Oscillation in RAT - Some tracks

[LaloFP]

Hi guys!

Im building a RAT and my protoboard version is having the popular oscillation among DIY RATs.

BUT I start this thread because Im facing something weird: I read a lot about the RATs oscillation and everybody said that the problem was on the opamp feedback, and it seems logical, but I can reproduce and stop the issue in other section.

The issue was the same, in the high gain part of the DIST pot, the pedal started to introduce an oscillation and the DIST pot moveing changed the pitch.

Then, I noticed that moving the protoboard changed by little the oscillation pitch too, so I started to moving every component and disconnecting some of them. Then I realised that disconnecting a cable to the hard-clipp diodes stopped the issue at all.




I repeated the test to ensure that it was really affecting the scenario and it was.

Then I changed the diodes position and the issue dissapeared too: (then I put everything in the other position and the oscillation was back...)



BUT THEN!!!! I realised that if I press on the white cable that is on the left, in some positions it inserted oscillation too, but with a higher ptich. I pressed very hard on the protoboard first to ensure that the only thing that I was actually moveing was the cable and not other component.

So, in my test I can get a RAT pedal with and without oscillation, only by changing a few components position fisicaly .

Why is that? Are all those DIYers with oscillating RATs getting the issue only because of a fisical design and components disposal?

I can make a video if you wish.

Thanks!!!!!!!!!!!!!!!!

[Brisance]

Parasitic capacitance.

[GibsonGM]

Every (unshielded) component leg and wire run is a radiator of energy.  Every pulse of current through them results in a burst of 'signal' being radiated into nearby components - this is electromagnetism we're dealing with at heart.  Get the phase and amplitude right, and you get oscillation.   Add in the parasitic capacitance of the darn breadboard...the metal plates inside it act like small caps and become part of your circuit.  Caps change the phase angle of a signal, so provide that end of the oscillation deal.

People regularly use an actual chopstick, or piece of wooden dowel, to move wires around inside amps to see if they can remove/diagnose this problem.  It's why everyone is always talking about "layout" on here - how you lay the parts out can REALLY affect how they interact (or hopefully not).  The higher the gain you are using per stage, and across stages, the higher the chance that you'll COUPLE your signal into a place you don't want it.  And distortions like the Rat are pretty high gain      

You just rediscovered a very cool electrical phenomenon, Lalo     You basically answered your question, yes....physical design, layout...pushing a simple circuit to its limit, make the RAT easy to experience oscillation.

Even cooler - some things oscillate so rapidly we can't hear the actual problem, just its symptoms (VERY crappy tone, ugly harmonics...), and some oscillate so slowly that the circuit 'breathes' in a very odd way...

[LaloFP]

Oh GibsonMG, thank you!!! a really nice explanation 

I was planning to draw my own layouts, but now I realize that I need to learn much more! (and I like it)

Thank you GibsonMG 

[GibsonGM]

More than welcome, Lalo!   Keep with it, the little lessons like this will eventually start to make sense and you will just do things "well" because you started out, made some errors, and learned.   We all do.    Even an old hand sometimes has a feedback problem, and it can be annoying.

[karbomusic]

Which opamp did you use? I only ask because I blindly laid out my own RAT PCB a few weeks ago and have seen zero oscillation issues, nada, none, zilch and it has even more gain that a stock RAT. I ask because I'm hesitant to think I should be so lucky and think more could contribute than just layout; I'm suspicious that some other factor contributes to the sensitivity of the layout and moving wires isn't actually addressing the root cause.

[LaloFP]

Karbomusic, I used the LM308N, but I also bought the OP07 and TL071 to test the audio diferences.

I didnt test that yet, Im going to try tomorrow and tell you if the IC changes something on the oscillation!

Also, Im more interested in the hole effect of layout and all the stuff that GibsonMG told about, because like he said in this case is a notorious noise, but I dont want to make subtle changes in my pedal sounds because of ignoring this topic. And I think that the stock originals maybe got "layout problems" that affect and make the sound that I want.

So I definitely need to learn all of it haha

And thank you again GibsonMG for the encourage! 

[duck_arse]

that GibsonMG is always very helpful.

Lalo, I can't see you pictures at the moment ( ), but if you clip the diode leads to half the length I last saw them, and swap/re-arrange their positon on the BB with the filter pot, you might solve some of yr osc probs. see how far karbo has his diodes from the op-amp?

and if that doesn't work, use a vise like karbo.

[karbomusic]

Also, Im more interested in the hole effect of layout and all the stuff that GibsonMG told about

By all means do and don't allow me to distract from that. I'm somewhat new myself so wanted to add a few questions thoughts.  On a side note, I was wondering if a non LM308 chip might lend to the issue so I learned something already.

[LaloFP]

Also, Im more interested in the hole effect of layout and all the stuff that GibsonMG told about

By all means do and don't allow me to distract from that. I'm somewhat new myself so wanted to add a few questions thoughts.  On a side note, I was wondering if a non LM308 chip might lend to the issue so I learned something already.

Oh no! that doesnt distract at all!

Im so angry... I just was testing with the opamps and... I DONT HAVE A LM308... I trusted the seller when he told me that he got some... is a LM307...

Anyway, the oscillation is the same with the OP07 (a bit different, more in the highs, but pretty much the same)

A question: In your image I see 2 transistors. Did you use this schematic? because it has some errors:
http://www.muzique.com/schem/pcrat1.gif

that GibsonMG is always very helpful.

Lalo, I can't see you pictures at the moment ( ), but if you clip the diode leads to half the length I last saw them, and swap/re-arrange their positon on the BB with the filter pot, you might solve some of yr osc probs. see how far karbo has his diodes from the op-amp?

and if that doesn't work, use a vise like karbo.

Thank you Duck_arse, but I see that in the karbo´s RAT the diodes are next to the op-amp. Or I didnt understood you

And I moved the diodes to the other corner of the BB and the results are the same, but weird: I got the oscillations if I connect the diodes to the bottom ground line of the BB. If, in the same position, I connect the diode´s leads to the upper ground line of the BB the oscillation is gone (Like in the image I posted). Obviously, both ground lines are connected with the white cable I told that produced oscillation when moveing.

And what do you mean about the vise? A big piece of metal, not connected, can 'protect' me of that oscillations? (I think Im saying stupid things haha   )

[karbomusic]

A question: In your image I see 2 transistors. Did you use this schematic? because it has some errors:
http://www.muzique.com/schem/pcrat1.gif

Ugh, I don't remember... I peeled through schematics for a few hours as I built up mine in Eagle but very likely very close to, if not that. Where are the errors? There is another version of that schematic that is in the wild on the internet that looks just like it but the guitar input is on Pin 2 and the negative feedback loop on pins 1/3, that one is definitely wrong.

I also forgot, I'm using the UA308, I did use a LM308N right after I soldered the board up to test it on the scope, it then swapped it out for the UA308 just before I boxed it up:

[duck_arse]

lalo - I also built a rat many moons ago, using an LM307. and sold it to an old hand, so I wouldn't worry too much about that. just don't buy LM308 from that seller again.

and it is I say the stupids around here. karbo, and some others, often use their bench vise to hold a board for photo takings. I'm implying that the vise will stop yr oscillations, you see, if you squeeze hard enough.

and yes, I realise you can't get the diodes as close on the breadbord as karbo. but if you cut down their leads, and move them as close to the IC as you can get, then put the filter pot next on the BB, you won't have the long white wire to worry about.

[LaloFP]

lalo - I also built a rat many moons ago, using an LM307. and sold it to an old hand, so I wouldn't worry too much about that. just don't buy LM308 from that seller again.

and it is I say the stupids around here. karbo, and some others, often use their bench vise to hold a board for photo takings. I'm implying that the vise will stop yr oscillations, you see, if you squeeze hard enough.

and yes, I realise you can't get the diodes as close on the breadbord as karbo. but if you cut down their leads, and move them as close to the IC as you can get, then put the filter pot next on the BB, you won't have the long white wire to worry about.

now I understand
and yes, Im going to avoid long cables from now
thank youu duck_arse

Ugh, I don't remember... I peeled through schematics for a few hours as I built up mine in Eagle but very likely very close to, if not that. Where are the errors? There is another version of that schematic that is in the wild on the internet that looks just like it but the guitar input is on Pin 2 and the negative feedback loop on pins 1/3, that one is definitely wrong.

the main error is the input buffer stage (that isnt in the original), and then got other value diferences