how to get as much clean power as possible from 386 chip

[birt]

there must be a way to find this out without having to build a 386 amp right?

for example a ruby. what resistor do i need to use between pin 1 and 8 to get as much power as possible without distortion?

[Mark Hammer]

First off, you have to ask yourself about what the circumstances are under which a 386 might distort.  One reason is certainly the gain it is set to by the pins 1-8 loop.  But another is whether the power supply is sufficient for what one is attempting to do.  In many instances, folks are trying to get the sort of volumes out of it that seriously tax the power supply.

There is also the matter of what you can reasonably expect in bass response from a speaker small enough to be moved by that chip.  I suspect that despite OUR typical use here, it was not at all intended to be used with speakers over 3" or so, and "clean" bass in a 10" or 12" speaker is not all that likely.

It is also only fair to note that many of the implementations of the 386 as a mini-amp are pretty minimalistic.  There was a schematic/project in Electronic Musician some years back using a pair of 386s as a stereo headphone amp.  (I can post it some time over the weekend but in the interim try EM's archives.)  It was noticeably more complex than what we see posted as the Gem/Ruby etc, with more caps involved.

[Krinor]

What are you using the chip for ?
In my experience the Ruby with the "Bassman" mod (found at Runoffgroove) produces a nice clean signal when run at 12V into a 8'' speaker.

[bancika]

I built little gem mkII with two 386 and nothing between pins 1 and 8 and it has plenty of power to drive 12" greenback speaker and it's very clean. See how small it turned out

[dano12]

Just an addition to the great comments already posted, it also depends on which 386 you use. The National LM386 has several variants, each capable of handling different supply voltages. For example, the N4 can handle up to 18 volts.

Not sure of diffs on the JRC 386...

[birt]

i am using it as the driver for a telephone speaker in a pedal. (the speaker is taped to a carbon mic) the ruby has 2 volumecontrols and i want them both fixed at the highest clean volume possible so i wont have any distortion of the chip.

then i want to control the volume after the chip as a drive control to go from clean to a distorted speaker/mic. i think this gets more interesting sounds then a distorted amp going into the speaker.

it's also a way to go from gain and volume before the chip to drive (after the chip) and volume (after the mic)

it will operate at 9 volts

[ambulancevoice]

ive read that more power gives you more clean headroom with the lm386/ruby amp
also, why dont you try puting a 1k trimpot to replace the gain pot in there, put the volume max, and slowly turn the trimpot until it gets to its max clean vol and leave it set there

[Paul Perry (Frostwave)]

I can't agree with Mark about driving a large speaker.
The larger a speaker is, (usually) the more efficient it is, and the less likely it is to distort on peaks, especially bass peaks.
But yes, these 386s sound absolutely awful when clipping & usually it is the power supply voltage that sets the limit.
Mind you, getting the 'maximum' power from these tiny chips requires a VERY GOOD heat sink.
By the time you have given it enough heatsink, you might as well have used a more appropriate higher power audio amp chip - of which there are many.

[micro]

Anyone tried using a chip with more power like the LM384 ??

[birt]

i don't need that much power to drive a small telephone speaker i was just wondering if there was a way to calculate what fixed resistors to iuse intead of trial and error (trimpot) or if anyone has ever searched for the same thing. i'll use a trimpot and see what happens

thanks for your input

[Steben]

But yes, these 386s sound absolutely awful when clipping & usually it is the power supply voltage that sets the limit.

Well, others seem to claim they sounds superb, which is audible in many soundclips. But I haven't succeeded in making them sound like that. Rather awful up till now indeed... The latest I used was a FET buffer and now it sounds really clean-nice, but the gain is too low to get distortion. What's the secret behind those fancy sound clips (a hidden overdrive pedal)?

[bancika]

What speaker did you try it with. When I made my first ruby I tried it with two 4" speakers from PC and it was horrible. I turned it to little gem mkII when heard that distortion...didn't even give poor thing a chance to try with real speaker

[birt]

the project is stopped until i have 386 chips again. i fooled around with it a bit too much and its toast

[Paul Perry (Frostwave)]

the project is stopped until i have 386 chips again. i fooled around with it a bit too much and its toast

This would appear to be an argument in favor of using a chip amplifier - say a TDA series - that is designed to supply plenty of clean audio power, if that is what you want.
Certainly, many on this board have succeeded in making a tiny combined overdrive/amp from a 386; my own inclination is to separate the functions, even if it adds another cubic inch or so

[birt]

haha, i get your point. but i don't need that much volume you know. i just don't want distortion. i might as well use a transistor amplifier
now where are those OC22's i had? Add some Ge mojo :p

[petemoore]

  Get away from the LM386's distortion unless you decide it is needed.
  clean is a bit more trouble but then you can have clean to compare to distorted.
  9v goes only so far to provide amplifier supply, if you want bass/clean at 'x' volume, 9v soon becomes too small.
  Check the max input voltage on your chip, maybe look at a different chip that uses a split DC supply.
  See GGG, and GEO for amp and power supply info.

 

[Barcode80]

my ruby sounds GREAT overdriven, in fact my coil tapped humbucker is too hot to get clean on it. sounds excellent though.

[birt]

i found out it's got plenty of clean power for my purpose. i need to go check for other speakers tough. i need small (2" or so) speakers that don't buzz with the guitar sound. they don't have to sound good as long as they can handle the power. i think a low cut around 100Hz might also help to keep them from buzzing and eventually breaking.

[Elektrojänis]

The larger a speaker is, (usually) the more efficient it is, and the less likely it is to distort on peaks, especially bass peaks.

I agree with you...

It is an urban legend that you will need a bigger amp to drive a bigger speaker. It's simply not true. The speaker sensitivity (efficiency) is the key. Bigger speakers usually have a better sensitivity than smaller ones (there are other things too that affect the efficiency). If one tries several speakers with the same amp, the most sensitive speaker will be loudest. 12" guitar speakers are usually quite sensitive. Many have a sensitivity of 100dB/W/m (100dB with one watt of power when measured at the distance of one meter).

i was just wondering if there was a way to calculate what fixed resistors to iuse intead of trial and error (trimpot) or if anyone has ever searched for the same thing.

There is no way to calculate that if you don't know the level of the input signal. The gain doesn't actually affect the maximum volume level before clipping. It is always the same. With more gain, you get more output as long as it is not clipping, but you would get just the same result with less gain and more input. With guitar signal the input level depends on so many things that the trying it with a trimpot is really the only good way to go.

Well, others seem to claim they sounds superb, which is audible in many soundclips. But I haven't succeeded in making them sound like that. Rather awful up till now indeed... The latest I used was a FET buffer and now it sounds really clean-nice, but the gain is too low to get distortion. What's the secret behind those fancy sound clips (a hidden overdrive pedal)?

The secret is probably the speaker/cab used. Speaker shapes the tone. Some speakers shape the tone more and some less... Some shape it to sound better with distortion and some worse. Usually the ones that will not produce too high treble will sound better with distorted guitar sounds.

Different 386-chips might sound different too. If you compare the equivalent circuit on NJM386 and JRC386 datasheets, they look different... They might work the same under normal operating conditions (when not clipping), but it is a different matter how they work when operating conditions are not normal (clipping).

[Steben]

Well, I finally got a decent tone out of it. I swapped the fetzer stage with a simple fet buffer - couple cap - and straight into the 386, no more controls. And voila! - loud! and clean tone. The speakers are still the same though. 2 x 6" 4 ohms (= 8 ohms). Sounds a bit trebly, yet no buzz and misbiased cut-off any more at clipping. Really buesy and nice clean up with guitar controls. I love this little thing from now on. At all folks: don't give up on 386... find the error!
When I put a wool overcoat over the speakers, it even gets really warm.  Mojo filter...
By the way, it seems my volume pot on the guitar needed a replacement. Very odd.

A question: I used 2N5457 as fet buffer. It puts out about the same as the J201's I tried. Are there fet's that can be used and have higher output? I mean: closer to unity gain?