Dennis from this Forum has designed a circuit using the lm 3886 as driver.
He write this to me with a circuit attached (can't attach it with this e-mail im not home)
My solution is
> Build a module on a plate with 2X MJ15015 and 2X
> MJ15016 and 1X LM3886, with the plate isolated from
> the chassis.The LM3886 is to be the isolated package
> version.All four outputs are to be mounted to the
> plate with no insulators.This eliminates the thermal
> resistance of the mounting washers.The expense of
> these washers can be more than the transistor they
> mount!The LM3886 will then track the temperature of
> the added outputs and actually protect them through
> the internal SPIKE circuitry!The MJ outputs are
> by 0R47 resistors in the +/- rails of the LM3886
> each MJ having a 0R22 emitter resistor.At 400W/4R
> will cause 4A (peak) to flow into the LM3886 and 5A
> (peak) in each of the MJ outputs.Because the current
> gain of the MJ outputs begins to go down at higher
> currents the circuit will have inherent short
> protection from the LM3886.This limit may be
> if desired, by using a 0R68 instead of the 0R47, and
> paralleling with values in the range of 3R9, 4R7,
> etc, until the exact current limit you desire is
> reached.Very high power at 2R is possible by adding
> more MJ output pairs.
> >that this idea could be adapted to using the stereo
> >LM1876 with two MJ15003/15004 pair for a 300W amp,
> >possibly the plastic MJL21193/21194.
I haven't build the circuits yet, But Dennis write that it works well, not high end ,but a good easy low cost sulution
The TDA 2030 circuit was published in one german as well as in Elektor. But it will be nice to use the same idea for LM 3886. If Mr. Dennis can publish circuit and pcb layout along with his experience then many like me will be interested.
Since dennis kleitsch sent the circuit to me by normal mail (his scanner didnt work)
I try here to sent the circuit
Dennis write this to me :
The maximum voltage on the smaller LM1875 chips
is only 60V as opposed to 84V on the LM3886.The
circuit would be different with the stereo chip as it
does not have stereo power supply pins to drive the
output transistors.It would be much simpler to use two
chips.Yes, you could use the MJL21193/21194 plastic
outputs, but I wouldn't.The current gain is so high
that the protection in the LM3886 wouldn't kick in,
and they cost about twice as much as the
MJ15015/15016.Even though the supply is marked +/- 40V
it is actually a single, floating 80V, and only a
single filter capacitor.The (ground) reference point
is obtained by a resistive voltage divider and only
has to supply the input bias for the LM3886, about
I realy hope that it okay that I have put your(dennis) circuit on the page
I still havent build the amp my self but om going to try.
To me it seems to be more easy than all the matching with 0.1 resistors and servo used in the parralel/bridge application for lm3886.
Does any one know of any low cost transistors that can replace the mj15015/15016 in plastic housing, so they dont need to be isolated.
Have you considered a driver IC like the <a href="http://home.eunet.cz/rysanek/pdf/tda7250.pdf">TDA7250</a>
Theirs a 100W per channel version <a href="http://www.wenfu.net/audio/tda7250.html">here</a>
<hr width="95%" align=center>
<i><font color="#804000">Have trouble with the sice of the attachment because the computer saved it as bmp, I try to sent it soon as posible again
Have you tried <a href="http://download.cnet.com/downloads/0-10077-108-38759.html?tag=upd">irfanview</a> converts to/from a very broard range of Images
file types, and the resize routine is very very good, best of all its free.