Question about the little gem

[Joecool85]

On runoffgroove.com

Is the 25ohm rheostat the volume control or is that just connected to the ground and there is a pot doing the volume control?

[RLBJR65]

http://www.runoffgroove.com/littlegem.html
Yes the rheostat is the vol. cotroll.
Pots and rheostats are similar but there are many differences.
Maybe these links will help.
http://www.kpsec.freeuk.com/components/vres.htm
http://www.fanbus.com/thelab/Rheo_Tips/rheostat_tips.htm
http://www.geofex.com/Article_Folders/potsecrets/potscret.htm

Richard

[Mark Hammer]

ANY power amplifier will deliver its output to as many speakers as you want to tack on to the output, whether in series or parallel.  Assuming the combined load is not so low that the amp burns up from the heat, the amp will distribute the power across the various portions of that load depending on their impedance.  You've likely read specs on amps and seen that the amp will supposedly deliver twice as much power to a 4ohm load as to an 8ohm load.  Well, if the amp was looking at a 4ohm and 8ohm at the same time, it would not split the power up evenly between them.  Since the lower load presents less of an impediment, the amp will naturally deliver more current to that side.

In the Little Gem, the amp sees a couple of impedances on its output.  One of them is the so-called Zobel network made up of the 10R/.047uf network.  It also sees another set of impedances made up of the series resistance of one side of the 25R pot (the leg leading up to the wiper), and then the parallel resistance/impedance of the speaker and whatever the resistance of the other leg of the pot is to ground.

If the rheostat is up full, the Gem sees an 8ohm speaker in parallel with 25ohms of fixed resistance.  Naturally, it will deliver the lion's share of its output to the 8ohm path.  The parallel impedance of those two paths is going to be less than 8ohms.  If the rheostat is turned down a bit, say by 10 ohms, then what the amp sees is 10ohms, and after that 8ohms in parallel with 15ohms.  The output is still disproportionately in favour of the 8ohm speaker, but it hogs less of the output current than it did when facing off against a 25ohm alternate path for the output signal (where the speaker would win hands down).  Not only that, but where the amp was previously looking at the parallel resistance of the speaker and 25R (by definition less than 8ohms), it now has to fight 10ohms before itever gets to the paralel resistance of 8ohms and 15R.

So, as the rheostat gets turned down, two things happen: 1) the amp itself sees a higher load, hence delivers less power, and 2) a larger share of that output is diverted to the path between the wiper and ground along the rheostat.  As the wiper passes the point where the leg of the potis less than the impedance of the speaker, more output gets turned into heat in the rheostat than is turned into cone movement in the speaker.

Make sense?

[Paul Marossy]

I was going to say earlier that it was like a master volume control, but wasn't really sure. Is that kind of like what it is doing?

[Joecool85]

Seems to make sense.  Any reason you couldn't use a pot though?

[vanhansen]

So, as the rheostat gets turned down, two things happen: 1) the amp itself sees a higher load, hence delivers less power, and 2) a larger share of that output is diverted to the path between the wiper and ground along the rheostat.  As the wiper passes the point where the leg of the potis less than the impedance of the speaker, more output gets turned into heat in the rheostat than is turned into cone movement in the speaker.

Make sense?

This last bit sounds like the rheostat is really a very low voltage attenuator, to put it in layman's terms.

[Paul Marossy]

Any reason you couldn't use a pot though?

I think so - the current going thru it would burn up a regular 1/2 watt pot.

[Joecool85]

K, well that makes more sense now.

Now here is a big one, what ohm rating speaker should I use?  the mkII goes down to 4ohm it says, but this is just the regular little gem.  So is 4 ohm ok, or would 8 be better?

[Mark Hammer]

The speaker impedance selected would depend on the chip used, and the supply voltage.  If you look at the datasheets for the 386 series and their equivalents, you'll see things like power output into so many ohms across a range of supply voltages. 

Since the chip is not designed to tolerate tremendous current draws, and since output power will be a function of the speaker load and supply voltage, you'll see indications in the datasheet that a 4 ohm speaker with a 6v or possibly 9v supply isn't such a bad idea but you'll also see things like specs for an 8 ohm speaker at 12v supply and nothing about 4ohms at the same supply voltage.  The gist is that since the chip might attempt to deliver an unsafe amount of current at 12v to a low speaker load, so you should use larger speaker loads when powering at the higher end of the recommended supply voltages.

The LM386 series has several different suffix designations: 386-1, 386-2, etc.  The 386-4 is rated to be able to deliver much more wattage/power (safely) at its highest supply voltages than the more common 386-1.  On the other hand, those ratings assume absolutely NO heatsinking, so you may be just fine with a 4ohm speaker, a 12v supply, a garden variety 386-1, and some heatsinking.  You can easily fabriate usable heatsinks for chips from little bits of aluminum and squirt a teeny bit of the thermal compound that came with your CPU fan between the chip and heatsink.

Again, always remember that the volume will not correspond precisely to the impedance.  Yes, the amp candeliver more power to a lower-impedance speaker, but a 12" 8-ohm will always be louder than a 3" 4-ohm, and an efficient 4" 8-ohm in a decent cab that retains bass can easily be much louder than a same-sized, but less efficient, 6" 4-ohm in a tiny cab.

[Joecool85]

Alright.  Well, I guess now is the time to start digging through some old radios and stuff and see what I can grab for a speaker.  (I can't find a good place to buy a tiny speaker anyway.)