More Little Gem modification

[javacody]

I just couldn't get the amp to sound the way I wanted with the tonestack. It seemed to just suck the life right out of my tone. I removed the tonestack and made a few mods and now have the following:

Rev 1.4

This sounds pretty good except for a few things:

eardrum piercing highs
lack of lows
a strange undertone that sounds like an electrical hum whenever I play with the gain fairly high.
[/list:u]

I think that the lack of lows comes from the fact that my 8" speaker isn't in a cab yet, but any ideas about the other two things?

[Joep]

increase the 0.047 to taste, this one cuts the bass, since the input impedance of the LM386 in low. Try 220n  or so (0,22)

The LM386 tents to oscillate real quick, I put an 10n...100n cap across the +9v and gnd to fix this. This is causing the strange noise you hear...

Hope this helps,

Joep

[Joep]

I breadboarded a Little Gem a few day ago and I was really impressed with the sound. Real nice, low cost, experimenting with instant good results!

[B Tremblay]

Sometimes using the inverting input (pin 2) can stabilize the 386.  Remember to ground pin 3 if you do use pin 2.

[javacody]

I used pin 2 originally (as per your awesome project at Runoff Groove), but it caused phase issues with my buffer, so I had to switch to pin 3. (I had originally had the signal coming from the drain on the JFET, but switched it to the source when I had the original oscillation issue, but I wanted to experiment with some of Jack Orman's buffer ckts, so I switched back to pulling my signal from the drain and switched to the pin 3 for the input on the lm386.)

Also  Joep, I had that that I had large enough caps on the +9 to ground already? Or do you mean something different (see c1 and c4 in my schem).

[Peter Snowberg]

Keep in mind that larger caps will pass bass frequencies better, but they are not as effective as the frequency climbs. Years ago when the world was manly TTL logic, I would routinely use 3 or 4 caps in parallel to keep rails clean. Often this consisted of 10uF + 0.1uF + 0.01uF. Try adding a 0.1uF and see what that does.

Another thing to consider is the use of a coupling cap in the pin1 to pin8 loop. Use something big like 22 or 47uF. I'm not sure if upsetting the bias of the input stage with a DC path there has anything to do with it, but I always use a cap in that loop.

You may also want to try a series resistor between the buffer and 386. That will alter the gain structure, but it will cut the highs too. If you look at tube amps with big resistors in the stage coupling, you'll see bypass caps around them to restore the highs. With the low gain boost of your buffer you won't see as much of this effect because you have less gain to "throw away", but it's a good thing to keep in mind.

Once you get the shrill highs hammered out, you may want to try some other tone controls like the BMP. Now you see the reason I was suggesting bypass switches . If you want the control of a tone stack, you have to give it some extra signal to eat for the processing. The speaker and the enclosure it’s in designate much of your tone, but as you have experienced, the whole thing is a matter of balancing lots of things to make harmony.

Take care,
-Peter

[javacody]

Thanks for the help guys! All excellent advice.

Here is the revised schem with most of the suggestions:



I implemented the suggestions one at a time so that I could tell what the effect was. First was a 0.22uF cap on the rail (as this was the closest to 0.1uF that I had). I gave it a listen, definite improvement. Next, I added a 47uF cap  between pin one and the pot. Much better. This also really seemed to tame the highs for some reason? I still had way too much oscillation, so I gave what B Tremblay said some thought and switched the output of the buffer back to the source and then I moved the input of the 386 from pin 3 back to pin 2. Again, more improvement. I can almost turn the gain knob all of the way up now without the sound turning "farty". Last, I decided to add a 0.01uF cap on the rail (mislabeled in schem). Again, a minor improvement.

Peter, after the second mod, the highs were tamed significantly, so I elected not to use a resistor (yet!). I figure that when I build a cab for this speaker and put some grill cloth in front of it that much of the high end will be damped. That way I also don't lose any more gain.

Thanks to you guys, I have most of the problem licked, but do you have any more tips to clean this up a bit more? At almost full gain, it sounds like a misbiased fuzz face, only grittier.

Oops, I just realized I forgot to ground pin 3. I'll try that next and see how it works.

[Peter Snowberg]

One more filtering thing to try... Do you have a cap on pin 7 at all? That would be a good place to put another 47uF, or anything really. I can't say how it will affect your tone exactly, but it may cut down on the oscillation.

You're now firmly into the territory where a scope would be of much benefit. A picture can be worth 1000 words when it comes to ultrasonic oddities. Pick up a used one if you can spare the cash.

I'm going to be watching intently for what happens when you put the speaker in a closed back cabinet (assuming that's what you're doing).

Take care,
-Peter

[javacody]

Peter, B Tremblay's advice about grounding pin 3 was dead on. It took care of the rest of the oscillation.

Also Peter, I'm thinking of doing a 4 8" speaker cab (sort of a mini-bassman). Would you recommend that this be closed back? Also, how would I wire the 4 8 ohm speakers for 8 ohms total?

[javacody]

Also, I can't thank you guys enough for your help, especially Peter. I've learned so much about all of this stuff and had a blast doing it!

I can't wait to work my way up to a sansamp G2!

[Peter Snowberg]

You're very welcome. Now just relize the flip side.... somebody else is going to come along with the same questions you had; then it's your turn to pass the knowledge along. I have Forrest Mimms to thank for my 386 tinkering.

I know just about zip when it comes to speaker cabinets so all my stuff is open back (yes, I'm cheating ). I only know what I've found empirically when it comes to boxes, but I've noticed the bass going way up when you're dealing with small speakers in ported enclosures.

For wiring four, 8 ohm speakers.... make two parallel pairs (4 ohms resulting per pair), and then put those pairs in series to get back to 8 ohms.

That’s going to be a great unit!

Take care,
-Peter

[javacody]

Duh, just like 4 resistors!    I'm a little slow on the uptake sometimes, but I catch on eventually.

[b_rogers]

i have learned on my first 2 effects (FF&smash drive) how important routing of wires is. if you have a rats nest it WILL oscillate..both pedals did this at full drive settings until i shortened and rerouted the wiring..neither of my projects would  have been possible without help from the great guys on this board..thanks again