Mark Hammers LM380 Mini Amp

[salocin]

Hi,

I've been playing with the LM380 chip, and have been taking elements from Mark Hammers LM380 Mini Amp schematic (http://hammer.ampage.org/files/Miniamp.png).
I have two questions relating to this schematic that I was hoping somebody could answer for me:

1) If I were to change the 3.9uF Cap in the gain loop of the preamp section, what effect would this have? 3.9uF seems to be a non-standard value (at least down here in Dunedin, New Zealand) so obvious substitutes are 3.3uF and 4.7uF values.

2) The schematic says to ground pins 3,4,5,7,9,10 and 11, though I'm pretty sure this should be 3,4,5,7,10,11 and 12. This is just a mistake on the schematic right? (Just looking for a sanity check here...) Datasheet for the LM380: http://www.national.com/mpf/LM/LM380.html

Cheers,

Nic

[slacker]

welcome aboard

Changing the 3.9u caps to 3.3u or 4.7u won't have any effect on the sound. It will technically change the frequency response slightly but the frequencies affected are well below guitar or bass frequencies so you won't hear it.
Looking at the datasheet I think you're right pins 3,4,5,7,10,11 and 12 should be grounded. On the datasheet pin 9 is shown as NC or no connection, which should mean it's not used for anything inside the chip so there's no harm grounding that as well.

[salocin]

Thanks very much.

Now I've just got to wait 7-10 days for the DIP Heatsink to arrive before I can test the thing without fear of frying it!

[Mark Hammer]

welcome aboard

Changing the 3.9u caps to 3.3u or 4.7u won't have any effect on the sound. It will technically change the frequency response slightly but the frequencies affected are well below guitar or bass frequencies so you won't hear it.
Looking at the datasheet I think you're right pins 3,4,5,7,10,11 and 12 should be grounded. On the datasheet pin 9 is shown as NC or no connection, which should mean it's not used for anything inside the chip so there's no harm grounding that as well.

I put together a combo 380-based amp and Highway 89 preamp for my son's friend as a birthday present a few weeks ago, and perfed the whole thing.  Initially, the 380 ran very hot and the signal level was very low.  Then I too noticed the discrepancy between the PCB, datasheet, and drawn schematic accompanying the article.  The schematic drawing is incorrect and the PCB layout IS correct.  In addition to pin 7, the pins to be grounded are the middle 3 pins on each side: 3/4/5, and 10/11/12.

And slacker's comment about the cap and bass response is correct.

[Morocotopo]

Follow Mark´s advice. That is, follow the layout.
I made this amp, but I used a LM384 chip. It´s the same pinout as the LM380, but 5W instead of 2,5W. It works, I powered it with a 18V pwr supply. It needs a heatsink, I made one from 1mm aluminum sheet, similar as the one shown in the datasheet.
At first the chip got very hot very fast, I had ultrasonic (and not so ultrasonic) oscillations. I cured that with a 330pF cap between pins 2 and 6 of the first opamp. It´s a clean amp, although it has (I think) clipping diodes in the preamp section, it sounds about the same with or without them. I used TL071 opamps.
Trough a 8" guitar speaker of so-so quality (not very efficient), it sounds great on clean, not very loud, nice high end and lows, a bit fizzy with a od pedal in front. I tried a lowpass crossover between amp and speaker, (inductor and cap) but the sound got "dirtier" in a bad way.
Trough my Fender hot rod Deluxe´s speaker (12" Eminence) it sounded WAY WAY louder, darker, more body and oomph, much better with a od pedal in front (as a matter of fact, really nice  )
Remember, I used a 18V pwr supply (if you do, make sure all your parts can withstand 18V, especially caps) and a LM384 chip, so your results may vary.
One thing to take into account: maybe you can tailor the freq response with the zobel network at the power chip output (at least i think that´s what the 1R and 0.1uF cap are...) to adjust the amp according to the speaker used.
Any suggestions Mark? 

I get a little hum, but I didn´t box it up yet, maybe that´s the cause.
I´d post a picture, but I don´t have a digital camera, not even in my cell phone...

Morocotopo

[Mark Hammer]

The 384 works with a higher range of potential supply voltages so maybe stick with 12V for the 380 for the time being.  I use 12v because it is an octet of C-cells, and I can buy them at the dollar store in 4-packs for a buck.

Mine has a 6" or 6-1/2" Marsland speaker that is VERY loud.  So, although a bigger speaker will move more air, one doesn't HAVE to think purely in terms of bigger speakers. A modest-sized speaker in the right cab can have a surprising bass punch.  I suspect most folks who think about small power-chip "practice amps" base their opinions on commercial products that tend to:

a) use open-back cabs instead of closed-back or ported cabs (less wood and shipping weight)
b) have a shallower profile than they might (deeper cabinets can produce more bass, but the packaging ends up costing more)
c) use speakers that are suitable for a broad range of output ratings (and sometimes a little too stiff for a lower-power amp)

Really and truly, your best friend might be in understanding a bit about tonestacks and how to tune/taper/revoice them to your needs.  Whipping up an improvised cabinet with a cheap speaker you found MAY get you a decent tone, or may get you something a little off-base for your tastes/needs.  If you can adjust the tonestack to give boost and cut in the places where that cab/speaker combo needs it, you be in good shape.

Personally, I don't know enough about Zobel networks to monkey with them.  I leave 'em as is, but that's me.

As an aside, Thomas Henry had a project in EM some years back for a "studio" amp using a 384 and a front end not unlike the Pocket Rockit and a zillion other gain/master/3-band amps.  He also used a passive crossover and woofer/tweeter combination.  I can post the schematic if there is enough interest.