Thanks. I see what John means by large knobs.
Wow a Benchmark DAC1. That's one unit people like or hate.
I plan on contributing a more in-depth response to this thread, but not till tomorrow.
13 or 14 NE5532 opamps, 7818 and 7918 voltage regulators. Up-sampling. Ceramic caps in series with signal flow. Full with with "audiophile" hated" ingredients, and still sound soo good. (it really does sound good). This put some things in perspective doesn't it?.....
With kind regards,
Bas
13 or 14 NE5532 opamps, 7818 and 7918 voltage regulators. Up-sampling. Ceramic caps in series with signal flow. Full with with "audiophile" hated" ingredients, and still sound soo good. (it really does sound good). This put some things in perspective doesn't it?.....
With kind regards,
Bas
Yes, I really, really understand.
I was a factory rep for Benchmark and helped getting the DAC1 to market.
At first, I got into pro audio which was very successful.
Spent a lot of time with the people making our music and then we went to the consumer.
After Stereophile's placement into product of the year, we did well.
Apart from the debate over component selection/implementation, the one thing that stands out with Benchmark are the people who run that company.
They are committed to their customers.
Allen Burdick, president at that time, is a GEM among men.
John Seau, VP of Engineering at that time is, I believe, now acting president and still head engineer. He carries on that tradition.
Can't say enough good things about them and I hope for the greatest success for them.
Last edited:
When I state that the Benchmark could be 'improved' I am VERY serious. As a professional audio designer, I have done upgrades of SONY, Revox, Studer, etc over the decades. To think that a name manufacture gets EVERYTHING right, is absurd.
Often in a case like this, the designers are primarily digitally inclined, and the analog section is just an 'add on'. Any serious designer who uses ceramic coupling caps is ignorant. It only takes a SIMPLE measurement of the ceramic cap to show its distortion creation. Is distortion a good thing to add with a PASSIVE component?
Now don't get me wrong. Good designers are sometimes blind to these things. For example, I once showed the VP of Beverage in the '70's the distortion generated in a ceramic cap in my office. Later, I asked Herald Beverage if he changed the cap. He said no, and I then refused to work for him.
Later, at HK, I was called on ceramic cap distortion by the senior engineer at the time working there. I just took a ceramic sitting on his bench and used HIS test equipment to show the distortion, made by the ceramic cap. Yes, I rubbed his nose in it.
I also recommended to HIM that he make HIS preamp's output stage, class A. He refused, and so it goes, even on this website.
Often in a case like this, the designers are primarily digitally inclined, and the analog section is just an 'add on'. Any serious designer who uses ceramic coupling caps is ignorant. It only takes a SIMPLE measurement of the ceramic cap to show its distortion creation. Is distortion a good thing to add with a PASSIVE component?
Now don't get me wrong. Good designers are sometimes blind to these things. For example, I once showed the VP of Beverage in the '70's the distortion generated in a ceramic cap in my office. Later, I asked Herald Beverage if he changed the cap. He said no, and I then refused to work for him.
Later, at HK, I was called on ceramic cap distortion by the senior engineer at the time working there. I just took a ceramic sitting on his bench and used HIS test equipment to show the distortion, made by the ceramic cap. Yes, I rubbed his nose in it.
I also recommended to HIM that he make HIS preamp's output stage, class A. He refused, and so it goes, even on this website.
once again with the broad brush John? - you well know its not "easy" to measure np0/c0g distortion - at signal levels it will be below -100dB - and of low order
if DA really "explains" audiophile cap quality the Pease' measurements show np0/c0g very much better than silver mica - which is loved by the tube tweak crowd and the low K ceramics are better than some brands of polypropylene
I would still try to use polystyrene 1st in filter apps - but the fused tin foil caps with leads terminated on the foil ends are too inductive to soak up DAC switching glitch signals - their SRF will be in low MHz at best
for audio signal coupling/DC blocking the numbers suggest low signal drop across the cap is sufficient for low distortion - larger values of capacitance may trump capacitor "quality" - but again no reason not to use at least polypropylene dielectric
although I've just read where the "complex DA" of pet/mylar is desirable in some apps - from a "name" designer - with audiophile tube designs in the market
know anywhere to get low value non-inductive extended foil polystyrene - or even teflon in few 100s of pF - the they seem to be mostly made at much higher values today - maybe NOS, Phillips boxed caps?
if DA really "explains" audiophile cap quality the Pease' measurements show np0/c0g very much better than silver mica - which is loved by the tube tweak crowd and the low K ceramics are better than some brands of polypropylene
I would still try to use polystyrene 1st in filter apps - but the fused tin foil caps with leads terminated on the foil ends are too inductive to soak up DAC switching glitch signals - their SRF will be in low MHz at best
for audio signal coupling/DC blocking the numbers suggest low signal drop across the cap is sufficient for low distortion - larger values of capacitance may trump capacitor "quality" - but again no reason not to use at least polypropylene dielectric
although I've just read where the "complex DA" of pet/mylar is desirable in some apps - from a "name" designer - with audiophile tube designs in the market
know anywhere to get low value non-inductive extended foil polystyrene - or even teflon in few 100s of pF - the they seem to be mostly made at much higher values today - maybe NOS, Phillips boxed caps?
Last edited:
I used a 0.1uF C0G combination made by 3x33nF// to try and see if it is obviously bad subjectively against DC coupling and Teflon, PP film, PIO caps very recently. The amplifier's impedance was high enough (120K Ohm). Did many iterations. Teflon seemed very near to DC coupling to me, the C0G proved more vague, was like a softer Auricap (had Auricap too), but obviously better than Wima MKP 10, or basic SCR that I had there. Its not conclusive, its non controlled, but if the C0G it does not bother me in that way of check, I doubt it will ever will.
P.S. Benchmark DAC-1 has an SMT lytic bypassed by two progressively smaller SMT ceramics (don't know what class). I know it because I have one. Its the only coupler, and does obviously better if upgraded.
P.S. Benchmark DAC-1 has an SMT lytic bypassed by two progressively smaller SMT ceramics (don't know what class). I know it because I have one. Its the only coupler, and does obviously better if upgraded.
Attachments
I found that the mechanical behaviour of a cap is important. Micas have the advantage to be extremely rigid. The type i am using gave me the best results so far. The biggest i can get is 10nF. That is usable in RIAAs and active filters. I use them as coupling cap in my line level filter. I got very good results as decoupling cap in my new MC stage that has a frequency response to 50Mhz with a gain of 30dB, so gain bandwidth product in that stage is 2GHz. Yes, it can be done and the Micas stabilised HF oscillation very well. Maybe they are not all made equal.
Hi John:
As you may know, I was a factory rep for Benchmark.
I don't take your comments on the DAC1 as a negative personally.
In fact, I welcome it.
Obviously, Allen Burdick and John Seau (Benchmark) were constantly being
asked many questions by me.
They were always glad to speak with me at length, but my knowledge, especially at that time was limited regarding the things you bring up.
Things like their use of those op amps and the ceramics, etc. were always cumming up. Not because I mentioned them, rather audiophiles.
Allen once told me that although he was aware of all these things, he said that in the end it's how they all worked as a whole.
Not that the following proves anything, but I never heard any comments from the pro audio guys with very few exceptions.
They plugged it in, listened and said better than most for the price.
Regarding the analog output quality, from your expertise and my lack, I was only able to discuss with clients, at that time, the published data, as well, measurements in the field and the press.
The specs which at least to me, are a fairly large set and I could never find typical measurements that did not appear as excellent.
So, either these type of measurements are not comprehensive or other things, as I have pointed out in other threads, are not forthcoming.
As for its digital performance regarding locking-to-signal and decoding to analog, it appears to be very robust with my experience in varying studio interfaces.
However, and this is what I am going to write about tomorrow, but briefly here.
Even as their rep, I had immediately pointed out that, once I had it up and running, that suddenly most of my CDs sounded worse!
The minority of the CDs sounded much better.
A question arises here as it did then regarding linear translation of source data from many commercial CDs.
To many who are not at the recording venue, a question remains as to what is accurate.
Once I got into the studios with it and did some comparisons, things were different.
Even then, evaluations were often not practical in a comprehensive manner.
Unless I have the opportunity to test what you hold to be representative of an excellent DAC, I truly can't say what that is... for now.
Since many of my evaluations of audio are really sourced from my own compositions and studied with regard to how well they will translate to most consumer setups, the DAC1-for now-serves me well.
I do wish to here what you think can be done to it to improve upon the performance at a reasonable cost or as I have requested in another thread.
Recommendations please for a DAC that outperforms the DAC1 at around 700 dollars used.
Any thoughts?
As mentioned above, I look forward to expressing my view on the whole matter of audio electronics as I continue with this thread.
Where else can I get this amount of intellectual attention
I've pointed out that high op amp GBW is "linearizing" in I/V converter applications
the "noise gain" low AC impedcance shunt at the DAC out also cuts peak delta_V at the op amp input
the ADA4898 offers a unique new linearized input stage which even better deals with DAC switching glitches/large I steps
and as far as I can tell Hawksford's nested loop op amp I/V topology isn't being used - I see no reason to avoid really good op amps rather than the really cheap ones he called out:
http://www.diyaudio.com/forums/digital-source/168721-other-hawksford-iv-nested-loop-op-amps.html
the "noise gain" low AC impedcance shunt at the DAC out also cuts peak delta_V at the op amp input
the ADA4898 offers a unique new linearized input stage which even better deals with DAC switching glitches/large I steps
and as far as I can tell Hawksford's nested loop op amp I/V topology isn't being used - I see no reason to avoid really good op amps rather than the really cheap ones he called out:
http://www.diyaudio.com/forums/digital-source/168721-other-hawksford-iv-nested-loop-op-amps.html
Last edited:
Dear,
I was a bit sarcastic about the DAC1. But the point is still valid. For the money you get a very good product, that has a very low noise floor, and sounds good and detailed. Everything can be improved, since every commercial design is a compromise to start with. (every designer knows though, that those compromises can be excited and excites the mind to look out of the box)
The point I wanted to make is, that designs with NE5532's and 78XX/79XX doesn't have to be bad. I think the people at Benchmark have very good PCB design skills, and that is for sure one of the reason why it sounds so good.
Btw. Guess what Burr Brown advice in their reference application for the PCM1798.. Not even their own highend OPA627's but NE5532's
My company was committed a few years ago in modifying the DAC1 and we sold tons of them modified.
The first thing I disliked about the DAC1 where the three paralleled capacitors (one ceramic, one electrolytic of 1000uF! and one film) in series with the signal flow. We removed them and designed an add on PCB with a dedicated (but simple) DC servo circuit which replaced the caps's. The DC servo was designed in a way to maintain the differential stage equal impedances with the added servo circuit, to keep common mode noise rejection.
The stage Benchmark use behind the DAC is a 1:1 copy from the AD1853 app. sheet. Benchmark placed those three capacitor between the output of this stage and a unity gain amplifier. After this there are a bounce of opamps for the outputs enz. all DC coupled.
Removing those caps was one of the biggest sound improvements. A certain "strain" that often get blamed by the NE5532's was simply gone and where caused by those cap's
Second, The I-V converters we replaced with AD797's. In my personal opinion still the best opamp for I-V conversion. We tried also I-V conversion with a single transistor as transimpedance amplifier which worked very well. End of all the AD797 won.
Those who are interested, I still have al documentation of our mod's.
See photo: On the lower right the DC servo circuit. On the top the Tentlabs Clock. The rest speaks for itself. The DAC1 has a huge ripple in the power supply lines because of the low capacitance. We solved that the brute way with adding capacitance The AD797's are placed on little PCB's that make from two single one dual SOT8 opamp.
With kind regards,
Bas
I was a bit sarcastic about the DAC1. But the point is still valid. For the money you get a very good product, that has a very low noise floor, and sounds good and detailed. Everything can be improved, since every commercial design is a compromise to start with. (every designer knows though, that those compromises can be excited and excites the mind to look out of the box)
The point I wanted to make is, that designs with NE5532's and 78XX/79XX doesn't have to be bad. I think the people at Benchmark have very good PCB design skills, and that is for sure one of the reason why it sounds so good.
Btw. Guess what Burr Brown advice in their reference application for the PCM1798..
Not even their own highend OPA627's but NE5532's My company was committed a few years ago in modifying the DAC1 and we sold tons of them modified.
The first thing I disliked about the DAC1 where the three paralleled capacitors (one ceramic, one electrolytic of 1000uF! and one film) in series with the signal flow. We removed them and designed an add on PCB with a dedicated (but simple) DC servo circuit which replaced the caps's. The DC servo was designed in a way to maintain the differential stage equal impedances with the added servo circuit, to keep common mode noise rejection.
The stage Benchmark use behind the DAC is a 1:1 copy from the AD1853 app. sheet. Benchmark placed those three capacitor between the output of this stage and a unity gain amplifier. After this there are a bounce of opamps for the outputs enz. all DC coupled.
Removing those caps was one of the biggest sound improvements. A certain "strain" that often get blamed by the NE5532's was simply gone and where caused by those cap's
Second, The I-V converters we replaced with AD797's. In my personal opinion still the best opamp for I-V conversion. We tried also I-V conversion with a single transistor as transimpedance amplifier which worked very well. End of all the AD797 won.
Those who are interested, I still have al documentation of our mod's.
See photo: On the lower right the DC servo circuit. On the top the Tentlabs Clock. The rest speaks for itself. The DAC1 has a huge ripple in the power supply lines because of the low capacitance. We solved that the brute way with adding capacitance
The AD797's are placed on little PCB's that make from two single one dual SOT8 opamp.With kind regards,
Bas
Attachments
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.