LM3875 troubleshooting

[Grandma´s_SUB]

So, whats the problem. Ive made gainclone with mentioned chip.Ive connected all as on scheme is, but that scheme was total minimised gainclone, which means no 100nF caps included at positive and negative rail.

+ and - voltage after greatz bridge and caps is 40.2V, which is 80.4V total.

So ive connected speaker on amp and suddenly it was burnt.Chip was very hot. The DC voltage was disaster and breaktrough in chip happened (resistance between -Vcc and GND is about 0.5 Ohm)

1.How have i burnt everything, if voltage was ok and connections were OK?

2.What *must* you do, to get DC voltage/chip breaktrough/big voltage at amp output (the most common mistakes)?

[moamps]

Are you sure you insulated the chip's tab from the heat sink mass?

Regards,
Milan

[Minion]

The Gainclone is allready Minimalized so why would you want to make it even more Minimal?? You are just asking for trouble because the Things you have Omited from the design are There for a reason Like to prevent Oscilation and for the Chip to run Stable and to prevent DC Offset....

You should go back and look at the Data sheet and and rebuild the Chipamp properly as per the Datasheet .....

I think what is happeneing is the Chip is Oscilateing which is caueseing it to quickly overheat and causing huge DC Offset, also you are running at close to the Max Voltages that the Chip can handle so you had better have a Big heatsink also the 100K Pot at the Input is too big ,It should be 10k to 50k max I think...

You should also add the 0.1uF caps to the supply pins close to the IC and add a 1k resistor in series with the 22uF Cap to ground off of Pin 8 and add a 220K resistor from the input to ground so there is a Path for DC to leak to ground and maybe sure you have enough Filtering in your Powersupply (at least 4700uF per rail per Chip)

You should be able to do these things without much trouble and it should solve yer problems....


Good Luck


PS: I hope Grandma doesn"t Punnish you to badly for screwing up her Chip amp...

[Grandma´s_SUB]

Originally posted by moamps
Are you sure you insulated the chip's tab from the heat sink mass?

Regards,
Milan

Sure.The Ohm-meter has shown infinitive resitance.

[moamps]

In that case, your amp may have run into oscillation, which is not entirely unusual for amps working in the inverted mode. Next time, you may wish to test the amp first by connecting it to a resistor. Once you're positive everything is fine, you can hook up your speakers.

Regards,
Milan

[MikeW]

Why are you biasing the output? after the feedback resistor you need to have a resistor to ground or you will have infinite gain. The output will go to the rail.

[MikeW]

http://www.national.com/images/pf/LM3875/01144901.pdf
Try this from the national website.

[jackinnj]

When you force an amplifier into Class-A all sorts of nasty things happen -- you might just as well have turned it into a space-heater.

[Grandma´s_SUB]

Originally posted by MikeW
Why are you biasing the output? after the feedback resistor you need to have a resistor to ground or you will have infinite gain. The output will go to the rail.

You mean this (take a look at attachment)?

Originally posted by jackinnj
When you force an amplifier into Class-A all sorts of nasty things happen -- you might just as well have turned it into a space-heater.

You are talking about that 1.5k/2W resistor?I didnt include it in my project.

And the most important; that scheme isnt my work, ive found it somewhere on the net.

[Grandma´s_SUB]

And why there is no connection between GND and feedback resistor in NUUK´s design?