Motorola Hi-Fi power amplifier

[audioguru2]

Hi Guys,
We have a nice new audio power amplifier project:
http://www.electronics-lab.com/projects/audio/041/index.html

I think the author makes an error when he says to put a current meter on the output and adjust the trimpot for zero current reading. I think the adjustment is to reduce offset voltage at the output to zero so there isn't a DC voltage on the speaker.

I don't see any temperature compensation for the output transistors like nearly every other audio power amp includes. Temperature compensation prevents thermal run-away. As the output transistors heat-up during normal operation, the heat causes their base-emitter voltage drops to decrease, causing them to draw more current. The increased current causes them to heat-up more and draw even more current and around and around it goes getting hotter and hotter.

I also don't see any emitter resistors for the output transistors like nearly every other audio power amp includes. Emitter resistors would help equalize the gain of the output transistors and they reduce the tendency for thermal runaway.

A 741 opamp? I won't say anything until someone carefully listens to the amp. ;D

 

[Miles Prower]

I think the author makes an error when he says to put a current meter on the output and adjust the trimpot for zero current reading. I think the adjustment is to reduce offset voltage at the output to zero so there isn't a DC voltage on the speaker.

The instruction is ambiguous at best. He could mean to connect the ammeter in series with the speaker, then adjust for as low a current as possible. Regardless of whether you adjust for minimum DC voltage across the load, or minimum DC through the load, I would prefer to make that adjustment into a dummy load, not the speakers.

I also see a half fast attempt to implement a Sziklai pair. As for the output stage, Doug Self referred to it: "...Darlington configuration, the latter implying an integrated device with driver, output, etc in one ill-conceived package." The Darlington package might be good for voltage regulators, or a quick and dirty solution where you aren't particularly concerned with fidelity. Otherwise, avoid. Same goes for the 741 (I checked to see if the MC1741 specified here might be an enhanced version. It's not.)

I won't say anything until someone carefully listens to the amp.

Definitely. "I am very contented with this amplifier. It gives a very good sound quality.
Have fun with it!" Who knows? May be it does. I've seen similar claims made for some of the strangest designs from those who should know what they're talking about.

 

[audioguru2]

Hi Miles,
I wouldn't put an ammeter on a very low output impedance amplifier if its offset voltage wasn't zero. The only current limiter is the fuse. (the fuse in your meter);D

Years and years ago when audio amps started using opamps in them, the designers demonstrated extremely low distortion at 1kHz and a very flat response from 20Hz to 100kHz. But they got a bad reputation because of their poor sound.
Then someone discovered that the amps with old 741 opamps in them had high distortion above 10kHz and others said it didn't matter since we can't hear the 3rd harmonic at 30kHz and above.

Transient InterModulation (TIM) distortion was invented and was extremely high on the amps that used old 741 opamps, because they can't slew fast enough. The poor sound was because they couldn't slew as quickly as the many transients in music. Besides the severe distortion caused by slew-rate limiting of the output (triangle waves), the inputs are also severely overloaded because the negative feedback is delayed.
An old 741 opamp is slew-rate limited above 9kHz. Modern (for the last 20 years) opamps at 100kHz. An old 741 opamp doesn't have much gain at and above 20kHz where high gain is needed for high negative feedback, modern opamps have plenty.

What do you think about the thermal runaway issue with this amplifier? The use of darlingtons makes it worse. ;D 

 

[Miles Prower]

I wouldn't put an ammeter on a very low output impedance amplifier if its offset voltage wasn't zero. The only current limiter is the fuse. (the fuse in your meter)

I wouldn't either. That's not what I believe he's telling us. It would seem he wants one lead of the ammeter to be connected to the output, with the other lead being connected to the speaker. That way, the resistance of the voice coil will protect the meter. Of course, too much DC will burn out the voice coil. An ammeter is cheaper than decent speakers.  :  He should have specified the use of a 10 Ohm dummy load to make that DC offset adjustment. In any case, it's far better to put a voltmeter in parallel and measure DC offset directly.

An old 741 opamp is slew-rate limited above 9kHz. Modern (for the last 20 years) opamps at 100kHz. An old 741 opamp doesn't have much gain at and above 20kHz where high gain is needed for high negative feedback, modern opamps have plenty.

That's why I wanted to see if the 741 used in that project was an "enhanced" version. Upon Googling it up, the specs hadn't changed a bit. Same rise time, same 0.5V/uS slew rate, etc. (Although I like 'em for use in voltage regulators, comparators, undemanding audio applications (audio stage of longwave comm receiver, where you don't want much more than 3.0KHz anyway.) They're dirt cheap, and easy to get, easy to use.

What do you think about the thermal runaway issue with this amplifier? The use of darlingtons makes it worse.

I have seen Darlingtons that incorporate on-chip thermal run-away protection. If these don't, then I'd guess it's only a matter of time till the finals blow. I could see one or the other latching to the rail. That could bring a 10W resistor to red heat in seconds. That would take out the finals, and the speakers.  :'(  (I believe he's had this problem already. He includes a fuse in series with the speakers. This, too, is a no-no from an audio quality perspective.) Regardless, I wouldn't use them anyway from an audio performance stand point. I'd also like to see a couple of resistors between the Darlintons and the output. The Sziklai Pair topology does include these.

I'd have to see some test results on this design. If that thing is hi-fi, I'd have to hear it to believe it. (Then, again, I'm not necessarily the one to ask about that anyway.)

 

[Alun]

Am I the only person who's missed this or am I just having a dumb moment? ;D

ow does the signal get form the output of the op-amp (pin 6) to the complementary pair?

The output isn't conected to the input or the complementary pair in any way so how can it possibly work?

 

[audioguru2]

Hi Alun,
We have a 200W amp in our projects section that does the same thing. Except it uses a 10W wideband power amp IC instead of a lousy 741 opamp.
http://www.electronics-lab.com/projects/audio/013/index.html

When the output of the opamp goes positive, it drives a massive (!) current into its load, 47k R11. The current increases the opamp's positive power supply current by the same amount, passes through T1 and developes a huge (!) voltage across R7 which is fed to T3.
The same thing occurs with negative swings of the opamp increasing its negative power supply current and driving T4. T3 and T4 amplify the small voltage changes across R7 and R8 and their own temperature change by quite a lot. ;D

 

[ante1]

This seems like an odd way of doing this I guess!? :

 

[audioguru2]

Hi Ante,
It is a simple way to reduce the supply voltage of the opamp and still allow the output transistors to have their full supply voltage swing. ;D

 

[ante1]

Hi Audioguru,

Yea, I can see what you mean but it’s like using an “overload” to play with!  ;D ;D