This is a fantastic discrete OPA designed by AMB labs: A20
The α20 Line Stage Amplifier
I did an exhausting comparison with Pass B1 in multisim (ah, this multisim...). B1 has an amazing bandwidth up to 1MHz and A20 up to 0.6MHz. The issue is that, B1 has no gain instead A20 has gain. Both, can't be used in inverting mode like a consistent OPA.
The α20 Line Stage Amplifier
I did an exhausting comparison with Pass B1 in multisim (ah, this multisim...). B1 has an amazing bandwidth up to 1MHz and A20 up to 0.6MHz. The issue is that, B1 has no gain instead A20 has gain. Both, can't be used in inverting mode like a consistent OPA.
while discrete op amps may be fun to design there is little possibility of matching the last decade's best monolithics in any performance parameters except extreme low noise input suitable for mc phono input or for higher current, or V output swing
the shown circuit's 1K series input R wastes the noise potential of the discrete jfets
and composite op amp techniques allow using +/-15 Vs, 500 mA CFA output op amps with the best for the application input op amp
for line level source and output you're just fooling yourself if you think you come close to today's best monolithics used in composite amp circuits for audio frequency accuracy
monolithic IC designers have more expertise, better tools, transistor processes and the A/DSL, Ultrasound and other high speed, precision analog signal processing markets have given them incentive to push op amp capabilities much harder in the last decade
the shown circuit's 1K series input R wastes the noise potential of the discrete jfets
and composite op amp techniques allow using +/-15 Vs, 500 mA CFA output op amps with the best for the application input op amp
for line level source and output you're just fooling yourself if you think you come close to today's best monolithics used in composite amp circuits for audio frequency accuracy
monolithic IC designers have more expertise, better tools, transistor processes and the A/DSL, Ultrasound and other high speed, precision analog signal processing markets have given them incentive to push op amp capabilities much harder in the last decade
Last edited:
Technically you are right jcx. I think monolithic op-amps nowadays are absolutely superb. Having said that, the 30 year old SA5532/4 ain't half bad either.
However, you are talking primarily to enthusiasts here, and no matter how much engineering expertise you use to try to point out the benefits of going 'monolithic' , people are still going to go discrete and to their ears, it will sound better. Remember that 'dirty sand' letter to Sterophile a few years back from one Charles Hansen about the LM4562? LOL
However, you are talking primarily to enthusiasts here, and no matter how much engineering expertise you use to try to point out the benefits of going 'monolithic' , people are still going to go discrete and to their ears, it will sound better. Remember that 'dirty sand' letter to Sterophile a few years back from one Charles Hansen about the LM4562? LOL
Hey guys, i'm not a "Golden Ears" freak 
. Bonsai, he knows me and can confirms that i am "Measurements" freak . This A20 discrete, is quite unprotected against ripple noise due to absence of any CCS. To be noiseless, it needs a perfect power supply. I don't remember how many IC based devices (x-overs, mic-preamps, compressor-limiters, DI-boxes, Noise-gates etc) i made in my 25 years career on PA systems. I know very well that IC implemented devices are noiseless yet with ordinary regulators - 78xx, 79xx.
I like very much discrete for their "dynamics" in low frequencies. I like their "tight bass" reproduction. This can be proved, with a 100Hz square wave test. ICs present tilt, instead discrete no.
. Bonsai, he knows me and can confirms that i am "Measurements" freak . This A20 discrete, is quite unprotected against ripple noise due to absence of any CCS. To be noiseless, it needs a perfect power supply. I don't remember how many IC based devices (x-overs, mic-preamps, compressor-limiters, DI-boxes, Noise-gates etc) i made in my 25 years career on PA systems. I know very well that IC implemented devices are noiseless yet with ordinary regulators - 78xx, 79xx. I like very much discrete for their "dynamics" in low frequencies. I like their "tight bass" reproduction. This can be proved, with a 100Hz square wave test. ICs present tilt, instead discrete no.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.