Depends on the preamp. Most separate preamps are much better than the stock unit. With volume control mods however, it can be quite marvelous. It trounces my fully modded Mark Levinson #38, which I thought was quite transparent. Have to keep getting up to adjust the volume, but it's worth it for the sound quality.
Lots of info on this. Please do a search.
For me, direct into amp is superior. Many others think otherwise.
Thanks for the replies.
My preamp is the ModWright, so it would be interesting to see what the ModWright add/subtracts.
I will be trying both RCA and Balanced outputs direct into my amp and see the results. Cables are Slinkylinks. See how things go...
Skipping a preamp is just like adding one .
What I mean here is synergy . Just as preamps need to be matched to an amp , output impedence to input impedence , so does your DAC's output impedence need to match your amp's required input impedence range . Also , the quality of your volume control needs to be considered as mentioned above .
Ok, I've done the RCA and AES/EBU route for comparison...and the winner is [drum roll]...
Anyone want to hazard a guess???
"Ok, I've done the RCA and AES/EBU route for comparison...and the winner is [drum roll]..."
One of these is an analog connection and one is a digital connect. I don't understand what you are doing?
I was under the impression that you were asking if it is better to have a DAC (w/volume control) connected directly to an amp or through a pre-amp. In this scenario you would be using analog outputs from the DAC in both configurations.
Yeah, CMK didn't mean AES/EBU, I'd guess. He meant the balanced analog outputs. As I recall, from the Stereophile review, the impedance of the unbalanced output is quite high, so it won't be surprising if you hear a difference.
Bob_reynolds - the output impedance of the balanced outs depends enitrely on the jumper settings. The place that I set them makes the impedance 30 ohms, just right.
Sorry for the confusion, yes its the analog outputs - RCA and balanced (isn't this AES/EBU?) from the Benchmark, directly into the RCA/Balanced inputs on my Sanders poweramp.
After Saki's post, I checked the input/output impedences of both the Benchmark and the Sanders. For the Benchmark, the difference in output impedence between the RCA and balanced is that the latter is double that of the former. For the Sanders, there was no difference in input impedence between both RCA & balanced. Consequently when using RCAs, the sound was double the loudness of the balanced termination.
Since the ModWright SWLP is only line level, no balanced connections, I could only compare the RCA direct vs via the preamp. I'd say its a close call, with a slight preference for going through the preamp.
When comparing sound quality of the RCA vs Balanced direct to the poweramp, there was no contest. The Balanced output is significantly better than RCA. Everything that you would normally associate with balanced - blacker/bigger soundstage, deeper bass, more solid images, is all there with the balanced outputs. The difference is so obvious I was wondering if I had upgraded to another DAC. Not only was I getting more information, it was totally natural and musical. If you are still using the RCA output, you have NOT heard the true potential of the Benchmark.
Steve, my comment was about the "unbalanced" outputs. Though Benchmark specs them at 30 ohms, JA measured this:
All the outputs preserved absolute polarity, and the source impedance was a low 60 ohms from the balanced jacks, less than 1 ohm from the headphone output. The unbalanced source impedance was on the high side at 1230 ohms, but this should pose no problem with normal preamps.
Does running direct to the power amp in balanced configuration produce a better sound than going through the Modright preamp unbalanced?
Certainly that's what I've been trying to say. The Benchmark is essentially a balanced DAC, so the "correct" or optimal way to use it, is via its balanced outputs. The difference is not subtle.
I wonder if the output impedance of the balanced outs suits your amp better than the impedance of the rca outs ?
Might be worth thinking about , for others , rather than accepting a blanket conclusion that one connection type sounds better than another .
What really matters is that you found what works best for you . Congratulations and enjoy the music .
Well certainly the impedence of the balanced outputs match my amp's input. Actually with the Benchmark, its a simple case of turning the attenuator pots down to use the RCA outs, but then you won't be able to enjoy the superior sound the balanced outputs produce.
My point was not so much that XLR cables are better than RCA, rather that the Benchmark's XLR outputs are far superior to the RCA outputs, so if those of you still using the RCA outputs, I would suggest you try the XLR connections on the Benchmark. Some of the common criticisms of the Benchmark - titzy highs, lean sound, are completely gone! You'll wonder why you never tried it in the first place.
I'm really enjoying the music, thank you.
PS: WARNING TO REMEMBER TO SWITCH THE TOGGLE SWITCH AT THE BACK TO "VARIABLE" OR YOU MIGHT RISK BLOWING YOUR AMP/SPEAKERS.
Bob_reynolds - this high RCA output impedance is only in the older versions, more than three years old. The new cisrcuit board has new buffers and does the voltage divider between buffers rather than at the output pins. The new RCA output impedance is 30 ohms, a bit too low IMO.
Cmk - the DAC-1 converts single-ended to balanced at the output drivers. I would not call this a balanced DAC, but I suppose there are other interpretations.
Steve, thanks for the update. I emailed Benchmark the same question yesterday.
Just curious... Why do you consider 30 ohms to be too low?
I'm not technically inclined, just a bit of tech knowledge, so I can't tell what's really balanced or not, but I do know this, the Benchmark's XLR outputs are a quantum leap over the RCA. Something which 2 others I know have heard, and confirm the same findings.
Since you obviously know more about the internals of the Benchmark, maybe you could shed some light on why there's so much difference between the two?
Bob_reynolds - when I first started modding DAC's I put no additional resistance on the outputs. What I found was that with high-Q or low-loss cables connected, this evidently caused transmission-line type reflections on the cable which could take a while to damp-out whenever there was a HF transient in the music. My theory is that these reflections cause the output drivers to go non-linear at times, which affected the audio range.
If I approximately matched the characteristic impedance of the cables (like you would do in a RF system), the sound improved dramatically at the high-frequencies. Most analog cables are in the range of 75-120 ohms characteristic impedance, never as low as 30 ohms.
There are other engineers that have tried this and reported similar results.
Steve, thanks for your explanation. I'm way over my head here, but I thought transmission line issues don't manifest at audio frequencies. That's why no one has been overly concerned about the lack of a 75 ohm RCA connector. But, you think that at the upper end of the audio band impedance mismatch reflections to exist.
Bob_reynolds - I dont believe that the impedance mismatch and the resulting reflections directly affects the audio band. This is a secondary or indirect effect of some sort. May not happen with tube outputs.
Cmk, this is great info. As my amp CJ 2500a has RCA only I'm wondering if it worth getting some adapters just to try the XLR outputs on the Dac1 directly to the amp?
Adasilva - on the recent DAC-1's, using the balanced outputs as single-ended is the same as using the RCA outs.
Adasilva, definitely worth getting adaptors. If you listen to digital only, you have all to gain and nothing to lose. I'd even go so far to say that it compares favourably with SACD and good analog.
Have you actually tried to listen to both outputs before concluding that they are the same?
I ask this because one of the users who reportedly heard the improvement used adaptors on the XLR output to drive his amp and he heard a drastic improvement in sound. In his words "Tried this last night. Jaw dropping detail[ed] and dynamics."
See also the following links:
cmk: running balanced into your Sanders amp sounds better than running single-ended, so you conclude that we should all listen to the Benchmark via its balanced outputs. Isn't is also possible that the improvements you are hearing are a result of running the Sanders via its balanced inputs? Maybe this is not about the Benchmark at all but about the Sanders.
Have you actually tried to listen to both outputs before concluding that they are the same?
Of course. I mod a LOT of these, so I listen to them continually.
The folks that are reporting better results probably have the older DAC-1 that has high-impedance RC outs. The newer units have low impedance on both Balanced and RCA's, and separate buffering as well. The buffers/drivers are identical on both.
Since last trying the Benchmark directly to my amp I've changed it's power cord to the Cardas GR and the nickel plated outlet to the Oyaide SWO-XXX. All the time running through my tube preamp to amp.
Trying it again directly to amp with the new PC and outlet and I must say I'm very surprised by the results (rca to rca). Perhaps the GR power cord and the SWO-XXX reduced the harshness and quieted the background noise. But for the first time I'm not unhappy with the sound of the Dac1 to directly to the CJ2500a.
Oh, I've also added the DH Cones. Thanks Cmk....
Signal path inside is identical with or without volume control. In one case front volume pot is active and in the other trimpots are engaged instead. Additional components like preamp or extra set of cables cannot make it sound better but often add a little bit of distortion making sound "livelier". Very old Benchmarks had problems with output impedance - fixed in earlier revisions. I have DAC1 driving directly (XLR) Rowland Model 102 with great results. Protect the switch on the back to prevent accidental full power to the speakers.