oscillation in Ruby amp? help!

[mcnowinski]

All,

I recently constructed the Ruby amp according to the wonderful article at runoffgroove. I have two 3.2OHM speaker attached in series to the circuit. From the beginning, I have noticed some distortion from the speakers even at the default gain setting (20). It only gets worse with increasing gain.

Yesterday, I finally had a chance to take a look at the signal on my o-scope. The source looks fine, as does the output of the JFET buffer. The output of the LM386, though, has some distinct strangeness, particularly on the negative portions of the dynamic signal. Upon closer look, I can see a much higher frequency component superimposed on the main audio signal, especially when it dips below 0V.

Is this the LM386 oscillating?

I have followed the schematic exactly as published at runoffgroove, but have seen some other LM386 applications on the web show higher cap values to ground at pin 7 (BYPASS), as well as some that apply RC filtering networks to the batter supply.

Any ideas would be greatly appreciated!

Thanks,

matt

[B Tremblay]

You can add the RC network between pin 5 and ground, as seen on the Little Gem schematic.  That may cure the oscillation.

What type of 386 are you using?  Also, how are you powering the circuit?

[mcnowinski]

B Tremblay,

I'm using the LM386-3, and it is currently being powered by a single 9V battery.

I will try to add the filter at the amp output to see if it improves the situation. Thanks for the tip!

-matt

[B Tremblay]

Is it a fresh battery?  Are you sure that the unused input (pin 3) is grounded?

[mcnowinski]

B Tremblay,

The battery is fresh and pin 3 is grounded.

I'm happy to report that by adding the 10OHM resistor and .047uF cap to ground from the LM386 output pin 5 (as shown in the Little Gem and the LM386 app notes) the oscillation has gone away completely.

Thanks for your help.

-matt

[B Tremblay]

Great!  I'm glad it worked out for you.

[Ge_Whiz]

A good place for me to put in my tuppence worth, as I promised Brian and Gary some feedback on my build of the Ruby.

I was a bit disappointed with the higher-than-expected distortion levels from this amp, but I had fiddled about with a few component values in the preamp stage and was running the circuit from a 12V supply (a small gel accumulator), so I couldn't claim that I had reproduced the circuit as given. Still, it worked despite the distortion and I used it quite happily yesterday on a friend's boat for nearly three hours on its small battery, which still has juice to spare.

Today I read this thread and went straight off and added the resistor/capacitor combination at the output and the result is 100% improved. Now I have full control over the sound, and it's a real beaut.

Brian and Gary, I would strongly recommend that you at least add a footnote about this mod. The rest of you, if you haven't got around to powering your portable/practice amps with small 12V gel batteries, try it soon.

[B Tremblay]

Thanks for the report, Phil!  I have been contemplating whether to add the RC to the Ruby schematic, so it's good to hear that it was necessary in your build.

[newfish]

Supreme bump on this thread!

Just built one of these tiny beasts (case is drying as I type this...)

I too had oscillation ight from the get-go when both 'Gain' and 'Volume' were up at full, so I thought I'd post a couple of observations...

In trying to get rid of the oscillation, I tried using pin 3 of the LM386 as the input - and immediately lost the clarity I had with pin 2.
I grounded pin 2 during this trial - otherwise I'd have got nothing, right?

I've managed to get of oscillation on all but the very end of the controls by the following...

Reduced the 3K9 to 2K2 - since there was no oscillation with the volume turned down, why not drop the volume a fraction?

Increased the Pin-5 RC network's resisitor gradually until I found something more useable (in this case, 1K2), and finally added a 680R Resistor in-line with the wire coming from pin-8.

This has pretty much stopped my oscillation (and I'm using a fresh 9v battery too!) - and still gives a great sounding headphone / bench amp.

If these changes help anyone, that's all good.

Thanks to Runoffgroove for providing schematics, and to Aron for this splendid list!

[grolschie]

...I've managed to get of oscillation on all but the very end of the controls by the following...

Reduced the 3K9 to 2K2 - since there was no oscillation with the volume turned down, why not drop the volume a fraction?

Increased the Pin-5 RC network's resisitor gradually until I found something more useable (in this case, 1K2), and finally added a 680R Resistor in-line with the wire coming from pin-8.

This has pretty much stopped my oscillation ...

Replying to an old thread. Did any of these mods change your tone? Was the oscillation heard as a crackly distortion? Thanks in advance.

[newfish]

Hi,

The oscillation I got was more of a <clears throat> "HOOOOOONK!" whenever I played a note.

Not pretty at all.

For what it's worth, I made another Ruby from scratch on a knocked-up-in-my-shed PCB, and all oscillations have gone.

Might have been my ropey layout on Vero.

New PCB version is running from 12v, and works as well for my 'leccy Violin as it does for the Telecaster.

Also drives my Friend's 2 x 12" a treat.

I can post PCB layout if anyone's interested?  It's a hand-drawn affair, but I'm willing to share...

[grolschie]

Thanks for that. Yikes. Mine's not that bad, just crackles on the decay somewhat, well maybe it's crackling all the time but I only hear it when the note fades. Made on vero with an LM386N-1.  I wonder, does the IC type affects things?