Questions about the LM3886

Thread Starter

bloguetronica

Joined Apr 27, 2007
1,541
Hello guys,

I'm working on a new version of Taurus, that will have a few more components and will use the LM3886 instead of the LM3875. The datasheet suggests using a 0.7uH coil in parallel with a 10Ohm resistor, all in series with the load, so it will prevent HF oscillations due to the inherent capacitance of very long cables. I figured that a air core coil having 7 loops 8mm diameter of 23AWG wire would do. But what wattage the resistor should have? Is 1/8W ok? Or should be 1/4W or even 1/2W, on the safe side?

Also, the datasheet suggests a capacitor Cc of 220pF to eliminate HF induced radiation. Is it OK to use ceramic for this?

LM3886.gif

Also, I'm planning to add a crowbar to protect the speakers against DC current case the LM3886 fails, but I want it to be simple since I don't have much board space. Is it a good idea to use DIAC-TRIAC crowbars for this?
 

SgtWookie

Joined Jul 17, 2007
22,230
The datasheet suggests using a 0.7uH coil in parallel with a 10Ohm resistor, all in series with the load, so it will prevent HF oscillations due to the inherent capacitance of very long cables. I figured that a air core coil having 7 loops 8mm diameter of 23AWG wire would do.
I get about 450nH for your dimensions and number of turns.

But what wattage the resistor should have? Is 1/8W ok? Or should be 1/4W or even 1/2W, on the safe side?
Consider using a 2W resistor, as the body will be larger in diameter, and require fewer turns to make your 0.7uH/700nH inductor. You can use Wheeler's Formula to get "in the ballpark" for the number of turns vs coil ID vs coil length; it's within a few percent.

Also, the datasheet suggests a capacitor Cc of 220pF to eliminate HF induced radiation. Is it OK to use ceramic for this?
Sure.

Also, I'm planning to add a crowbar to protect the speakers against DC current case the LM3886 fails, but I want it to be simple since I don't have much board space. Is it a good idea to use DIAC-TRIAC crowbars for this?
Have you considered using non-polarized poly caps to block DC?
 

Thread Starter

bloguetronica

Joined Apr 27, 2007
1,541
Thanks SgtWookie!

Have you considered using non-polarized poly caps to block DC?
The problem is not about blocking DC from the input to the output. Yes, I could use a poly cap in series with the input, but NS suggested adding a cap (Ci) somewhere in the feedback network. That cap will reduce the gain at lower frequencies acting effectively as a DC blocking cap. And makes a better work since it is not in series with the audio line, as seen here:
LM3875.jpg

The problem happens when one of the output transistors inside the LM3886 gets damaged and shorts. How to prevent it from toasting the speaker?
 
Last edited by a moderator:

Audioguru

Joined Dec 20, 2007
11,248
Many amplifiers use a relay to disconnect the speaker while the amplifier is powering up and when there is a fault that produces DC into the speaker.
 

Thread Starter

bloguetronica

Joined Apr 27, 2007
1,541
Many amplifiers use a relay to disconnect the speaker while the amplifier is powering up and when there is a fault that produces DC into the speaker.
Yes, but the LM3886 doesn't produce a power on thump and I want to keep the project compact. I don't have space for big relays.

I was thinking more about this:
Crowbar.gif

But I don't know which diac and triac should I choose. Ideally, I want to fire the diac case there is a positive or negative offset exceeding 2V for more than one second. That prevents false triggering. But I don't know how to choose the values. Is there any literature? I want to design this myself, but I need some help.
 
Last edited by a moderator:

Audioguru

Joined Dec 20, 2007
11,248
What will you short-circuit with the triac? The amplifier's output to ground? Then you will blow up the amplifier if it doesn't have short circuit protection.
You cannot use the triac to connect the speaker to the amplifier (and to disconnect it) because it will cause severe distortion.

You need an amplifier to drive the gate of the triac, not a 30V diac. You show an AC light dimmer circuit.
 

Thread Starter

bloguetronica

Joined Apr 27, 2007
1,541
What will you short-circuit with the triac? The amplifier's output to ground? Then you will blow up the amplifier if it doesn't have short circuit protection.
You cannot use the triac to connect the speaker to the amplifier (and to disconnect it) because it will cause severe distortion.

You need an amplifier to drive the gate of the triac, not a 30V diac. You show an AC light dimmer circuit.
The LM3886 has short circuit protection, but has no DC protection.

The TRIAC will only act if everything fails, thus protecting the speakers.

There is no point protecting an already blown amplifier against short circuits to the ground, since one of the trannies is already shorted to one of the rails.

This is not a dimmer. This is a circuit that actually came from a book named "Audio Power Amplifiers Design Handbook", and it is actually used in many good amps.

Once again you didn't suggested a practical or concrete solution, nor presented an alternative, though the diac should be replaced since it will require 30V to conduct. At least that is a good start.
 
Last edited by a moderator:
Top