but I'm lazy    If it's faster or wider, then count me in, I'll just build your version top to bottom

[bioroids]

Hi!

I don't really remember how the range compares with the original, I did those tweakings almost a year ago.

I think it has a pretty good range the way I do it, thoug it's faster than I like, and I'd could use it slower also. In all cases, you can switch between versions just by changing a few caps (the two 1uF on the LFO were 0.47uf originally) and a resistor (R16 was originaly 1M or 2M2 I think).

Also beware that this pedal doesn't have exatcly "sine" and "square" waves, it's more like "square" and "rounded edges square". Sounds pretty nice anyway.

Luck!

Miguel

[StephenGiles]

Fantastico!
Stephen

[audioguy]

Has anyone drawn a layout for this new noise free schem?

Thanks!

Audioguy

[John Lyons]

Has anyone done up PCB layout for this yet? I was just about to use the tone pad version but this seems like an improovement for sure! I'd love to get it right the first time.

John

[sta63bmx]

Can someone post up this schematic again?  I cannot find it at the site and I would like to try it.  If I run my Pulsar alone or at the very end of my signal chain, it does not tick (woot), but I would like to be able to move it around on the pedalboard.  Putting it last just kind of seems like a lame treatment of the symptom, not the problem.

When he says "decouple" does he mean to add a filtering section for the lfo like this?  It's a little unclear to me.



When you're done laughing at the painty goodness, please advise.

[Mark Hammer]

No.

The added components are meant to form a lowpass filter that will reduce the "treble" showing up in the supply.  Think of the supply as "crackling" every time one part of the circuit suddenly demands current for its needs.  The circuit that bioroids supplied (and is used in many other circuits too) uses a simple RC lowpass filter on the power supply to eliminate that crackling, in much the same way that you might turn the tone way down on your guitar and treble down on your amp if you had a crackly volume pot on the guitar.  There is certainly more to it than that, but its a handy way to think of it.

The lowpass filter is formed by having the small-value resistor in series with the supply (between +9v and the LFO on your drawing), and the cap going from where the supply feeds the circuit/chip to ground.  Look at the tone control on the Proco Rat (also a lowpass filter), or the simple cab simulators on many of the ROG emulations of different amplifiers, and you'll see the same configuration.

The "rolloff" is given by F = 1/(2*pi*R*C).  Here, a 1k resistor and 470uf capacitor yields a 6db/octave drop in fluctuation in the power supply starting around .33hz.  What this means is that sudden changes in available current in one part of the circuit caused by sudden demands in another part will be smoothed over if they occur faster than a certain rate/speed/rise-time.  The extra 1nf cap is for those superfast blips, and to compensate for the quirks of larger electrolytics.  It is common to have large, medium, and tiny caps in parallel in stereo power supplies.

I realize it is a little odd to think of the power supply as having "tone", but whether transistors, chips, LFOs or gain stages, its all just fluctuations in current anyways, isn't it?

[bioroids]

Listen to Mark, he's the man!

This is what I meant (I took the liberty to edit your schem):


Anyway, this is an improvement on the noise issue, but it doesn't remove the noise completely. It also depends on layout, and the quality of the LFO coupling cap (4.7uf IIRC should be tantalum). After doing all this, I just switched to other trem design, as this is inherently noisy it seems.

Luck

Miguel

PS: lol the spell checker always makes me laugh, it tries to change LFO for UFO

[sta63bmx]

Heeeeeeeeeeeeeeeeeey, that was a copyrighted schematic...lol j/k  I think I'm with you.  In all our digital circuitry in the lab, we usually have 0.1 uF caps between power and ground to dump any transients off the line.  So any crackling and high-frequency transients in the power lines just get dumped right to ground?  Thanks a lot for the info.  I apologize for digging up an old thread, but was keen to get that schematic.  I'll give it a whirl.

I'm building a trem pedal for a girl, and she (understandably) had issues with the noise.  I may just build her an EA trem or tremulus lune instead and keep the pulsar for weirdness' sake.