Can a "Ruby" style amp be made sparkly?

[liquids]

Mark - I'm not totally clear on what you are saying in your post.  But my limited knowledge of audio is that you need exponentially less power to drive ahigher frequencies - and frequency and volume are, to the ear, exponential, not linearly related/perceived. 

My wife's laptops speakers, tiny, driven by a little bitty preamp/amp, drives me nuts, because all it can produce (or at least, to the ear) from an arms length away are loud obnoxious high-mids and highs.  Those frequencies shoot across the house when doors are open.  I tell her to plug it into the hi-fi, please, or (now with Win 7) go in and add the 'bass boost' EQ setting so it evens it out a little more...only a little.

Bass frequencies , on the other hand, need a lot of wattage (and speaker cone area, relative to distance from the speaker) to produce the same apparent volume of 1W driving 'only' high frequencies.

Anyhow, sparkle should not be an issue with the ruby amp.  If it's a hi-fi amp chip, it should be able to reproduce all audio frequencies decently or per datasheet spec.  As for speakers, no guitar speakers produce much above 6k anyway, harsh/sparkly/bright/ice-pick fender amps included.  Normal speakers reducing sparkple shouldn't hinder a thing until you start driving instrument amp speakers designed to be dark or bigger than 12"....or car audio speakers that are small but designed to be woofers or sub-woofers.

I think the issues here is that the ruby amp can serve two purposes for guitarists around here as an amp - one, a low wattage power amp, that is EASILY driven to clipping at low volumes that don't piss people off 3 houses down.

Under clipping, the fact that there is no mid scoop passive EQ that we are all used to between guitar pickup and amp keeps gain high, and makes the distortion produced  more pleasing, since it's comparatively mid-heavy (flat), so it's not percieved as harsh. 

If you want clean and sparkly, and you plug your guitar into a ruby amp, you get a polytone/jazz amp - flat and dark. All the more so if you aren't driving it with a low output impedance source. Why?  Because the low input impedance of the ruby is going to suck the highs right out of your sound, and also, because there's none of that mid scoop EQ we're all used to having in a guitar amp and is oftentimes inherent even (or unless) you dime the mid pot.....unless you add one of those passive EQ circuits we all have in our amps before the (flat EQ) ruby circuit.

[liquids]

(double post, sorry)

[Mark Hammer]

What I'm saying is that ultimately the guitar is attempting to move a physical object via the Ruby.  It takes a certain amount of current produced by the amplilfier to overcome the inertia of the speaker piston, and producing high frequencies requires that there be enough power to produce near-instantaneous acceleration.  Clearly it takes more horsepower and time to get an 18-wheeler up to 60mph than it does to get an SUV up to that speed, and more horsepower and time to get the SUV to accelerate to that speed than it takes for a Lexus coupe, and more for a Lexus coupe than for a (insert favourite small 2-seat Italian sport convertible here).  In the universe of speaker cones, a typical 30-60W 12 inch speaker is like an 18-wheeler, and a 3" is like a Chevy Cruze.

It DOES take more wattage to reproduce lower frequencies...in their proportion to the rest of the signal.  Sluggish speakers will easily be coaxed into moving slowly at modest power, but there won't be that THUMP unless there is sufficient power to make the cone move instantaneously to the initial low frequency transient.

It is also true that HF content does not require as much power to reproduce in their true proportion to the overall signal, but you still need a minimum of power to overcome speaker inertia.  In hi-fi systems, that problem is addressed by dividing the frequency content into larger and smaller drivers.  The tweeter and/or midrange, are lower-mass drivers, and require less wattage to accelerate appropriately and reproduce the content in its true proportion.  It still probably requires more wattage than it ought to, ideally, if only because smaller drivers may be lighter, but they move less air, compared to a big honking woofer.

The average 250mw-1W micro-amp, whether it is a 386-based Ruby, a Smoky, or a tube-based ZVex Nano-Amp, will sound bigger into a larger cab, feeding a larger speaker, if only because the extended range of a 12" in a proper cab has more body than a 3-5" in a puny box.  But you won't get the thump produced when feeding the same cab at least 5-10W.

On a distantly related note, however, I'll mention something I've noted here previously.  There is no reason why a person has to use only ONE path between pins 1 and 8 on a 386.  Internally, there is 1.35k fixed resistance between pins 1 and 8.  If bypassed with a 10uf cap, one gets a boost in gain across the spectrum.  But that's a boost for the whole spectrum.  If one bypasses the internal resistor with a smaller-value cap, that boost is applied to a smaller portion of the spectrum, and the smaller the cap value the more the boost is restricted to higher frequencies.  I first stumbled onto this in an Nuts and Volts construction article from the late 90's, that used a pair of 386s and bypassing with a small cap to yield a crisper sound for people with hearing loss who wanted to listen to the TV.

The upshot is that one could link pins 1-8 with a 10uf cap and a series resistor (though less than 1.35k, obviously) for some full spectrum boost (but less than the full 200x), and then nudge the highs more with a smaller-value cap linking 1 and 8.  So, say a 10uf+220R, in parallel with a .1uf cap.  There will still be the limitations of the chip I described earlier, but you should get more bite.

[liquids]

I understand what you are saying much more clearly now.

And cool point about using pin 1/8 like the emitter of a BJT/inverting pin to center rail connection of an non inverting op amp for frequency control...given the internal resistance set at ~1.5k...

hard to know how much the effect of a 10uF & 220R in series between pin 1 & 8 would 'reduce' overall boost as compared to just strapping the cap straight across...might be wise to just start with no connection, and begin with a small value cap straight across the pins (start with 1n or something) and increase until the high end frequency range is noticably enhanced - decrease from 1n to find out how small is 'too small' a cap to hear any difference for reference...then increase the cap value until it either is too harsh to you or you're more or less approaching a full frequency boost...an educational process.

Then, if you did reach a point where Cap value X was 'as much added highs as I can take,"  and you still wanted more gain, yeah, find a 5k or 1k pot+series resistor or something...rig it like a variable resistor (tie the wiper to one of the outer lugs), put the pot it in series with a BFC (47uF+) between pin 1 and 8 and dial in an 'full frequency gain boost' from max pot resistance to min resistance until you don't need any more....and there you go...if you want to leave it there, measure the pot and sub a resistor, or keep the pot pat of the circuit....
Should note that invariably the big freakin cap method to get gain - if it works, as it sounds, like a bypass cap on a cathode/emitter/etc, will affect how the chip itself enters clipping when driven...for better or worse...that is, driving the input with more gain but no cap will sound different than driving the input with less gain but with the cap, all else being equal...but a useful experiment.  I'd avoid the big cap if you're after a mostly clean low volume tone, however.

There's more than one way to 'skin this cat' (and where the heck did that cliche phrase come from?), but that's a fine and interesting one.   

If high frequency boost (via smaller value cap between 1 and doesn't yield the sparkle desired, a mid scoop/passive EQ route aforementioned is a good next approach that is less a subtle tweak and more an complete re-voicing, I'd suspect.

[liquids]

Also - I should look at datasheets and search before I ask this - but is it possible that there might be some benefits to 'stacking' two of these chips? 
Maybe you'd be able to drive lower impedance loads, or drive the same load more cleanly, or get an increase in wattage? 
Or just blow up both chips?