In yesterday’s post we covered the problems with IC op amps concerning their huge amount of gain. Problem is, you simply cannot control how much gain they have and therefore are left to take what you get and simply deal with it for better and (mostly) worse.
The second biggest problem IC op amps have is their power supply voltage which is typically 30 volts across the whole op amp. Why is this limiting? For a couple of reasons: headroom and linearity being the two biggest.
We first ran into the headroom issues when we starting building phono preamplifiers back in the early 1970′s. At that time we used a pair of IC op amps as I’ve mentioned before. The two op amps were used to form a passive RIAA equalized phono preamplifier. This layout is fairly simple: the first op amp is connected directly to the phono cartridge and sends the amplified signal through an equalization network of a couple of capacitors and resistors. The second op amp then amplifies the equalized signal to the proper level. The problem with a passive phono preamp is that the high frequencies coming off the phono cartridge are “hot” by a factor of 10 – meaning they are really loud and we ran into headroom problems almost immediately.
So here’s the headroom issue: any amplifier I’ve ever worked with starts sounding somewhat worse the louder the signal that goes into it. You’ve probably heard this yourself on a good system. The orchestra or group plays really loudly and the sound that used to be open and dynamic starts to close down and sound a little less open and dynamic. That’s just the nature of amplifiers in general – the closer the signal gets to their maximum output the more compressed they sound. To fix this designers only have a couple of choices: they can lower the gain of the stage so the output doesn’t get so high or they can raise the power supply voltage, thus moving the maximum loudness level of the stage up. This is known as increasing the headroom.
In our second generation phono preamplifier we moved from an IC op amp that was limited to 30 volts power supply to a discrete version we designed that could handle twice the power supply voltage: 60 volts. The difference in sound quality was amazing and we spent a number of hours comparing the discrete version at 30 volts vs. 60 volts and became instant devotees of higher voltage as a means of opening up the sound.
The next big problem with running lower power supply voltages is one of linearity. Every amplification device has a range of operation where it is more linear to a signal – meaning that it is more faithful to the original – step outside those bounds and it becomes more non-linear where the output isn’t the same as the input. This area of linearity is controlled by the amount of voltage across the device. Yep, power supply voltage.
So if you put 15 or 20 volts across a transistor the linear region will be a percentage of the total voltage – double the voltage and the percentage stays the same, but of course the area of linear operation doubles. Pretty simple. If you want to read more about this we ran a series a while back you can brush up on here.
To this day all PS audio designs from the lowest cost to the highest cost, as well as a few other noteworthy brands in the high end, use discrete high-voltage op amps as opposed to low voltage IC op amps in the signal path. We do this because we can run significantly higher voltages which improves headroom and linearity, we can control the open loop feedback so we don’t get any harsh sound, choose which combination of transistors gives us the best sound – like FET’s in the input and bipolars in the output – and have complete control over how open, musical and alive the sound is. If we need to voice the circuit a little differently it’s easy with a discrete op amp and basically impossible with an IC op amp.
Where are these op amps used in high-end audio? Pretty much every DAC ever made has at least two: one to convert the output of the DAC to a voltage and the other to amplify that voltage, preamps, integrateds, power amplifiers, phono stages etc. As a rule of thumb, most tube circuits are NOT op amp blocks (although they could be) and most power amplifiers don’t use IC op amps because there’s too much voltage required.
So, to end this discussion I think it’s important to reflect on what we’ve learned – that if an audio designer wants to be able to manage every aspect of his/her sound, the functional block called an op amp is a perfect format – but we as music lovers would all be better off if designers spent the time to design their own instead of choosing an off the shelf solution that was easy – yet limited.
When you’re thinking about making the purchase of a DAC, a preamp or any type of source equipment, remember that in 99% of the solid state designs a functional op amp will be used in the signal path. Now you want to ask what type is it.
There may be many good reasons designers choose an IC and some may sound extremely good. I just believe they can sound better if the designer takes the time and care to go the extra mile and be discrete baby!
Forward to a friend and help us engage more readers
Recent Comments
Paul McGowan: Excellent post Mark, as usual. I agree with you completely on this point and it really brings to the...
Soundminded: I always like these postings that stand back and try to look at more than one or two trees but...
Bassman23: It’s important to remember that both microphones and loudspeakers are members of that class of...
Lawrence Schenbeck: I had the same impression of the “fate” theme when I came back to the Fourth, to...
Soundminded: First impressions of the first movement, four ideas seem evident. Romeo and Juliet Overture stands...
Soundminded: This two disc set is from analog recordings originally made in 1975-1976. It’s DG 453 088 2....
Lawrence Schenbeck: You have to admit, though, the third movement seems a long ways away, expressively, from the...
Soundminded: I’ve got a set of the Karajan recordings of 4,5, and 6 on CD and I’ll be listening to...
stimpy2: I believe that Andy Narell’s pans were crafted by Ellie Manette who is I believe the preeminent...
stimpy2: I enjoyed reading reading this article very much and have been at several of his concerts and up close...
Paul McGowan: Sure. The idea is to build an active differential low pass filter whose rolloff is low enough to not...
Paul McGowan: No worries – but I don’t actually have any control over where these post or when,...
stimpy2: I just joined the group and was under the impression (notice that I am not using the word assumed) that...
Rick Dieterich: I have been searching for a new sub(s) for my system and one brand I’m considering is the...
Soundminded: Does your use of the term “servo” in this context mean frequency dependent negative...
Dan Gravell: Great comment! It comes down to how your music player works. Personally, I don’t see filenames...
countessouroboros: I realize this post is a bit old now, but for future viewers I feel I need to point out that...
Paul McGowan: Thanks Scott!
sbrinck: Paul The Munich audio show was the bomb !! I am extremely envious ! If you ever decide to quit your day...
Paul McGowan: Yes the apertures are located in the 1/8 space corners so they should be as effective as possible....
Soundminded: I’m quite curious and I must admit somewhat skeptical about how effective these frequency tuned bass...
Paul McGowan: Aging memory Mark. I wrote what I remembered and then went back and actually measured it.
The Q...
Soundminded: Paul, on April 27 you posted “The main listening room at PS Audio is approximately 16 feet wide...
oliver T. Finch: Simple. Treat the cause and there won’t be any symptoms.Regards.
Paul McGowan: You’re certainly correct and I’ll do my best to give a full report.
Paul McGowan: Thanks! Glad you’re enjoying.
petewilson: As suggested above, it’s a Good Idea to fix the cause(s). But, in general with real-world rooms,...
Soundminded: In response to the question about treating the cause or the symptom in medicine it’s my...
jb4: Thank you for very interesting posts and video’s. I totally agree with your approach, cure the cause,...
oliver T. Finch: Excellent posts. I look forward to them everyday. Very informative too.I am being convinced more...