Crap, so I turned the pot all the way down and I felt the to-92s, I guess both overheat and this is with pot all the way down and ALSO, the two Opamps overheat...the heat is ALOT. I likely fried the opamps and to-92s. the to-220s are fine. I am not worried about frying them because I have a dozen opamps of the same variety to test, and alot of the to-92s...So what the heck is up with the 6v cells?

 

Sounds like you wired the 6V batteries backwards...

 

dang, I had the leads from the 6v batts to the opposite terminals...the batts were wire correctly but I sen the pos to the neg neg to pos on the circuit board...I am going to re experiment..

 

Okay, so no overheating it was a simple fix but I still get the led on reguardless of switch position....When I measure each power supply terminal on the power supply input terminals(connected to batts) from each to ground I get the typical 6v measuring from one side but not the other while the switch is open but it shows while closed to this goes to show that I wired it wrong for the switch because the GND and one side is powering the led and neg supply side...and it's 6 volts so it isnt like the GND is eliminated with the switch...It has effectively worked before where the led was off while the switch was open and so it is strange that is suddenly stoped working...I guess i should check the wiring again.

 

Okay So I finally resolved it not working at all and the LED being on reguarless of the switch...I moved some wires around and I believe there was likely some wires touching or something loose...

Okay something that bothers me is that the voltages measured on the power output transistor collectors, and the batteries themselves while playing music, is UNBALANCED. I don't know why...this is very likely the same reason the speaker still idles it's cone position away from center(offset)

I am using cheap 741 opamps, should I play with the offset Pins or something could this be it?

 

Okay So I finally resolved it not working at all and the LED being on reguarless of the switch...I moved some wires around and I believe there was likely some wires touching or something loose...

Okay something that bothers me is that the voltages measured on the power output transistor collectors, and the batteries themselves while playing music, is UNBALANCED. I don't know why...this is very likely the same reason the speaker still idles it's cone position away from center(offset)

I am using cheap 741 opamps, should I play with the offset Pins or something could this be it?

If your circuit is wired correctly, and all parts are good, your speaker should not be biased (unless you have your pot wiper all the way to the left on your schematic). The op amp offset adjustment is not needed.
I would change your circuit to a single supply, as in the attachment. This will avoid passing large DC current through the speaker in the event of a fault (which is apparently what you have). High current through the speaker in your DC-coupled circuit means high current through one of your power transistors.
Don't use a 741. The frequency response and crossover distortion will be horrible. TL072 is not ideal, but it will WAY better than a 741. If you need to use two separate op amps, use TL071.

 

I was always under the impression the tl7/8x were decent opamps. What would you say is good? Alos, is there any way to adapt that rail to rail circuit posted earlier to this circuit?

Oh and could I just use a capacitor on the output in series to remove a dc offset?

And If I used a tl072, and diodes for biasing would you say it is ideal enough? What if I use a 741 as the input buffer and a burbrown opa 24xx(forgot the model)for the 2nd op amp?

 

Anytime you have a op amp crossing over a bias thresh hold on your transistors (as you have drawn) there will be cross over distortion. Op amps have a limited slew rate (speed), even good ones, and there will be a dead space as one pair of transistors turn off and the other pair turn on. You need a way to slightly bias the transistors just barely on so they can turn on as soon as the signal is in their ball park. This is the reason for the diodes shown in Figure 3.

Just my 0.02¢.

BTW, please host your pictures locally, not on some other web site. We have many options to do this. The funny thing is, the image shown is off an early posting, before I learned how to do this. So do what I say, not what I did. If you need help just ask, I'll show you how.

 

One other point, use a inverting op amp and make a linear H bridge. It will double your potential output power. I'm going to design a really simple Class D audio amp in the near future, they also can be a good way to do this, as they can be very efficient.

An article I wrote a while back for the ebook.

Pulse Width Modulation

 

I was always under the impression the tl7/8x were decent opamps.

They are. The 741 is kind of archaic though.

Oh and could I just use a capacitor on the output in series to remove a dc offset?

You can, just as Ron H has done in his circuit. There's a slight downside... it will roll off low frequencies at some point. If you can get the output to where it is exactly zero or at least within 1/10 V you can omit the capacitor and, thus, no loss. If the output point is higher than that, I would work on resolving that issue first.

 

Bill I understand the issue with the cross-over distortion and slight bias need...I was just experimenting with my first class B and I wanted to hear the distortion or measure it and see it then add diodes later. I am more concerned getting the output power and general function working then adding the biasing..

BTW, I am interested in the idea of an H-bridge...I have read a little into it, can I use it with this design and would I include the darlington setup for each bridge component? How will it work, will I use 4 power output transistors to double the possible current? It would be awesome to double the output ability. I know of class D amps I am just starting this project as a learning experience. So i'd love to stay with it for now but increasing power output potential would be great though.

I am not so sure I want to switch to single ended supply because I have already soldered everything in place and ran ALOT of wires under neath my protoboard.

I asked earlier, does the oscope measure peak to peak, peak, or RMS voltage?

 

Okay guys, The way the circuit is, I just hooked up a dual 12V supply, previously I had issues like overheating of a transistor when using dual 12 but now it works fine! I measured a max of 5 Volts from my scope before clipping on music(highest number I seen) and I tested with a 6 ohm speaker. I just need to know what voltage scopes measure p2p peak rms? My power at this point would be little under 5watts if it was RMS I believe but I doubt it is. And to calculate RMS can I use RMS voltage over nominal speaker impedance and multiply that current by the RMS voltage for RMS output or does using nominal speaker impedance give me the peak current output?

Another good note, my LED blew lol and I noticed my DC offset dissapeared!! I believe it was probly because the LED but I did have it in paralell, paralell to both rails not center ground. I am still curious why it clips after a max of 5 volts. I understand a 3 volt per rail reduction from the opamp, and a 1.4V per rail from the darlington output but that should leave me with 7.6V per rail before clipping right? If my batts/power supply couldn't handle the current draw, it would start clipping from that alone correct? Well I am using 4 6V cells so I figure they will have no prob with current demand, reasonably.

 

Actually I was wrong I measured .96V DC offset at the output terminals. It seemed to have dissapered with the last speaker I tested. And I just got 5v just this instant and transistor overheating...must be something like close wiring...geeze..

EDIT: Alright I noticed a GND connection was touching my negative rail so I'm glad I noticed that hopefully there are no more crazy issues like that again...

 

You know why I don't like helping you, Chunk? I make suggestions, and you always come back with "What if I do this? Or that?". You are a constantly moving target.

@Bill Marsden: I don't advocate a class B audio amp. Chunk is trying to see what class B sounds like.

 

Hi Chunk.

Just spent a while trying to catch up on this thread.
Would you mind posting a schematic of your circuit at the present.?

I have another suggestion for you for a future project. A current limited bench power supply.

 

You know why I don't like helping you, Chunk? I make suggestions, and you always come back with "What if I do this? Or that?". You are a constantly moving target.

@Bill Marsden: I don't advocate a class B audio amp. Chunk is trying to see what class B sounds like.

What I can't come up with ideas or ask questions to learn? I'm not here to put together a scientfair project I am here to LEARN. The mroe questions I ask the more I learn. I have learned alot from the threads I have made and I am greatful for it.

Hi Chunk.

Just spent a while trying to catch up on this thread.
Would you mind posting a schematic of your circuit at the present.?

I have another suggestion for you for a future project. A current limited bench power supply.

lol I have one jerk I built an SMPS power supply but it's single ended...I used a voltage regulator smps package for an easy design it is variable aswell but like I said won't work. I wanted this project to be portable anyway.


On another note: I figured out my scope was measuring RMS on the default readout so I calculated about 5w rms of output. That is if the speakers nominal impedance is correct to use in the power equation instead of a reactive impeadance value. I still have that anoying dc offset, when no signal is playing, I get about a volt output from the second opamp and about .20v from the first opamp. When I turn the pot down to reduce "volume" the idle offset voltage output from the second opamp reduces. Well, the 1volt value is with the pot all the way "down". The first opamp idle voltage output dosn't change with pot adjustment. So, does this mean thaat the problem could be fixed with a change in the pot? I have added a 1k in the feedback loop in series and nothing on the other side I am not sure how it should be done.

 

Okay after thinking about it Ron I have decided to do the single supply design....One of the big advantages is because I can easily boost the voltage with my DC to DC boost converter and as long as my op amp can handle a decent amount of voltage I can push it upto the max. I just got back two op amps I sent to a guy and he sent them back I let him barrow devices with them in it. Boosting the voltage I can try and push for 10-15 Watts hopefully. This project turned out to be a guitar amplifier for a friend....I build subwoofer as a hobby so I am design an enlclosure later and build a subwoofer.

I have a tl082 and a Burbrown OPA 2134 IIRC...I am going to try the schematic posted earlier by you Ron H.

Ron the only reason I didn't want to listen is because I was on my last proto board and could tear it apart but I am going to radioshack to get a new and larger one...

 

Can the opa2134 Really take a supply voltage of 36v? That means I could possibly swing about +15V-15V on the output minus the VBE drops right? This would be awesome!

 

Ron in the single supply schematic, does the 2.2mF and 1mF indicate millifarad meaning uFx1000? Or microfarad? I am just trying to make sure.

 

Ron in the single supply schematic, does the 2.2mF and 1mF indicate millifarad meaning uFx1000? Or microfarad? I am just trying to make sure.

mF is millifarads. If I'm not mistaken, you had 2.2mF caps on each of your power supply rails in one of your schematics.