Yes, you need the push-pull transistor circuit at the output (without the 100k resistors) to provide additional current to drive the speaker.
With a +/- 15 volt supply The 741 itself can only supply around 5 or 6 ma into a 2k ohm or greater load which equates to 10 to 12 volt swing. But as the load resistance drops so does the voltage swing.

What would the voltage be across an 8 ohm load with 5ma's of current?
If you added transistors with a gain of 20, what would the current and output voltage be for an 8 ohm load?

EDIT: Also, as frequency increases the output of the op-amp may decrease. (Look up gain-bandwidth product).
The 500 ohm load was high enough to allow a reduced but useful output signal.
BTW. Have you tried the circuit without the 100k resistors ?

 

It would take .35 mA to produce 2.8 voltage across the 8 ohm load. The voltage drop would have to be 35 kV. Okay that makes sense lol.

If I am hearing sound through the weighted summers, do i even need this circuit configuration (my teacher said i would need it for a speaker)?
...

Check your units

 

Yes we did. We changed them to 90 Ohms because if we didn't add the resistors in, the transistors heats up very quickly but we didn't try to go lower than that because we ran out of time. We hear good sound through the speaker as we were hearing at the weighted summers but saturated.

voltage across the 8 Ohm load would be 40 mV with current at output at 5 mA

I'm guessing you mean the gain of the current by 20 (Not sure). This would raise the current to 100mA and would make the voltage 800mV.

"The 500 ohm load was high enough to allow a reduced but useful output signal."
Do you mean that it lowered the output current, so it was showing us a good voltage but if the 500 ohm was the speaker, we wouldn't hear any sound because of a very low current.

 

Check your units

Lol. Ooops

 

Limited time right now.
Yes, those transistors will get very hot trying to drive an 8 ohm load much above a volt or so RMS.
Calculate the power dissipated for 2.8 volts P-P into 8 ohms versus lets say, 5 volts or 10 volts P-P.
Next, look up the power rating of those transistors.

EDIT: Is the lower pnp transistor correctly wired as mentioned earlier?

 

So, the power dissipated for 2.8 V P-P is .35 A and it would be .625-1.25 [A]. The schematic is updated.

http://www.fairchildsemi.com/ds/2N/2N3906.pdf

http://www.fairchildsemi.com/ds/2N/2N3904.pdf

I think the max amount of current i can send through these is 100 [mA] based on description of the circuit.

So i would probably need better transistors?

 

The transistors you need have to not only handle more current, but also more power. Probably in TO-220 case with a radiator. Now power transistor usually have small beta. So darlington transistors would be good. Say TIP120 (NPN) and TIP125 (PNP). But this makes it even more important to forward bias them to avoid crossover distortion.

 

Our lab professor didn't really answer our question, so we went to another professor and look at our project for 10 seconds and said you need better transistors.
So we are going to try that out.

Thanks for all the help.

I will keep it updated if it works after that.

 

So, the power dissipated for 2.8 V P-P is .35 A and it would be .625-1.25 [A]. The schematic is updated.

http://www.fairchildsemi.com/ds/2N/2N3906.pdf

http://www.fairchildsemi.com/ds/2N/2N3904.pdf

I think the max amount of current i can send through these is 100 [mA] based on description of the circuit.

So i would probably need better transistors?

Power(watts) = volts x amps, not just amps.

Yes you need higher power darlington transistors, as alexfreed mentioned.

Do you understand what the term crossover distortion refers to?

 

Right power formula i know, i was on my phone when i sent it.

Crossover distortion is the change from one transistor to another. In a very basic case, from -.7 to .7 [V] there is the flat spot on the a Vout/Vin graph because of the voltage drop of a "diode" (model in a transistor). Then you add in the feedback in order to get rid of this. So I assume this well be larger with the darlington transistor (Guess).

 

Good guess. 3 standard silicon diodes in series between the bases of the output transistors and a couple of resistors to the + and - power rails will help a lot. And as previously mentioned using a faster op-amp, like a cheap NE5532 or TL072 will help at higher frequency. If the op-amp doesn't support a high enough slew rate, it won't be able to quickly swing between the -1.4 and +1.4 volts. With 3 diodes the gap of about 0.7 V will remain.

 

So we cant get our transistors to work correctly. When we plug of the transistors like seen in the schematic, even with diodes in between the bases (looked at schematic online), the transistors are heating up to where they will smoke if you keep it on. I am not sure if the 90 Ohms resistors are not large enough, and we are running into crunch time on our project. If we change the 90 Ohm resistors to a large value resistor will that change our output current to the speaker?

We don't have much experience with transistors as this is a very new concept for us. From U7 to the speaker is not working correctly and our transistors have very high ratings. NTE184 and NTE185. Our group is very lost

Thanks for the help

 

First of all, the transistors used are not darlinton ones, so 3 diodes will be too much, but 2 will be OK. If they get hot WITHOUT diodes between the bases, you must either have mis-wired them or there is parasitic oscillation. Otherwise there should be zero current at idle (as on schematics shown). There should be no 90 ohm in collectors but small resistors in the emitter will help to stabilize the idle current. If it is too high, the voltage drop on the resistor will reduce the base-emitter voltage and hence the current. The Hfe for these transistors is about 50, so to get 1 amp on the speaker you'll need 20 mA from the op-amp. The 741 can drive 25 mA, so you may get away with it, at least at low frequency. And what will be the voltage drop at your 90 ohm resistor? Exactly, 90 volts you don't have. So this can't work with collector resistors.

 

At idol they work fine. Not heating up, but as soon as we start to put an input voltage, they start to heat up and if I put a resistor at the emitters, that makes a voltage divider with my speaker and then I don't get enough voltage to hear my signal.

 

Do you have the transistors on isolated heat sinks ?

EDIT: Please post the schematic as you wired it with the bias diodes.

 

No
Would I have to? If so is there anyway to power the speaker without this configuration. If not, we will probably just use a speaker that is power from inside of itself. Because a heat sink seems a little much for this project and don't have enough time to order one.

 

I used this schematic. Not by computer so cant load my schematic and update it.

 

No
Would I have to? If so is there anyway to power the speaker without this configuration. If not, we will probably just use a speaker that is power from inside of itself. Because a heat sink seems a little much for this project and don't have enough time to order one.

Yes you need a heatsink for each transistor.
The heatsinks should not touch each other because those transistors likely have the collector electrically connected to the heatsink tab/hole. (These I think, use a TO-126 package which has this small metal area on the back around the mounting hole).

EDIT: A heatsink can be a simple small piece of aluminum or even a fender washer (has a small hole compared to the diameter of the washer). You can mount them together on the same metal heatsink only if you use special insulators and hardware designed for this.

 

Used a different schematic for powering speaker. Will show last schematic when I am by a computer. Thank you guys for all your help on my project. It was greatly appreciated.

 

Final Schematic. Has a tiny amount of crossover distortion but very very little to even notice.