Well, better than us having to read the AI slop all the time having to correct the same type of errors over and over. We have to do this for humans already.. leave the AI to deal with the AI
-
Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
The "DAC-as-Preamp" Trap: Why DAC Buffers Bottleneck High-End Amplifiers
- Thread starter SanojS
- Start date
- Status
Professional microphone cable has 0.22 mm and can be used for long distances, so not stupid. With sufficiently low output impedance you can use it for 200m or more.
- Joined
- Sep 4, 2026
- Messages
- 97
- Likes
- 135
I don't know anything about this stuff. This is where ASR really teaches me. So I have a question. Some source outputs, like a DAC, do not feel like they have the output to really drive a particular power amp even if they have a volume control. That's why I have a Benchmark DAC3hgc. It feels powerful enough to easily drive anything and has internal, via jumpers, gain adjustment. So is the issue really just a matter of proper gain staging between devices? Or is there something more I am missing?
Yes. That there is no issue whatsoever.
With 24dBu/12.2Vrms, it better does.
- Thread Starter
- #65
Yes you are right, I made a doodoo! Thanks!
This is what I was looking for! I couldn't make sense of what I was seeing... of course... face palm...
Finally someone who can give me right answer!
- Thread Starter
- #66
I do want to understand my original question that had nothing to do with slew rates, in which circumstances the DAC-only with op-amp at 200Ohm/20mA (as an example) does need a preamp... technically. I still think there is a limit where the current drive of the DAC will not be sufficient. This is where I am trying to understand if there is such a scenario, where the analog signal will reach the limit of the opamps...
A Class-A discrete preamp stage or even solid state can comfortably idle at 50mA to 100mA or more (like examples I am considering T+A P1000, Sugden Headmaster, etc...). they have much larger current reserves compared to the tiny op-amp chip inside the DAC.
Is this important or is it audiophile mumbo jumbo?
Or is this just a design philosophy question between GNF amps (like opamp using feedback loop to keep signal in check) vs open loop (aka "zero" feedback like a classic Class-A preamp with high power reserves)
A Class-A discrete preamp stage or even solid state can comfortably idle at 50mA to 100mA or more (like examples I am considering T+A P1000, Sugden Headmaster, etc...). they have much larger current reserves compared to the tiny op-amp chip inside the DAC.
Is this important or is it audiophile mumbo jumbo?
Or is this just a design philosophy question between GNF amps (like opamp using feedback loop to keep signal in check) vs open loop (aka "zero" feedback like a classic Class-A preamp with high power reserves)
Last edited:
Why would there be?
In general there really isn’t a major difference in driving an amp or pre-amp.. they generally all have high input impedance and offer an easy to drive load. There are always exceptions to the rule to be found, but there are not that many.
- Thread Starter
- #68
I was just making a joke... In any case a microphone cable is balanced and has twisted pairs, with opposite phase and error correction, which an RCA does not... so... not the same.
Of course. You only need a pre-amp if you have other analog sources. Otherwise, a DAC with built-in volume control is ideal. That is how I run my desktop system. On the main system, I have a Reel to Reel deck so have to use a pre-amp. But even there, I have taken the tape deck out of the loop and drive the system from the DAC/Streamer direct.
Remember, it recent times, I have tested DACs with 600 ohm load impedance and results show superb performance with well engineered DACs. No weakness is seen. That is a much lower impedance than just about any power amp you are likely to use:
Assume ~3V peak signal (that's what DAC outputs deliver, 2Vrms. 100Ohms output impedance, 10k load, 1nF cable+input capacitance (typical for any reasonable cable and cable length). Let's look at the start of a 100kHz sine wave (is that transient enough?). Let's assume electrically short cable so that cable capacitance can be assumed a lumped capacitance, no transmission line effects.
Vin is the voltage at the opamp output, before 100R isolation resistor. Vout is after it. I(R2) is the opamp output current.
Voltage slew rate is ~2V/us. Maximum current is below 2mA, often within the class-A range of many opamps.
The current slew rate at the sudden onset of the 100kHz sine is on the order of 20mA/us but that is a completely unrealistic scenario because such transient cannot be created by any DAC with any available music content. For a full-scale 100kHz sine the max current slew rate is ~1/20th of that, around 1mA/us.
All of that is easy cake for any competently designed opamp-based output stage in any modern DAC. An OPA1612 at gain = -1 has about -130dB distortion at 3Vrms output into 2k, throughout the whole audio range and beyond (blue curve).
Last edited:
- Thread Starter
- #71
Yeah, I’m in the same dilemma, which is why I originally posted my question and got a bit of a grilling!
I'm trying to understand the exact same thing. Gain staging is definitely part of the equation, but my theory is that output circuit topology and current reserves have an impact too.
Many DAC output stages rely on op-amps with heavy Global Negative Feedback (GNF). The feedback loop compares the output signal to the input to erase distortion, the idea is to give an accurate signal with a low output impedance (is my understanding).
The issue I can see happening if you pair that DAC directly with a demanding power amplifier that presents a low or highly reactive input impedance, the op-amp might climb outside its optimal current range. When load demand climbs, op-amp struggle with the load, leaving less feedback factor available to correct errors. This alters its harmonic distortion profile long before the chip hits a hard clipping ceiling.
This is exactly why some high-end DACs bypass standard op-amps entirely, opting instead for high-bias discrete Class-A output stages or solid-state buffers.
I am still not understanding exactly the parameters that control this behavior and what to look for. And in which scenarios I DO need a preamp. You found a solution by getting a DAC with more power reserve - Great!
Someone else is happy with DAC performance but add a preamp - Great!
I'd like to understand if I can predict based on components and parameters of specifications, if I can know beforehand!
What is frustrating to me is that manufacturers don't post optimal or even max current loads on DACs.
They only list the voltage and the static output impedance. For me, this does not tell the whole story...
No, the reasons for this kind of design philosophy are different. First and foremost, create a unique selling point for marketing.
One can design a 100.00% flawless output stage with simple low-cost opamps consuming only a few mA, and there are ways to reduce the distortion to levels way beyond measurable limits (using so-called composite amplifiers), with the actual distortion only limited by passive parts, connectors, cable itself.
- Thread Starter
- #73
Yes, this is your classic static bench test driving a 50 Hz tone into a flat 600-ohm resistor and proves the op-amp has excellent steady-state linearity.Of course. You only need a pre-amp if you have other analog sources. Otherwise, a DAC with built-in volume control is ideal. That is how I run my desktop system. On the main system, I have a Reel to Reel deck so have to use a pre-amp. But even there, I have taken the tape deck out of the loop and drive the system from the DAC/Streamer direct.
Remember, it recent times, I have tested DACs with 600 ohm load impedance and results show superb performance with well engineered DACs. No weakness is seen. That is a much lower impedance than just about any power amp you are likely to use:
I don't argue this point, I agree 100%!
But in real usage, a power amp and cabling do not present a flat, resistive load across the audio spectrum. A GNF-based Op-amp might struggle if it does not have enough current available to handle both the output load as well as its own internal feedback error correction in such a scenario. No?
No. See my posts.
- Thread Starter
- #75
Ah yes, very good point about composite amplifiers! I do love vintage class-A preamps, but mostly because they were the shit in my youth... and the classic way to handle it is with more power and current... I can add that the reason I love those old designs is that they're also highly serviceable by normals like myself, while the modern designs with loads of surface mounted and components requiring hot-air soldering are a nightmare... becoming land fill when they fail...
So what you are saying, is that indeed simple opamp output stage has an issue in current delivery, and the way to handle it is to add an additional opamp as a power reserve, if I understood your technical description correctly?
- Thread Starter
- #76
Yes got it! Saw your post after that answer...
- Thread Starter
- #77
Thanks! This is what I was looking for!
If I read it correctly, I shall pay high attention to the opamp circuit to ensure the DAC and it's analog circuit is designed like you say, I personally believe a lot of people pay too much attention on the DAC chip rather than the analog stage...
solderdude
Grand Contributor
Imaginary problem thinking solved ?
Anyone designing a power amp will do so considering to not load up the input as to cause the very problem you mention. This is the whole idea of a power amp! To make the difficult load of the speaker not appear on its input. The demands you talk about are handled by the power amp. You are worrying about problems that do not exist.
- Status
Similar threads
