Chinese LM3886 Board having problems

[mattthegamer463]

I bought this amplifier board, a 68W per channel LM3886 based amplifier with integrated power supply. With the board I am using a 420VA 25-0-25 toroidial transformer.

2 x 68W Watt LM3886 + NE5532 Audio Amplifier Board on eBay.ca (item 330375561759 end time 11-Nov-09 22:00:52 EST)

I have spent the past few months working on trying to get it to work correctly. So far I have blown up two boards trying to get them to work, and am on my third one now. The only reason I bought any more boards is because the chassis is built around the board's design and I don't want to re-do the chassis. Hindsight is 20-20, so please dont ridicule me about buying chinese junk off ebay. I'm on my last thread of sanity and I just want to get this thing going.

The problem with the first two boards was that I hadn't isolated the LM3886 chip. There were no instructions or any documentation included with the amplifier board, so I didn't think of isolating the chip. Now that that issue is fixed, there isn't any more sparks, but I have managed to liquify two 3A fuses at the 120V mains input. The amplifier will work for a while with a small 10W 8 ohm test speaker, but when i connected a 40W 8 ohm bookshelf speaker the fuse blows like its a dead short. The speakers are fused with 1A fuses, which did not blow. I am not sure if its the speakers that are causing this or if its something else. One of the 40W speakers is now damaged from this amplifier.

If anyone has any idea what I'm talking about or what is going on with my unfortunate amplifier, please post. Thank you.

[netbug]

Thats a pretty large tranny (va not volt). I could see it blowing a 3amp non-slowblow fuse. Try a 3.5 or 4.

[mattthegamer463]

Originally Posted by netbug

Thats a pretty large tranny (va not volt). I could see it blowing a 3amp non-slowblow fuse. Try a 3.5 or 4.

You're right, it has a lot of headway, but the circuit should not be drawing 3A @ 120V, thats 360 watts. The board shouldn't require that much power. Also, the way that the fuse blew and the conductor vaporized into little copper balls indicates that much more than 3A passed through that fuse very quickly.

[netbug]

On power up the tranny will supply as much as its capable of until the power capacitors are charged. That 20,000uF acts like a short until charged. Also looking at the pictures from the auction, I only see 2 diodes so I'm really confused on how they get the +V and -V. half wave rectification on each half the center tap??

Not sure why it would act differently with differnet speakers attached.... I'm probably barking up the wrong tree.

[madisonears]

Do a search for light bulb tester and use that circuit in series with the AC line in. It might not help you fix your amp, but it will probably give you a better chance to get it powered up without blowing fuses.

Peace,
Tom E

[mattthegamer463]

Originally Posted by netbug

On power up the tranny will supply as much as its capable of until the power capacitors are charged. That 20,000uF acts like a short until charged. Also looking at the pictures from the auction, I only see 2 diodes so I'm really confused on how they get the +V and -V. half wave rectification on each half the center tap??

Not sure why it would act differently with differnet speakers attached.... I'm probably barking up the wrong tree.

Perhaps a slow-blow fuse is the answer?

[Andrew Eckhardt]

Originally Posted by mattthegamer463

Perhaps a slow-blow fuse is the answer?

It's possible to stress or even open a fast acting fuse with large filter caps and no in-rush limiting, but I'm concerned that you were actually able to bake a speaker with a 1A inline fuse that didn't blow. You got something going on there. Once that amp is powered up you should be able to change speaker loads with no signal and nothing real fancy should happen. I'd power it up with no load and check for offset at least. If it's the caps inrush current blowing your rail fuses and you don't have a short or oscillation problem, it's going to happen no matter what is connected to the output of the amp.

[poynton]

Originally Posted by netbug

On power up the tranny will supply as much as its capable of until the power capacitors are charged. That 20,000uF acts like a short until charged. Also looking at the pictures from the auction, I only see 2 diodes so I'm really confused on how they get the +V and -V. half wave rectification on each half the center tap??

Not sure why it would act differently with differnet speakers attached.... I'm probably barking up the wrong tree.

Full-wave bridge!

The fuse will blow because of inrush current.

The speaker damage could be due to oscillations.


Andy

[mattthegamer463]

Originally Posted by Andrew Eckhardt

It's possible to stress or even open a fast acting fuse with large filter caps and no in-rush limiting, but I'm concerned that you were actually able to bake a speaker with a 1A inline fuse that didn't blow. You got something going on there. Once that amp is powered up you should be able to change speaker loads with no signal and nothing real fancy should happen. I'd power it up with no load and check for offset at least. If it's the caps inrush current blowing your rail fuses and you don't have a short or oscillation problem, it's going to happen no matter what is connected to the output of the amp.

Whoops, I neglected to mention that the speaker died before the 1A fuse was put in place. Sorry.

As far as I can hear, there is no oscillation problem. When the unit is functioning correctly, there is no audible hum out of the speaker. I haven't put it on a scope yet though.

Checking for DC offset is still a good idea though, I think. I will check it out tomorrow.

[Andrew Eckhardt]

Originally Posted by mattthegamer463

Whoops, I neglected to mention that the speaker died before the 1A fuse was put in place. Sorry.

As far as I can hear, there is no oscillation problem. When the unit is functioning correctly, there is no audible hum out of the speaker. I haven't put it on a scope yet though.

Checking for DC offset is still a good idea though, I think. I will check it out tomorrow.

The trouble with rail fuses is they are "designed" to protect the amplifier, not the load. Depending on the behavior of the IC one rail might make it to the load if that's all it's getting. So without knowing anything about the IC ('specially if it's shorted), if you use rail fuses, use an inline fuse too, or watch out.