What is wrong with op-amps?

Status
Not open for further replies.
Jump to Latest
#3,102
mondogenerator said:
In my latest new effort to make the simplest possible composite power amplifier, I have also had similar successes and failures, with both monolithic and discrete.

It is these failures that have taught me (and probably many others in the own lifetimes) what these devices are capable of, and what the limitations are.

Imagine if you will indulge me:

LF353 op amp, directly coupled to PP discrete output stage, biased simply, with a single diode at each Base (TIP41/42). NFB from output returned to the inverting op amp input, using appropriate Rf and Rg for Av of 10-11.
Supplies limited to +/- 15V...

What could possibly go wrong?

😀

Besides running out of current to drive the bases?
Resultant limitation to max V swing.
Measly 3 W before the onset of clipping.
Jeez at first, I didn't even thermally couple the diodes...why does 3 watts seem.more like 20 watts of thermal runaway?

Hahahah. It's all a learning curve. I revel in my mistakes. I welcome the next funk up. Do I need some more scrapbook calcs? Definitely. Iterative work only gets me so far. Next to work out why I couldn't iteratively get RC coupling to work. Needs some maths lol. Then a driver stage to drive the bases better and load the LF353 less, probably BC441 haha. (Parts bin cheapskate build)

But for 3 watts and less than 10 components....it sounds great. I'd still like to have the ability to measure THD so I could see how much my ears and brain ignore.

Expectation bias? Maybe. Probably. But I had fun.
Click to expand...

Speaking of an expectation bias... ;-)

If you connect all bases together, no bias, to the input of your opamp, there should be huge nasty crossover distortions, right?

Not necessary... Add a resistor (something like 47 Ohm) between bases and emitters and listen... And compare... 😀

Sometimes it is better to discard odd entities, searching for Graal. 🙄
 
#3,103
famousmockingbird said:
Nothing fancy, most will scoff but it suits me fine.

I mainly listen to ripped CD's that are .flac files or stream via Spotify. My computer is old but works fine for me, it has a decent soundcard:

http://www.analog.com/media/en/technical-documentation/data-sheets/ad1884a.pdf

My turntable isn't anything fancy either. Denon DP-300F w/ a 2M Red cart.

My main speakers are Klipsch La Scala's and I usually run either my home brew tube amps of various topologies. I got triode strapped sweep tubes single ended, a nice Tubelab SE board w/ 45's in it. Mullard 5-20 style push pull but with less open loop gain and overall less feedback. For SS I have a Slewmaster CFA I made that I usually use that to power my Klipsch RB-81 II's. If I am feeling frisky or having a party I supplement La Scala's with a subwoofer, two 15's powered by a Peavey 2U rackmount two channel PA amplifier.
Click to expand...

thanks for being upfront and sharing...
 
#3,104
Waly said:
Been there, done that. More caps are on the board back side, including Vcc to Vee. Not my call for SOT23 vs. SO8. Those 6 layers eval boards are expensive, almost x10 the cost of the chip.

I agree, it sounds a little thin in mid bass, also I've noted some sibilance on high pitched female voices. Good for punk rock, Nina Hagen sounds great, fails short for opera.

I guess I have to check the OPA695, perhaps it will improve the sound 😉.
Click to expand...

Interesting.

And being quite serious, to what do you (or anyone) attribute the "sibilance" to??

_-_-
 
#3,105
Wavebourn said:
Is anything stellar under the Moon? 😀

That's the question of the topic! 😉
Click to expand...

No, the question, restated slightly is if opamps have any deficits - in this case when used for audio.

Which when the topic is beaten into the ground from several directions ends up being a question of the audibility of differences between opamps and/or does such an audible difference exist.

Fun is fun, but when 20-30 posts go by with nothing substantive, what's the point?

_-_-
 
#3,111
morinix said:
Just curious, why did you use the 4558 and not the good old standby 5532?
Click to expand...

Because I took TOA 216 console as a donor, it contained almost all 4558 OPAMPs. Instead of replacing I just biased them by 7k5 resistors, that was enough. Why should I fix what is not broken, if there were other changes needed to give better sound and convenience?

The same with the reververator... If Behringer reverb is nice, why should I pay extra money for Lexicon label?
 
Last edited:
#3,112
So Bear I'm all ears, what did you mean by this comment? If I took it wrongly just correct me, no problem at all and we can move forward.

The basis? Having done it? Having been present when others were trying to do it so that there was no difference? Having been friends with a recording engineer for a (now defunct) highly regarded classical label... stuff like that.
Click to expand...
 
#3,114
Wavebourn said:
Because I took TOA 216 console as a donor, it contained almost all 4558 OPAMPs. Instead of replacing I just biased them by 7k5 resistors, that was enough. Why should I fix what is not broken, if there were other changes needed to give better sound and convenience??
Click to expand...

Hi Wavebourn, do you put the resistor to the negative rail so npn devices conduct? Or does it not matter so much?
 
#3,116
famousmockingbird said:
Hi Wavebourn, do you put the resistor to the negative rail so npn devices conduct? Or does it not matter so much?
Click to expand...

Wavey is doing things right. The 558 opamps use a lateral PNP transistor in the quasi complimentary output stage. These are very poor transistors compared the NPNs. By using a tie to the power rail it basically takes the transistor out of the circuit.

Replacing the chips with a 5532 would increase the chances of oscillation as they are not as tame as a 558. As this console uses Tamura input transformers there would not be any improvement in noise performance.

Now for a behringer reverb as it is used as an effects box the levèl is lower than the direct signal. Thus most distortion issues would be masked. The only issue would be the quality of the reverb template. Since they stole that from decent units it should be fine.

The specific console was really made at the height of moderate cost analog design. Newer ones were targeted at lower cost. Transformer inputs have a slight disadvantage in studio use for inherent distortion. For field use the RF rejection, noise free gain and real balanced lines are a plus.

Now Bear marches to a different drummer and this level of gear is less than what he normally plays with. Never the less the console in good condition probably is among the best ever 5% of live sound boards and as modified perhaps the top 1%.
 
Last edited:
Status
Not open for further replies.
Top Bottom