Ruby at 12V and through a 4ohm speaker??

[solarplexus]

Hi there,

I'm wondering if I'll fry the chip if I run the ruby on 12V powersupply and through a 4ohm, 35W RMS, 6,5" pionneer car speaker..... ?? 

Thanks,

Matt

[B Tremblay]

No, you won't fry it.  I'm using a 12V supply in my MacRuby and that has a 4 ohm speaker.  However, a car audio speaker will not provide the best sound.  Use a guitar speaker if you can.

[solarplexus]

Thanks for the quick reply!
well... I got a 6x9 pioneer as well 

I don't have any amp speakers available... I'll manage with this for now.

How about 18V (2 x 9v batteries)

would that fry it ??

thanks,

Matt

[remmelt]

My 386 was rated for 12V max. Check google for a datasheet of your specific chip.

[solarplexus]

it's a lm386 from smallbear so I guess 12v max 

how about 220V ?? 

12v it is.

Thanks for your help guys...

I will try to find a decent speaker though.

[Mark Hammer]

You have a couple of things to juggle:

1) There are several versions of the LM386, each with a different suffix.  The 386-4 can comfortably handle the most power output. Twelve-volt operation into a lower load is porbably grounds for using that one, rather than a 386-1.  Not sure about the power handling of the JRC version which has no power-related suffix.

2) You can devise some sort of heat sink such that the chip can easily survive whatever power output you demand that it provide into that load.  The object is to make sure the chip is not being asked to do something it can't without self-destructing.  Drawing, and passing, oodles of current, which is what happens when the supply voltage goes up and the load impedance goes down, generates heat.  If the chip has to handle ALL of it by itself, trouble may ensue.  If the chip has help getting rid of the heat, it can do more work.  Just be sure that you have good thermal transfer from chip to heatsink, that you provide space on the board for a decent heatsink (2 perfboard holes of clearance around the chip shold be fine), and that you don't inadvertently touch and short anything with the heatsink.

3) The output power is a function of what the input signal level and gain loop tells the chip to produce.  You can keep the input signal modest and the gain factor restrained, and expect the chip to be comfortable with that supply voltage and output load. You can adjust the amplification factor downwards, via the gain loop (pins 1 and 8 ).  See the datasheets for the 386 to see how to do it. There are reasons for wanting to run the circuit at 12v besides max power.  One of them is battery life, which will be longer when the supply voltage starts out higher in the first place.  So, if you can run it at 12v, without compromising the chip in the process, you're ahead.

[solarplexus]

Thanks Mark and everyone else.
Will check which kind of chip is the  386.
(the board is already soldered, just didn't have time to finish the wiring)

[nico13]

Here are some different LM386's specs:

                                      ABSOLUTE MAX VOLTAGE RATING               OPERATING VOLTAGE
LM386N-1 / LM386N-3            15 V                                                         4 - 12 V
LM386N-4                              22 V                                                         5 - 18 V

[sfx1999]

If you need a good guitar speaker:

http://www.tubesandmore.com/

Just remember, speakers need to be broken in.

Looks like a Jensen MOD was used on the Mac Ruby.

[Steben]

Since the thread cuts into 386's:
With the same voltage, do different types (1-4) sound different?
Or is the awe for the type "4" only related to higher power?

[solarplexus]

well I checked the board and it's a LM386N-3 so... not more than 12V

[TheBigMan]

Since the thread cuts into 386's:
With the same voltage, do different types (1-4) sound different?
Or is the awe for the type "4" only related to higher power?

I've never done a comparison but I'd guess at the same voltage they all sound the same.  I use a 4 in my Ruby, it'll happily run off 9V but I quite often run it at 18V.  Into a 2x12" loaded with G12H30s it sounds fantastic, almost like a Twin.

[Mark Hammer]

They will "sound different" to the extent that they elicit different behaviour from the speaker, but I would expect no more than that.

I can't pitch it as gospel truth, but I imagine the 1-4 designation is essentially "grading" of the same fundamental product, in much the same way that transistors are often graded as A/B/C by virtue of their hfe range.  I imagine some, but not all, suffix designations of many of the chips we all know and love come about the same way, and are not necessarily designated on the basis of some deliberate internal design difference.

I'll let the folks here with more industry knowledge confirm or refute that.

[petemoore]

  Usually it's the heat where they fried on me.
 I made quick heat sink.
 Tested temperature by touch, slowly increasing current [by moving volume control CW, with fuzz on etc.], allowing full heating/testing for each Volume^ step, the chip sounded good to me when it wasn't in 'super heat generation mode' anyway.
 the chip and heatsink heated up, but not too bad.
 

[Mark Hammer]

Setting aside the possibility of making one's own, what sorts of commercially available heatsinks are there for things like 8 and 14-pin dips?

I used an interesting heatsink for a 380-based miniamp I made, but this is the only heatsink of its type I've ever seen.

[mydementia]

Would there be any benefit to stacking LM386's?
I remember a thread about stacking dual op-amps - seemed to have interesting results.
I assume we'd get twice the output AND headroom - does this make sense?

I've been building Ruby amps with a SHO in front lately for friends - sounds great... but could use more headroom...

Just wondering.
Mike

[slacker]

well I checked the board and it's a LM386N-3 so... not more than 12V

I run my LM386-1 equipped Ruby at 14volts, and it works fine and will happily drive the 4ohm 2 x 12 speakers off a Fender Twin. I think so long as you stay below 15 volts you'll be OK, or perhaps Mark's theory is correct and I've got lucky and got an LM386-4 spec chip thats been "downgraded".

[Mark Hammer]

One wouldn't "stack" them per se.  Rather, they would be run in parallel, in much the same way that one has several power tubes or power transistors running in parallel.

Of course, since power-amp chips are more complex than a single power transistor or pentode, and since running the chips in parallel requires additional passive components that may be unmatched and produce differences in phase between multiple devices, one has to wonder if its such a wise idea.

The alternative, of course is the bridged-mode amplifier.  That can be done with simple mono power-amp chips, such as in the case of the Lil Gem Mk II.  But there are a whole host of dual-channel power-amp chips that are designed in anticipation of being run in bridged mode, and can deliver surprising power with very simple designs and a suitable supply.  For example, the KA2206 (a 14-pin dip) gives you 2.3W into 4 ohms with a 9v supply, but will give you 4.7W into an 8ohm load with 9v if run in bridge mode.  The KA2209 (an 8-pin DIP) is a fairly low-power headphone amp that will give 110mw per channel into 4ohms when you're running it from a pair of AAAs or AAs in an MP3 player or whatnot.  Run it in bridge mode with a 6v supply, and you can squeeze 1.35W from it.  The NJM2073 (also 8 pins), often found in cheap plastic desktop computer speakers (the kind that often get left curbside on garbage day ), will give you 650mw into each 4ohm speaker when run in stereo from the usual 6v AC adapter they come with.  Run that sucker in bridged mode, as per the datasheet, and you'll get a full 2W into a 16ohm load (e.g., a pair of 8" speakers wired in series) on a 9v supply.  I imagine if you heat-sink it, and run it at 12v into an 8ohm load you're looking at even more.  And that from a wimpy little headphone amp.

I run my LM386-1 equipped Ruby at 14volts, and it works fine and will happily drive the 4ohm 2 x 12 speakers off a Fender Twin. I think so long as you stay below 15 volts you'll be OK, or perhaps Mark's theory is correct and I've got lucky and got an LM386-4 spec chip thats been "downgraded".

The heatsink is the most efficient route for heat to travel, but not the only part of where the heat can go.  There are other routes for any heat buildup to travel, and conceivably yours is not reaching that critical point over a sustained enough period where it turns into magic blue smoke.  For instance, most datasheets for small power-amp chips will recommend having sizeable ground areas as a means of heatsinking the chip "from underneath". 

As well, keep in mind that wattage output is a function of what the input signal level tells it to do and how you play, not just the supply voltage and chip specs.  It IS possible you have one which is "under-graded", but it is similarly possible you have not stressed the chip as much as you think you have.

[slacker]

It IS possible you have one which is "under-graded", but it is similarly possible you have not stressed the chip as much as you think you have.

Too be honest until I saw this thread and reread the datasheet I didn't realise the operating voltage was only 12 volts, I'd just seen 15 volts on the datasheet and hadn't noticed that it was the absolute maximum voltage 
I guess the reason the blue smoke hasn't escaped (yet) is that I don't run it flat out because I like to keep it as clean as possible, which is why I increased the voltage in the first place.