Home
AudioAsylum Trader
Digital Drive: FRENCH ART - more on tweaks by Claude G

Upsamplers, DACs, jitter, shakes and analogue withdrawals, this is it.

For Sale Ads

FAQ / News / Events

 

FRENCH ART - more on tweaks

213.228.27.88


[ Follow Ups ] Thread:  [ Display   All   Email ] [ Digital Drive ]
[ Alert Moderator ]

Hello ART and music fans, I have now carried on other tweaks on the ART DI/O with some – hopefully- interesting results. Warning, this is a long post!

POWER SUPPLY (+6)

I am now running on my self-made PS. It is a small quality transformator with 2 outputs of 9V/1.66A. I switched them in parallel in order to obtain a high quality supply with 9V and over 3.3A. Why not a big unit? First, if I believe that 2A (my previous PS) are just a bit underrated for a state of the “art” performance, a PS of around 3A proved to be enough to get the best of the DI/O under any circumstance (2 line filters in serial didn’t give any restriction). Bigger PS do only bring complication IMHO. For the technical freaks, my ART is now running (with line filters) on 9.62 V.

On the other side, wallwart of 3.4A are an improvement, but not the best achievable as their quality is no outstanding. At least, running to transformators in parallel cuts the impedance and resistance by 2, and this proved to work very well in my power amp (in response, control, and quickness in the highs).

And indeed, whereas a big supply improves mainly the bass and the stability over the range, and a bit less the highs, my quick PS brings the same performance all over the range and improves even further the highs and the rapidity of the response. The improvements on impulse, aeration and precision of the soundstage and the quality of the highs are well another +2 worth in comparison with my previous PS. However, bass and stability are nearly (a bit more impact) untouched, but there was little to improve this way.

I recommend highly this PS as it really makes a big difference on the sound, is easy (no intern tweaks) and cheap (around 15$ plus casing and small parts… and DIY).

My only concern remains still the deepest bass (my system begins to cut off at 24Hz in my listening room). The ART has now a nice and deep bass, but it seems that from 45Hz the bass is somewhat slowed down, and begins slightly to cut off from 35Hz (deepest notes of a contrabass for example) down. 20Hz is still there, but the impact from 45 Hz down is a bit shy / poor. Op amps? Caps? We shall see…err, hear!


LINE FILTER (+6)

Still the old one, already described. It gets one more point, because in conjunction with the new PS, it is now on its best without any restriction on dynamics on complex passages. The problems I had experienced were indeed linked to the poor PS, that simply couldn’t give all the energy in the bass that the filter allowed the ART to give (because the bass and highs improve a lot with a better current/PS).

Together with the new transformator, the results are outstanding and made a somewhat bigger sonic improvement then for example the existing one I experienced recently between an AN DAC 1 and DAC 3.1 signature.

This is a must, go for a good PS and line filter, I do not know how hardly I should recommend it!


OUTPUT CAPS C17 & C18 (+4)

They are there to avoid DC current to get into the inputs of the next unit (pre amplifier for example). These are awful electrolytic caps, 22uF/25V, quite easy to access on the output board (the 2 blue caps on the top).

As most units do have a DC blocking device on input, this is a redundant mounting, so I simply removed them and replaced them with 2 little bridges. The best cap is still no cap at all : a wire! Unless you have a very special set up, of course.

And what a difference! The soundstage improved in all directions, the noise went back, the precision and rapidity all over the range increased, the band extended in the high and low frequencies. The biggest improvement was that the silences were no more silences, that the notes remain longer, the transparency and accuracy was incredible and I could now listen to parts that only the very detailed Audiomat had revealed to my ears. An entirely new dimension, you rediscover your CDs.

However, when I did this, I expected mostly to remove the bass cut off. It did, bass is extremely differenciated and goes now down, but I still miss some impact. In fact, the biggest improvements came in the highs, with an incredible resolution. Middles won a lot on expression and I was seating in front of Cassandra Wilson, feeling her moving in every direction!

But there is a reserve. The original set up was outstanding, very musical, yet not the most precise. This tweak (be careful inside the unit, thin paths) did moved it in direction accuracy (Audiomat, ML), but did not improved the musicality (the same). The sound is now more direct, and the very highs may sound harsher on some recordings (where you barely noticed there was something with the caps). In fact, there is now no curtain between you and the music, so it is less forgiving on average records, and you also listen directly to the very average op-amp outputstage (in the original ART the caps are masking it) with some weakness on resolution, speed, bandwith, impact, musicality, S-tones aggressivity. But as the accuracy is higher, the discovery also, and the sound not more stressing, I would simply say the pleasure and musicality remains the same with (despite?) a lot more details, so a very good move and a good base for a future op-amp tweak?


GROUNDING

Some strange findings. As stated, grounding the unit made a difference, but only when grounded to a wall plug. This is the only unit to be grounded on my HIFI, as the others gave no improvement that way (not even my REGA).

However, I renewed a couple of days ago my entire power supply. My hifi is ran as ever on a separated HIFI deserved wallplug, the ART Ground is from the beginning on it.

I had a big surprise when taking out this wall plug : there is no ground it, neither on the others, just a “façade decoration”, only 2 in my entire flat are really on earth!! Dear “old” Paris…

Just to make sure I was not crazy when hearing such a difference, I plugged the ART on another not grounded plug, and again the difference was clearly there. I made a discovery : when the fridge was turning on, and the ART “nearly grounded” to this normal plug, nearly nothing happened, whereas when not connected it nearly switched off! So I am not crazy, but I still do not know how it works, but it does! Explanations and experiences are welcome.


X2 MODUS

Asking ART on that subject, they advise me that running on anything else then “external” would bring random errors. I do not know what happens with X2 as they stated it was not an upsampling device, however, I prefer it that way. Note that the guys from ART are really friendly and professional, but also that they stated that there was nothing (or nearly) do to in order to improve the sound of their DI/O… err, however, the European units now come with a 1.2A PS instead of the dangerous 0.9A one, a good point.


OTHER TWEAKS

It a pity I can not get a better scheme then the incomplete photocopy I have (missing half of the output stage down-page and the entire power supplies with regulation (?) and diodes up-page!).

DIODES
Special Thanks to Gary with his welcome warning on the diodes, I will follow blindly his advice and replace only the important ones for the sound that handle 15V or less. I want to use mine Schottky diodes SB 130 (30V/1A) for CR1, CR11, CR12, CR4, CR5. Is it OK now? No voltage problems or else?? Once done, I will report.

OUTPUT STAGE
The last step, unless somebody finds the tweak of the year, will be the output op-amp TL0-74. I have found some replacement for it (AD 713, TLC074…), but none can match the excellent OPA627 (but a tad warm and slow) or AD827 (very neutral, yet a bit dry). I am thinking of an alternative, so help is welcome.
Place is cramped, but if nothing appears to me, and if the power supply of the TL074 can stand it (how many mA reserve?), I might go for a radical solution : a montage made of 2 outstanding CMS AD825 (the best sounding ones ever?) and a dual op-amp in replacement of the old quad op-amp. Surely the best unless you go for a triode cathode follower, but that’s a bit more involving and not pratical. However, I expect as much improvements from this tweak, as my PS gave me! It is also a fine way to do a fine tunig on the sound, as it is the last change (?)

ELSE
I considered other tweaks, like the caps or the regulator. Anyway, as everything really works well on this side, and as it is a bit dangerous to try around, I would simply say : if it ain’t broken, don’t try to fix it! In France we say, unlike sex, the less you touch it, the better it is! You need to do the tweaks in a reversible matter, and one by one over a period in order to notice the differences.

Another subject is the digital board. It can easily be taken out, so maybe ART will one day bring an update?? That will surely be the best solution, if there are improvements to achieve in that area.

For the moment, however, tests with jitter killers did not improved the sound, and that means either our beloved unit is crap and do not notice the difference, or the CS4820 is doing a very good job. I bet the last one.

There are newer DACs and CODECS than our AK4524, like the AK4528 (not compatible!) from AKM. Having a close look, you will find this one is only better as ADC, its DAC part is the same. You might want a 192 kHz DAC, or one with better integrated op-amps in the output stage (ours seem to be average, but as they are not separable…), but do not forget that this is the heart of the sound, and it sounds outstanding good (better than the Hermes with a pure DAC from AKM!), so why going for something else and take lots of risks/alter the set up?

At last, I considered the clock. Our crystals are average (maybe 100 ppm?), this might affect the clock jitter (not the other jitters). I had some good experience with my Trichord clock, compensated in voltage and temperature in order to give 2 ppm. It worked well on small Pioneers (+4, but still a lot of money), very well with old DACs, but improvements were very low (+2 maybe) as soon as newer Delta Sigma DACs were considered. As our DAC is using a switched capacity technology (nearly not sensitive to clock jitter), I expect no improvements at all (?). Note there are 2 crystals in the DI/O.


Anyway, comments or ideas are welcome.

Read you soon, enjoy music


Claude


This post is made possible by the generous support of people like you and our sponsors:
  Kimber Kable  



Topic - FRENCH ART - more on tweaks - Claude G 18:24:10 10/10/01 ( 17)