Which toroidal is ok for the LM3886 and speakers 4 ohm?

[personal]

Hello to all, I have a one toroidal with 2X25V 6A and 4 MUR860. How I can use them for 2 LM3886 and speakers 4 ohm?
Thanks of your help


Someone of you has information of the Be' Yamamura?:confused

[Dxvideo]

I also work on a LM3886 x 2 @ 4ohm project. As i calculated on Overture Design Guide for 4ohm;

@ LM3886TF Pack Maximum Vcc/Vee = 26Vdc and it gives 63W RMS per channel with 1,52CW heatsink for each IC

@ LM3886TA Pack Maximum Vcc/Vee = 30Vdc and it gives 84W RMS per channel with 1,65CW heatsink for each IC

They also can work over that voltages but with really huge heatsinks (lower than 1,5CW) Under that situation,
For TF package 26/1,41 = 2x18v AC
For TA package 30/1,41 = 2x21v AC maximum trafo outputs.
* I dont add the diode losses to the calculations..
Nice days.

[personal]

Thank DXvideo,
but you have the scheme for my problem?
I am a novice
Very very thanks

[Dxvideo]

25 x 1,41 = ~35v DC without losses. May be you can get 32-33v for each rail. Theorically; i guess it works with 4ohm speakers but i recommend that do not turn volume pot to the end!
- Or you can reduce the voltage gain for; 18K-1K = 19AV
- May be you dont use a preamp, just connect them to the source.
- And a third solution; you regulate DC voltage with two LM338 then you can adjust the output voltage down to 28v.
But for the first two way you must cool the IC very strongly..
Have a nice day.

PS: I am not very experienced member of that forum. There are many "master" members (for ex:Pink Mouse, AndrewT, NicoRas). I guess you must ask them before apply my solutions.

[Nico Ras]

Originally posted by Dxvideo
25 x 1,41 = ~35v DC without losses. May be you can get 32-33v for each rail. Theorically; i guess it works with 4ohm speakers but i recommend that do not turn volume pot to the end!
- Or you can reduce the voltage gain for; 18K-1K = 19AV
- May be you dont use a preamp, just connect them to the source.
- And a third solution; you regulate DC voltage with two LM338 then you can adjust the output voltage down to 28v.
But for the first two way you must cool the IC very strongly..
Have a nice day.

PS: I am not very experienced member of that forum. There are many "master" members (for ex:Pink Mouse, AndrewT, NicoRas). I guess you must ask them before apply my solutions.

Dxvideo, thank you for your compliment. No masters just enthusiasts like yourself.

Dear Personal,

A simple solution is to use a capcitance multiplier as shown. This perform two functions, it will drop the voltage by about 3V and multiply the base capacitance with the combined gain of the transistors.

Because there is not a high volt drop across the series transistor it will not get very hot and moderate heatsink is sufficient.

[CarlosT]

What happens if you drop a relatively large (1 uF to 6.8 uF) non-polar film cap across V+ and V-? I see no harm since it is non-polar but what is gained?

Look at cap C7 in the attached schematic (from TechDIY's LM3875 board).

[CarlosT]

NicoRas is your schematic what is called a "snubberized" setup?

[Nico Ras]

No it is really a capacitance mutliplier. Or to put it simply it is a series regulator without regulation. But the capacitance multiplier smooths the output ripple to almost nothing.

[AndrewT]

Hi Per,
I just looked up the 3886 datasheet for you.
It says +-28Vdc for 4ohms.
That needs a 20-0, 20-0Vac or 20-0-20Vac transformer. NOT 25Vac.

4ohm and chipamps are not a good mix, if you want cheapness and simplicity.

I use the rule that a good amp will drive half the normal impedance without failure.

So, if you want to drive a reactive 8ohm speaker then the amp should drive 4r0 resistive load.

If 4ohm reactive then the amp should safely and reliably drive 2r0 resistive.

That ability will require an even lower voltage transformer.

I repeat 4ohm and chipamp are not a good mix.

[Nico Ras]

Hi Andrew,

I also had a look see the data sheet and I am in full agreemnt with Andrew on the topic of 4 Ohms.

Since he has the tranformer, what about changing the simple cap multiplier into a regulator replacing the 12 K resistors with 24 V zeners and the 470 Ohm to something like 2.2K.

Although heatsinking would be necessary the tranformer would drop under load when psuhing it and the regulator would not have to work that hard keeping more or less stable temperature rise.

Then again this may sound silly but what about adding two 4 Ohm 10W resistors in series with his speakers, would cost him a lot less and offer more protection. Besides, bass often sounds "nicer" with series resistance.