Is your target curve always the same, or does it vary depending on the recording?
In general I would assume DSP would give you much more flexibility than swapping buffer circuits in the hope one of them would be close enough to what you need, but there is one thing that a buffer circuit can add that a DSP oriented towards optimizing frequency response usually can't provide - distortion. My suggestion is to use DSP software that allows for various harmonic distortion plug-ins to find out the parameters of your desired target response, and only then figure out how to achieve it as part of your normal playback chain.
I wrote "one thing that a buffer circuit *can* add". Are you of the opinion that the possible audible differences between different buffer circuits is purely down to frequency response?
Julf,
I almost always use the same target curve, for all the music I listen which is very varied. Sometimes I make subtle changes to on the LF side of the band, but I usually end up coming back to my original curve.
THD and Freq. Response are both different kinds of distortion. Anything that changes the output from the input signal can be said to be distortion.
I think that if using an input buffer makes difference from using another one, or not using anything at all, there must be not only measurable but listenable distortion in at least one case.
Can just a change in freq response lead to listenable changes, like wider and deeper soundstage, or only tonal improvements?
If only lowering THD would lead to better detail, soundstage etc, then wouldn't be the best option to bypass the buffer and drive the power modules directly? Assuming the preamp has that driving capability of course.
I almost always use the same target curve, for all the music I listen which is very varied. Sometimes I make subtle changes to on the LF side of the band, but I usually end up coming back to my original curve.
THD and Freq. Response are both different kinds of distortion. Anything that changes the output from the input signal can be said to be distortion.
I think that if using an input buffer makes difference from using another one, or not using anything at all, there must be not only measurable but listenable distortion in at least one case.
Can just a change in freq response lead to listenable changes, like wider and deeper soundstage, or only tonal improvements?
If only lowering THD would lead to better detail, soundstage etc, then wouldn't be the best option to bypass the buffer and drive the power modules directly? Assuming the preamp has that driving capability of course.
Good - that makes things easier.
You are of course right, but often the term is reserved just for non-linear distortion (distortion that generates new signal components - harmonics and intermodulation products) as opposed to linear distortion that just emphasizes or attenuates signals based on their frequency.
Yes, as long as we verify that the audible difference is real, and not caused by things like expectation bias.
Most definitely yes. As far as I know, Bob Carver's famous Carver Challenge feat was accomplished purely by tuning the frequency response.
Yes - but what if *increasing* THD/IM leads to better detail, soundstage etc.?
It's not completely clear (at least) to me, what resources he used to null the difference between his solid state amplifier and the transformer coupled "high end" reference tube amplifier.
This can certainly be the case. Some time ago, a user in this forum compared his new made class D Anaview BTL amplifiers (which are close in performance to Hypex Ncore, highly undistorted units) to his OTL tube amplifier. He opined, without doubt, that the tube amplifier had more resolution in the high freq region than the Class D, while the latter had better detail in the lower freqs. From a THD specs perspective of both units, this did not make sense, unless what you mention abobe is valid.
I exchanged a couple of emails with Colin (Nord NC500), asking for the possibility of including a "buffer bypass" switch in a future version of his input buffers.
Would anyone else be interested in having this option in a NC500 design?
Sebastian
There's already a buffer bypass jumper on the Hypex OEM boards. Just order the cheap Nord amps if you want to do that.
Gain can't stay the same without the op amp and feedback resistors which are... the input buffer. Also, don't forget the low impedance of the amp module - yet another reason for the input buffer.
The volume control on the preamp does a pretty good job of adjusting gain. Unless you have a preamp with no volume control. And yes if your preamp can't drive a 1500 ohm load with ease, bypassing the opamp isn't a good idea. Any preamp worth it's salt should be able to drive a 600 ohm load with ease.
I'm using a Benchmark DAC2, which surely can drive the NC500 modules directly.
The Benchmark output stage uses LM4562 Op Amps, similar to the ones used in the standard Hypex input buffer.
I would like to try Colin´s buffer, with its options to exchange Op AMps, but still be able to have to option of driving the NC500 modules directly from my DAC.
I think my scenario is not very unique, on the contrary should be rather common.
Julf, I wonder why you mention the gain as a requirement. Only concern that I can think is the risk of switching "live" from direct connection (low gain) to the input buffer option, not sure how many dbs can that add to the chain...
Sebastian
The Benchmark output stage uses LM4562 Op Amps, similar to the ones used in the standard Hypex input buffer.
I would like to try Colin´s buffer, with its options to exchange Op AMps, but still be able to have to option of driving the NC500 modules directly from my DAC.
I think my scenario is not very unique, on the contrary should be rather common.
Julf, I wonder why you mention the gain as a requirement. Only concern that I can think is the risk of switching "live" from direct connection (low gain) to the input buffer option, not sure how many dbs can that add to the chain...
Sebastian
Keep in mind that the LM4562 opamp when used in the NC-500 input buffer is intended to be driven by a decent preamp, or DAC with robust output stage already. If you use that output stage in the DAC on its own as a gain stage for the DAC and the amp, it's working that LM4562 pretty hard. The better thing to do would be bypassing the LM4562 in the DAC and using Colin's Rev C board with SIL 994!
Last edited:
I would think that the Hypex input buffer is not difficult to drive at all, even high output impedance tube preamps could apply. Did you have any issue with your NC500 and preamps to make that claim?
Totally 100% agree with you, I would love to be able to bypass the LM4562 Opamps in the DAC, but it doesn't look like a simple mod to do, could not find anyone online who done it.
I'm talking about if you bypass the LM4562 on the Hypex input buffer, not driving the input buffer. If you want to bypass the buffer on the input board I would recommend a real preamp. And a good one at that.
It is only a requirement if you want to use the switch to compare the sound, but as SGK pointed out, it is not really possible.
Mike, the Benchmark is a "real" preamp. It can switch inputs, has good independent analog and digital volume controls, and the output stage is strong with low impedance. It even provides tools to optimize the gain structure.
A strong point and weakness at the same time, is that the Benchmark is really neutral. It is not a preamp one can rely on to add anything to the sound.
What, in your opinion, a "real" preamp would add to the Benchmark capabilities, apart from perhaps coloring?
Sebastian
Julf, are you sure the input buffer can't be configured for unity gain?
I know many buffers just adapt impedances. However, not sure if the Hypex (or Colin's) buffer can do that, and no idea what the sound impact of making them unity gain would be.
In my case, a unity gain input buffer would be even more desirable, not just for comparing sounds, but to optimize the gain structure with my DAC. I currenlty have to insert attenuation between it and the amplifier, to be able to use the digital volume control in the recommended zone.
Sebastian
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.