Helicopter Oscillating Noise: Poor filtering vs. Typical of Breadboards?

Started by Twhjelmgren28, September 17, 2020, 04:17:35 PM

Previous topic - Next topic

Twhjelmgren28

Some of you may've seen my schematic already in the Simulation subforum.  I've got a distortion breadboarded with some helicopter oscillating noises.  First, here is the schematic:




The only components not shown are a typical tone and volume pot off the output as well as power supply filtering.

The deal is, I'm getting intermittent oscillating noises - as the title states, sounds like a helicopter.  I wasn't getting any noise with my tele but when I switched to my strat, that's when it started.  For reference, my strat has vintage output single coils. 

I'm wondering if you guys think this is a case of poor filtering (see above schematic) or if it's semi-normal dealing with a breadboard.  As mentioned, I do have power supply filtering on my breadboard, it's just not shown in the schematic above. 

To make matters a bit worse, my breadboard is near my computer and dimmable lights.  I'm also running it into my DIY Princeton Reverb, although I've never had an oscillating issue with this amp. 

Any help is appreciated.
  • SUPPORTER
I'm a rookie:  Teach me your ways, wise sensei!

Digital Larry

Digital Larry
Want to quickly design your own effects patches for the Spin FV-1 DSP chip?
https://github.com/HolyCityAudio/SpinCAD-Designer

Twhjelmgren28

Quote from: Digital Larry on September 17, 2020, 04:27:33 PM
How's the Vref generated?  Is it filtered?

EDIT:  I think it actually is filtered.  I have filter caps between 9v and ground on my entire breadboard.  I then have two 330k's splitting voltage off that supply to bring VRef to the 1Meg...

For reference, I'm using 47uF electrolytic and .1uF caps to filter the entire breadboard.

Should I have different/additional filtering for the VRef?
  • SUPPORTER
I'm a rookie:  Teach me your ways, wise sensei!

blackieNYC

  • SUPPORTER
http://29hourmusicpeople.bandcamp.com/
Tapflo filter, Gator, Magnus Modulus +,Meathead, 4049er,Great Destroyer,Scrambler+, para EQ, Azabache, two-loop mix/blend, Slow Gear, Phase Royal, Escobedo PWM, Uglyface, Jawari,Corruptor,Tri-Vibe,Battery Warmers

Twhjelmgren28

I have not yet - I'll give it a shot and post back here.  I'll also try it through a different amp to rule that out.
  • SUPPORTER
I'm a rookie:  Teach me your ways, wise sensei!

Twhjelmgren28

So I probably should've tried that before posting.  When I turned off the computer and lights, I didn't hear the helicopter sound.  I think it may have been the computer b/c I was able to successfully turn my lights back on without any noise.

So in theory, when the circuit is placed in a grounded enclosure, that should block most of that noise right? 

My filtering seems okay?
  • SUPPORTER
I'm a rookie:  Teach me your ways, wise sensei!

duck_arse

try/add another R//C line filter before the R7, so that stage supply is isolated from the 741 supply.
" I will say no more "

Digital Larry

Quote from: Twhjelmgren28 on September 17, 2020, 04:32:17 PM

EDIT:  I think it actually is filtered.  I have filter caps between 9v and ground on my entire breadboard.  I then have two 330k's splitting voltage off that supply to bring VRef to the 1Meg...

For reference, I'm using 47uF electrolytic and .1uF caps to filter the entire breadboard.

Should I have different/additional filtering for the VRef?
Far as I know, yes... although Vref generally feeds high impedance loads, it should act as a ground for AC signals.  As it is, the impedance of the Vref source is 330k/2 which I'd say is too high.  First thing you could do is lower those resistors to 10k apiece and second thing you could do is to put a cap across the lower one.  This is all "rule of thumb" and may not have anything to do with what you're seeing.
Digital Larry
Want to quickly design your own effects patches for the Spin FV-1 DSP chip?
https://github.com/HolyCityAudio/SpinCAD-Designer

Twhjelmgren28

Quote from: duck_arse on September 18, 2020, 11:35:54 AM
try/add another R//C line filter before the R7, so that stage supply is isolated from the 741 supply.

When you say before R7, are you saying between V+ and R7? I'm totally game for doing so, just want to be clear I understand.


Quote from: Digital Larry on September 18, 2020, 11:51:04 AM
Quote from: Twhjelmgren28 on September 17, 2020, 04:32:17 PM

EDIT:  I think it actually is filtered.  I have filter caps between 9v and ground on my entire breadboard.  I then have two 330k's splitting voltage off that supply to bring VRef to the 1Meg...

For reference, I'm using 47uF electrolytic and .1uF caps to filter the entire breadboard.

Should I have different/additional filtering for the VRef?
Far as I know, yes... although Vref generally feeds high impedance loads, it should act as a ground for AC signals.  As it is, the impedance of the Vref source is 330k/2 which I'd say is too high.  First thing you could do is lower those resistors to 10k apiece and second thing you could do is to put a cap across the lower one.  This is all "rule of thumb" and may not have anything to do with what you're seeing.

Got it - so general rule is about 10k to feed an opamp VRef?

I'm not understanding where I would put the cap - are you saying between ground and the 10k leading to ground, I'd put another cap to form low pass filter?
  • SUPPORTER
I'm a rookie:  Teach me your ways, wise sensei!

antonis

Not so much as LPF rather than Vref ripple rejection..

P.S.
DO what duck_arse told you about J201 power filtering.. :icon_wink:
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

Digital Larry

Quote from: Twhjelmgren28 on September 18, 2020, 02:31:56 PM
Got it - so general rule is about 10k to feed an opamp VRef?

I'm not understanding where I would put the cap - are you saying between ground and the 10k leading to ground, I'd put another cap to form low pass filter?
I guess it depends on what is attached to the Vref.  You have both DC and AC to consider.  Any DC currents which come out of Vref will tend to pull the voltage one way or the other.  e.g. nobody would ever do this on purpose, but if you had an LED going to Vref and it was pulling just 1 mA, if the impedance of the Vref was 100k, then it would "try" to offset 100 volts.  You don't have 100 volts but it would go as far as it could.  AC impedance is probably more important and will be dominated by the cap usually.  The corner frequency below which the cap starts becoming less effective is 1/(2 * pi * R * C) where pi = 3.14 and C is in farads.  So, higher R lets you get away with a smaller cap, at the expense of being more sensitive to DC currents.

This is why engineering is so fun (sometimes) because often different forces tend to offset each other and there's no simple solution that is also the best.  You usually settle for a compromise.

Filtering Vref:

You have R1 = R2.
R1 goes to +V and R2.
R2 goes to R1 and GND.
Junction of R1/R2 voltage = +V (R2/(R1 + R2) = +V/2.
Put the cap across R2.
Digital Larry
Want to quickly design your own effects patches for the Spin FV-1 DSP chip?
https://github.com/HolyCityAudio/SpinCAD-Designer

Twhjelmgren28

Think I'm understanding. After reading through this thread a couple times, finding another thread on power supply filtering, and checking out some schematics, I'm feeling much more knowledgeable and will be adding to my filtering.

Thanks guys!
  • SUPPORTER
I'm a rookie:  Teach me your ways, wise sensei!

duck_arse

Quote from: Twhjelmgren28 on September 18, 2020, 02:31:56 PM
Quote from: duck_arse on September 18, 2020, 11:35:54 AM
try/add another R//C line filter before the R7, so that stage supply is isolated from the 741 supply.

When you say before R7, are you saying between V+ and R7? I'm totally game for doing so, just want to be clear I understand.


yes - up to 2k2 should be ok, and a cap to ground from the junction - 22uF, 47uF, whatever is at hand.
" I will say no more "

EATyourGuitar

this is unrelated but I think you should have the clipping diodes to vref and actually bias the jfet so that it is not always a half wave rectifier after a diode clamp. this circuit would be more usable without the half wave rectifier on the output.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

Twhjelmgren28

Quote from: EATyourGuitar on September 19, 2020, 10:29:22 AM
this is unrelated but I think you should have the clipping diodes to vref and actually bias the jfet so that it is not always a half wave rectifier after a diode clamp. this circuit would be more usable without the half wave rectifier on the output.

Oh man - just when I was feeling I understood what my issues were! 

So, are you saying anytime I have clipping diodes referenced to ground going to a transistor, I'm causing a half-wave rectifier?  Can you explain this more for me?  As you can tell, I'm still a beginner  :D

I'm really liking the sound from the circuit as it is but I'll try it referencing the diodes to VRef. 

EDIT:  So I did just try my diodes referenced to VRef and I see what you mean by more useable.  There is much more volume and a larger range of gain.  Leaving in the source resistor and cap in the schematic, I ended up with 1.3k from V+ to Drain.  Oddly, it doesn't seem to be simulating correctly.  The sim shows a nice large symmetric signal before the output cap and after the JFET, then a much smaller signal after the output cap...not sure why.

I am still very interested to know how I made a half-wave rectifier. My knowledge of rectifiers is pretty limited to tube rectifietrs and solid state bridge rectifiers (full-wave) in amps - and when I say limited, I mean very limited.  I understand what they do in general and applications as it relates to classic amps but that's about it.
  • SUPPORTER
I'm a rookie:  Teach me your ways, wise sensei!

Twhjelmgren28

Just bumping this post up.

Hoping someone can shed some light on my question directly above.  Wondering what constituted the half wave rectifier in my schematic and what that essentially means for the circuit.
  • SUPPORTER
I'm a rookie:  Teach me your ways, wise sensei!

11-90-an

full wave rectifier output: up wave - down wave - up wave -down wave- ...

half wave rectifier output there can be 2 options...
either
up wave, no wave, up wave, no wave, ...
or
down wave, no wave, down wave, no wave.

essentially, 1 diode by itself can be a half-wave rectifier already...

hope that helps... :P

I'm also not quite sure about what @EATyourguitar said...
flip flop flip flop flip

Twhjelmgren28

Quote from: 11-90-an on September 23, 2020, 09:21:49 AM
essentially, 1 diode by itself can be a half-wave rectifier already...

hope that helps... :P

I'm also not quite sure about what @EATyourguitar said...

So if I'm reading you correctly, since I have 2 diodes I'm NOT running a half-wave rectifier? 

I understand the bare minimum of what a half-wave vs. full-wave rectifier is - I'm just not sure how it came into play in my schematic.
  • SUPPORTER
I'm a rookie:  Teach me your ways, wise sensei!

antonis

Quote from: Twhjelmgren28 on September 23, 2020, 10:16:09 AM
I'm just not sure how it came into play in my schematic.

I can ensure you for no other person does..  :icon_wink:
(no half or full wave rectification here..)

By connecting clipping diodes to Vref instead of GND you raise clipping threshold up to 4.5V + Diode forward voltage drop so you result into higher output voltage (Volume) and less distorted signal..
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

Twhjelmgren28

Got it!  Thank you Antonis.

Appreciate the info - that is exactly what I was trying to wrap my head around! 
  • SUPPORTER
I'm a rookie:  Teach me your ways, wise sensei!