The LM386 is great if...

[mrspurious]

...you like listening to the Fox sports network. Anyone thinking about using this chip for a real project should take the comments on this forum seriously.

Now besides that, it does what it is supposed to - produces a high gain from a low voltage, and it is cheap.

I am having a lot of trouble with the TDA1521 based Velleman K4003 kit, and I'm asuming that it is entirely my fault for not using the correct type of power supply. I know a lot more now than I did when I started, and I may be able to salvage that project. To get some of the fundamentals down, I stopped by r-$hack and whith out any particular design, purchased the LM386, an 1000 uf electrolytic, 10K resistor, and small ceramic disk capacitor. The 9v batteries cost more than the components, I'm sure.

After an hour or so of preparation and looking at similar designs, I came up with this:



As you can see, the sad TDA1521 and board in background, missing one 4800 uf capacitor that I managed to boil -- I'm using this with the LM386 between ground and V+ . The other electrolytic is behaving as an output capacitor (it doesn't seem to make an audible difference). The 10K ohm resistor is used between 1 and 8 for the gain setting, with the small ceramic capacitor. Ground and non-inverting input from the chip go to ground.

The speaker is from a Beige variety Macintosh - from a previous PC mod intended to produce that authentic Mac sound on a PC. Sound quality of the amp really becomes less important when using this speaker.

I have some questions regarding Op-Amp circuits, being new to this, and I have looked on my own for some of these answers, I just need some confirmation.

1) It seems like it doesn't really matter if I use the inverting or the non-inverting input, as long as one of them recieves the signal, and one of them is tied to ground. I am guessing that by using the inverting input, the waveform of the output signal is opposite that of the input, and sounds essentially the same. Correct?

2) I cannot hear an audible difference between using the output capacitor or not, even though it is fairly large. Is it acting as some sort of filter, and I just can't hear it because of the small speaker, or does it do something else?

3) Can I sill use the 4800 uf capacitor even though it is probably leaking now and open to the air, or will it fail catistrophicaly, like a bad jar of peaches?

4) I havn't heard anything bad about the TDA1521, but few people on this board seem to be using it. Is it worth it to try and make the Velleman kit work, or should I move on?

[ofb]

3) Can I sill use the 4800 uf capacitor even though it is probably leaking now and open to the air, or will it fail catistrophicaly, like a bad jar of peaches?

DON'T use a bad cap, and i hope you're wearing safety glasses. even little ones can go with a pretty good bang and spread hot bits fast. be careful here. you can lose your eyes and scar your face.

[BlackDog]

Originally posted by mrspurious


2) I cannot hear an audible difference between using the output capacitor or not, even though it is fairly large. Is it acting as some sort of filter, and I just can't hear it because of the small speaker, or does it do something else?

The output cap. is there to remove the DC component of the output signal. -- As you are using a "single" supply (ie 9V battery), the output voltage swings between 0V and 9V, meaning that there is always some average DC component in the output signal, which can cause the damage to the voice coil in your speaker.

And throw away that bad cap. I agree that it could blow anytime.

Steve

[paulb]

Originally posted by mrspurious
...
1) It seems like it doesn't really matter if I use the inverting or the non-inverting input, as long as one of them recieves the signal, and one of them is tied to ground. I am guessing that by using the inverting input, the waveform of the output signal is opposite that of the input, and sounds essentially the same. Correct?
...
4) I havn't heard anything bad about the TDA1521, but few people on this board seem to be using it. Is it worth it to try and make the Velleman kit work, or should I move on?

The LM386 (and probably the TDA1521) are not quite the same as an op amp. But in general you are correct, inverting or non-inverting, the music is basically the same.

I would try tomake the Velleman kit work. Always handy to have an amp around, and you seem to be learning a lot from getting it going. Though you may have damaged the part, given the description of the cap.

[rudolph]

Yes, the Philips' TDA1521 and National Semiconductor's LM386 are basically power op-amps. Both are vintage, about 25 years old if memory serves.
Be wary of using the inverting input as your input. As this is the point at which voltage feedback is applied, it will be at a much lower impedance than the non-inverting input. (Think about it - it's negative feedback, which will reduce the voltage at this point by about 1+the voltage gain. So, voltage for a given input current is reduced, and so, from Ohm's law, must be the input resistance/impedance. A low input impedance will probably load your signal source. If it's a 50ohm signal generator, you won't notice, but an audio source will probably have a somewhat higher output impedance. Don't do it!
The TDA1521 was a very good chip in its day. I know, I worked for Philips at the time it was introduced, and knew the designer well. It was by far the best sounding chip amp then, and had novel RF suppression features in the output stage. Chip amps are notorious for generating RF noise. Try holding an AM radio near one! If you haven't destroyed it, do persevere.
You MUST use an output capacitor with the LM386 if you're using a single 9V supply! The DC voltage on its output pin will be at approx. half supply, or 4.5V - ish. If you DC couple, it will attempt to put nearly half an Amp through your 8ohm speaker! (of course the PP3 battery could not supply anything like this). Also, you'll only be able to reproduce, badly, the positive half-cycles of the signal, as the output voltage cannot go below zero.
It must sound awful! THD estimate IMHO > 40%
Hope this helps.
Mr Hucker.

[mrspurious]

Originally posted by ofb


DON'T use a bad cap, and i hope you're wearing safety glasses. even little ones can go with a pretty good bang and spread hot bits fast. be careful here. you can lose your eyes and scar your face.

I'm getting rid of it right now. I was lucky once, and I don't think I'll tempt fate.

Originally posted by BlackDog



The output cap. is there to remove the DC component of the output signal. -- As you are using a "single" supply (ie 9V battery), the output voltage swings between 0V and 9V, meaning that there is always some average DC component in the output signal, which can cause the damage to the voice coil in your speaker.

And throw away that bad cap. I agree that it could blow anytime.

Steve

Thanks for the advice. I feel like a nube asking a lot of questions on my first post, but the form here is very helpful. I'll be sure to post a pic when the TDA1521 is up and running.

MrSpurious

[rudolph]

Don't bother about that!
We all have to start somewhere. Never be afraid of asking questions. Even old duffers like me have a lot to learn still.
BTW, in your application, driving a PC speaker, 100uF for the output cap on the LM386 would be more than adequate. Always rate elcos at at least 50% more than the maximum voltage they are likely to see. 16V is OK in this case.

[BlackDog]

Originally posted by rudolph
Don't bother about that!
We all have to start somewhere. Never be afraid of asking questions. Even old duffers like me have a lot to learn still.
BTW, in your application, driving a PC speaker, 100uF for the output cap on the LM386 would be more than adequate. Always rate elcos at at least 50% more than the maximum voltage they are likely to see. 16V is OK in this case.

Rudolph speaketh the truth!