LM3886 amp build

[malcolm mellersh]

Having built the amp with all the recommended components I have a growling speaker output and a rd hot positive rail. If I unplug the speaker, the rail cools down. What's the problem?

 

[bertus]

Hello,

Can you post the used schematic?
That way we can see what is going on.
Is it the one in the attached PDF?

Bertus

 

[malcolm mellersh]

Hello,

Can you post the used schematic?
That way we can see what is going on.
Is it the one in the attached PDF?

Bertus

You can access the circuit from www.circuitbasics.com
A Complete Guide to Design and Build a Hifi LM3886 Amplifier

 

[Audioguru]

Your circuit has no power supply voltages.
Your schematic does not show the speaker connected to 0VDC.
The output of the amplifier should measure 0VDC.

 

[bertus]

Hello,

So, this is the page you are using:
https://www.circuitbasics.com/design-hi-fi-audio-amplifier-lm3886/

Bertus

 

[malcolm mellersh]

Your circuit has no power supply voltages.
Your schematic does not show the speaker connected to 0VDC.
The output of the amplifier should measure 0VDC.

Very true. The power supply is 29VDC negative and 29VDC positive. The input and output are both grounded

 

[malcolm mellersh]

Hello,

So, this is the page you are using:
https://www.circuitbasics.com/design-hi-fi-audio-amplifier-lm3886/

Bertus

Yes. That's it

 

[Audioguru]

Maybe you bought a fake LM3886 from ebay?

 

[bertus]

Hello,

Here is the schematic from that page:

Did you connect ALL the ground connections together?
Did you also install Rsn and Csn?

Bertus

 

[Audioguru]

The schematic in the video has a signal ground for the input but also has a power ground. It wrongly does not show them connected together.

 

[malcolm mellersh]

Hello,

Here is the schematic from that page:
View attachment 53153
Did you connect ALL the ground connections together?
Did you also install Rsn and Csn?

Bertus

All grounds together. I have checked and signal ground and power ground are radially led to a single point.
I have installed Rsn and Csn
Could I have a faulty Lm3886. If so, how do I check it out?

 

[bertus]

Hello,

Wich version of the LM3886 are you using?
There are the LM3886T and the LM3886TF.

Bertus

 

[malcolm mellersh]

Hello,

Wich version of the LM3886 are you using?
There are the LM3886T and the LM3886TF.

Bertus

It's the TF

 

[bertus]

Hello,

How large is the heatsink?
The TF version will likely need a larger heatsink, as the case has a plastic cover.

Bertus

 

[malcolm mellersh]

Hello,

How large is the heatsink?
The TF version will likely need a larger heatsink, as the case has a plastic cover.

Bertus

I'm aware of the heat sink. Why is my speaker growling etc.
It's not the circuit and its probably not the components. Its me.
I've made an error somewhere and I can't find it. Can you help?

 

[WHONOES]

Referring to the schematic as posted by Bertus, if that is what you have built then perfom the following, if not, can you sketch out what you have actually done.

Firstly, disconnect the input ground connection from the ground connection used for the rest of the amplifer and take it direct to the power supply ground.
Secondly, discard D1, Rm and Cm and then connect pin 8 to the amplifier ground.
Thirdly, capacitors Cs1 and Cs2 must be connected as close as is physically possible to the power supply pins of the amplifier.
Lastly, The ground connection for the amplifier needs to be as robust as you can make it.

Try these one at a time so that you will know which one (if any) eradicates your problem

 

[malcolm mellersh]

Thankyou for that.

The signal ground is already directly connected to the power ground. Input socket to input socket.
I have isolated D1, Rm and Cm and connected pin 8 to ground. No difference there, so something else is happening.
I will bestripping back and rebuilding the basic circuit (without frills)
Any further advise would be welcomed.

 

[WHONOES]

The signal ground needs to be disconnected from the ground on your board and connected directly to the power supply ground so that it is completely independant from the grounds on the board.

 

[malcolm mellersh]

It is. The signal ground is taken from the signal input socket in the amplifier casing and passed directly to the power input socket in the amplifier casing. It bypasses the board completely.
Power is provided from a separate power supply unit.

 

[WHONOES]

Can you accurately sketch out what you have actually built and post it on here. Doing so will help us and may even show you where you have gone wrong (if at all).