==>> Parallel 4x LM3875 for 2 ohm speaker - pls advise..

[metha]

Hi,

I have 10 of spare LM3875TF on hand, so planning to build a new amp that can easily drive 2-4 ohm speakers. Here is the questions:

1) Can it paralell 3(even 4?) of LM3875TF by using the circuit on the NS's AN-1192 app note?

2). How can I make it as an Integrated amp? (the only source is CD Player - NAD521BEE).

3) How many watts will it be? ( tranf 25-0-25 Vac )

4). Any other ideas to play around with these spare LM3875TF?

Please advise. Thank you in advance.

[tlmadsen]

Hi

See my response in this
thread:Can you tell me how to beef up PA100?

Have fun

THomas

[carlmart]

Originally posted by tlmadsen
See my response in this
thread:
Can you tell me how to beef up PA100?

On your comments there, making the output resistor higher to lower DC offset, maybe that's why Rowland used 0.2 Ohm 1% resistors (instead of 0.1 ohm) on his first LM3386 version and then went for 0.5 Ohm.

In the first case, the Concentra, as Thorsten Loesch quite rightly pointed out by the parts count, it's quite likely Rowland implemented an IGC. His latest amp using 3886 seem to be non-inverting, but they carry a multi-turn trimpot on each chip, probably to lower DC offset.

In neither case does he use a servo, as National recommended on his parallel/bridge application note.


Carlos

[Mattyo5]

When BriantGT gets a parallel circuit similar to the Concentra...he'll sell 1000 boards easy. Can't wait

-Matthew K. Olson

[cdoggy81]

"When BriantGT gets a parallel circuit similar to the Concentra...he'll sell 1000 boards easy. Can't wait "
Amen to that as that's EXACTLY what I'm holding out for!!!

[tecnofossil]

WHy don't you guys simply buy 2 of Brians kits?


-T

[arthur]

am i commit ip or copyright problem if i
disclose all circuit of model 8 and model 12 of rowland

[tecnofossil]

Originally posted by carlmart


On your comments there, making the output resistor higher to lower DC offset, maybe that's why Rowland used 0.2 Ohm 1% resistors (instead of 0.1 ohm) on his first LM3386 version and then went for 0.5 Ohm.

In the first case, the Concentra, as Thorsten Loesch quite rightly pointed out by the parts count, it's quite likely Rowland implemented an IGC. His latest amp using 3886 seem to be non-inverting, but they carry a multi-turn trimpot on each chip, probably to lower DC offset.

In neither case does he use a servo, as National recommended on his parallel/bridge application note.


Carlos

Hey Carlos, where did you dig up the info on the later version of the Rowland amp using .5 ohms rather than .2 of the earlier unit? Just curious. One of my "multi chip" amp circuits uses the 6 (LM3886) devices per channel like Rowland although I am using lead type components rather than SMDs. If I have a moment I will try and take a pic and post it. I decided to go this route because of the unpleasant load my ES panels present to the driving amp. (looks to the amp like a capacitor)

-Tom

[carlmart]

Originally posted by tecnofossil

where did you dig up the info on the later version of the Rowland amp using .5 ohms rather than .2 of the earlier unit? Just curious.

I don't seem to find the threads here, but they should be. Look at the photo.

Originally posted by tecnofossil

One of my "multi chip" amp circuits uses the 6 (LM3886) devices per channel like Rowland although I am using lead type components rather than SMDs. If I have a moment I will try and take a pic and post it. I decided to go this route because of the unpleasant load my ES panels present to the driving amp. (looks to the amp like a capacitor)

Well, most ES panels behave like that and are a heavy load to most amps. Paralleling the devices should do it.

Which circuit did you use to implement that multi-chip amp?


Carlos

[carlmart]

Originally posted by tecnofossil

where did you dig up the info on the later version of the Rowland amp using .5 ohms rather than .2 of the earlier unit? Just curious.

I don't seem to find the threads here, but they should be. Look at the photo.

Originally posted by tecnofossil

One of my "multi chip" amp circuits uses the 6 (LM3886) devices per channel like Rowland although I am using lead type components rather than SMDs. If I have a moment I will try and take a pic and post it. I decided to go this route because of the unpleasant load my ES panels present to the driving amp. (looks to the amp like a capacitor)

Well, most ES panels behave like that and are a heavy load to most amps. Paralleling the devices should do it.

Which circuit did you use to implement that multi-chip amp?


Carlos