Why is the LM3875 more poular than the LM3886?

[head_spaz]

Why is the LM3875 more poular than the LM3886?

[Peter Daniel]

Because it sounds better.

[Damon Hill]

I dunno, I think some people claims the LM3875 sounds better
and it's a bit simpler having no mute circuit. If you've already
got the LM3886s, I say use them.

I used the LM3886 for my initial amplifier, and it sounds quite good even into a difficult load (Acoustic Research AR-11 (four ohms) and Kimber 4TC cable). I chose it because I thought
I'd need the higher power for my speakers. It's wired
exactly according to the data sheets, including RL load
protection on the output.

Because this chassis is modular with the chips wired P2P
and connected via terminal strips all mounted on a heat
sink, I can use my extra heat sinks to build more versions
and I might try it with the LM3875 built to a true GainClone
configuration.

Waiting to see how the LM4780 PCBs turn out, and for
user comments on the sound quality (so far, positive).
I think I'll go that route for my 'finished' amplifier, with
three sets of inputs as I already have a suitable selector
switch in my parts collection.

I've also got something a great many DIYers don't: a
serious audio-oriented test bench and equipment.
No guessing if something's wrong or not.

--Damon

[jackinnj]

national did a nice piece on bridging the lm3886 . i think that the 3875's popularity is due to Thorsten's inverted design.

it is time for a bakeoff, however.

[BrianGT]

My theory is that the LM3875 is a much simplier chip to implement in a diy design, due to the reduced pin count over the LM3886.

The LM3875 only has 5 pins used: +V, -V, +IN, -IN and OUT.

The LM3886 has 8 pins used: +V, +V, -V, +IN, -IN, OUT, MUTE and GND. (2 +V pins, not a typo)

With the reduced pins, and slightly simplier circuitry (no mute resistor needed), the LM3875 is a simplier implementation, and thus more appealing for a bare minimum components.

As for how the additional mute circuitry effects the sound of the LM3886, I have no idea. I have a pcb design for the LM3886 very similar to the LM3875 design, but I haven't tried it yet.

--
Brian

[Alcaid]

Have anyone tried the LM3886 with a bare minimum of components, as in the LM3875 minimalist gainclone? (adding one resistor for the mute circuit). Was it stable?

[Peter Daniel]

I tried LM4780 in that minimalistic configuration and it was very stable.

[pinkmouse]

I was never happy with the 3886 in minimal configuration, it works much better with big caps and more in the way of support components, IMHO

[lgreen]

I have tried the LM3886 (schematic here) in a minimal design-- inverting config, and it was stable into speakers and 5 ohm power resistors. That said, I did add to this configuration, more caps (3300 uF per rail total for each of 4 rails) and a differential receiver with its own power supply, though these changes probably don't affect stability. The final result is shown here but put your sunglasses on before clicking the link.

[carlmart]

Originally posted by Alcaid
Have anyone tried the LM3886 with a bare minimum of components, as in the LM3875 minimalist gainclone? (adding one resistor for the mute circuit). Was it stable?

Yes, I did try it and it was absolutely stable. According to Panasonic's table, you can get more power from the 3886 for the same supply.

It sounds great in a minimalistic gainclone. Not surprisingly it was Rowland's choice for his commercial amps.

It's the definitive choice for powering subs, single or in parallel.


Carlos