I believe the 741 is just acting as a preamp for another unit.

 

I believe the 741 is just acting as a preamp for another unit.

It should but the TS was driving the speaker directly from the output of the 741 from his description in post #1

 

oh that's not gonna work... well... lol. it doesn't really have much of an output stage. It's not designed as a power amp. I was going by his diagram... sorry tldr

 

Zipped through the thread, still not clear:

Is the "speaker" a plain old 8 ohm voice coil speaker, or something else; something like an amplified speaker so commonly used with computers?

ak

 

@SamR try a class AB made up with some beefy transistor pairs like MJE3055/2955 or similar... you can do it with smaller transistors if you keep the gain low enough not to set them on fire.

 

An opamp has a maximum peak output of 20mA then its RMS current is 14.1mA. Then the power in an 8 ohm speaker is (14.1mA squared) x 8= 0.0016W which is almost nothing.
Adding an LM390 or LM386 will produce about the same 1/2W output power as the Alexa so why bother.
Why add a lowpass filter that cuts all intelligible sounds?

 

because he used 16k so at passband he's dropping everything on that instead of the speaker. I see what you're saying AG, the low pass is at just under 1K... lol. there's your real problem.

 

Guys you are missing the point it is just an exercise in learning according to the TS.

 

try a lower resistor value, but calculate the corner frequency... pick one in the audible range like 10k or 20kHz...

 

I put a 386 in and still having problems. Looked at the TL071 PDF and it is a pin for pin copy of the 741 and supply up to +/-18V. Thought it was a low supply voltage chip and it can be but doesn't have to. So I'll try one and it looks to be just the ticket to supplant the 741. Gonna take a break as I've been at it for a few hours. I take a fresh start back at it later with a TL071 and LM386 or should I use the LM380?. Thx for the input guys.

a plain old 8 ohm voice coil speaker

Yes, it is from an old early Technics surround sound BIG stereo system. It was the shelf sized rear speaker and has great audio out. Normally my Echo/Alexa goes into a pair of powered Bose desktop speakers that are independent from my computer which has some Altec Lansing speakers with a big subwoofer that my wife hates.

Why add a lowpass filter that cuts all intelligible sounds?

This is a lab exercise and it is part of the instructions for it to put the filter in and hear if it makes any difference. I can hear down to ~10hZ and this cuts off below ~1300hZ. My high end on my ears, primarily my left ear, quits around 8kHz and above. This thing was oscillating at one point and the wife came in to see what the racket was. I couldn't hear it but could see the oscillation on the scope...

 

your filter has cutoff of 1/ωRC or or about 995 Hz. Why not increase the cutoff to 5kHz where you will be able to hear and will roll off beyond that.

 

Need to see your complete schematic. The LM386 will oscillate without the 10 ohm and 100nf cap on the output.

 

The NE5532 is a reasonable low-cost op-amp for audio.

See https://www.analog.com/media/en/technical-documentation/data-sheets/LT1010.pdf#page=14

I've used the discrete circuit in figure 3 to boost the output of an OP-amp. The output and feedback are taken at two different places.

I've also used the "isolating a capacitive load" circuit.

Just never with audio,

 

You're trying too many things at the same time:
1) preamp
2) filter
3) power amp

Do try thing at a time.
Do you have a music source?
Do you have a computer powered amplifier?

Start by connecting the music source to the amplifier. Do you get decent sound output?

 

your filter has cutoff of 1/ωRC or or about 995 Hz

I actually measured the cutoff and with the components used it is ~1300hZ and that is a lo pass and I hear fine in that range. The problem I had was when I plugged the speaker into the circuit it flatlined, complete attenuation. Without the speaker, it looked fine on the scope and with the speaker plugged in before the filter it sounded fine. The resistor in the filter I guess was the "straw that broke the camel's back" in a sense.

The NE5532 is a reasonable low-cost op-amp for audio.

I have some and they are certainly a LOT quieter than the 741. One reason I used the "required" 741 the exercise called for was I was curious as to what the null deal was with the 741 so I was playing around with that also. I assume later model op amps don't need to be nulled?

 

You're trying too many things at the same time:
1) preamp
2) filter
3) power amp

Do try thing at a time.
Do you have a music source?
Do you have a computer powered amplifier?

Start by connecting the music source to the amplifier. Do you get decent sound output?

The previous lab exercises did just that. One exercise building using the preamp with a microphone. One exercise building active lo, hi, and bandpass filters and recording/plotting their outputs as the input frequency and voltage varied. The power amp is out of the blue and has not been covered at all so far in the book. I do have some experience using the LM386 though and have a few other model 2-5W amplifier chips. This was the first exercise connecting it to a music source. I assume the "power amplifier" was a piece of supplied lab equipment as it is only shown as a block diagram and there is no schematic for it other than wiring connections to it. Since I was able to drive a speaker without it previously I didn't need it before now.

 

Hello,

Did you have a look at this post?
https://forum.allaboutcircuits.com/threads/components-selection-guide.65137/#post-448192
In there you will find a various collection of amplifier chips.

Bertus

 

If you ever make a preamp with an audio low noise opamp driving a lowpass filter that is fed to a power amplifier driving a high quality speaker then a lowpass filter with a cutoff frequency of 1kHz will produce spoken vowels and boom boom sounds.

An AM radio has a highpass filter below about 250Hz from its cheap little speaker and its electronics has a lowpass filter above about 4kHz.
A good FM tuner in a hifi system produces sounds from 30Hz to 15kHz.
The best sounding hifi playing a CD produces all sounds we hear from 20Hz to 20kHz.
I do not know the bandpass of an Alexa. Maybe it does not produce sounds above a few khz.

Did you use your lowpass filter cutoff frequency at 1100hz to cut hiss from the noisy 741 opamp?

ALL single opamps have NULL pins to cancel the input offset voltage that every opamp has. The NE5532 is a dual so it does not have NULL pins.

Almost any audio circuit will oscillate if you build it on a breadboard with a tangle of fairly long wires all over the place.

 

No "normal" opamp, or even a high-performance one like the 5532, will drive an 8 ohm load directly to any listenable volume. As AG pointed out, an output current of 14 mA into 8 ohms is a delivered power of 1.6 milliwatts. The 5532 has a short circuit current of 38 mA, for an output power of less than 6 milliwatts.

ak

 

Did you use your lowpass filter cutoff frequency at 1100hz to cut hiss from the noisy 741 opamp?

No, it was just part of the exercise. I have used op amp voltage followers/buffers to remove EMI on low voltage scope inputs but then there is noise generated inherently in the 741 and other opamps.

I went back and looked at the scope shots and saw what I missed. The Vpp on the output is in the mV range. I really thought I had more Vout than that so that tells me a lot. I had a fair amount of gain but not that much input to amplify...