Digital Drive

Posted by Claude G (A) on August 10, 2002 at 13:07:50

Hi Music Lovers,

It has been a long time since I hadn’t open my ART again. There was no real need for it, recording quality being by far the limiting factor of CDs.

However, as I know have nearly finished my TT (just waiting for a Goldring 1042 cartridge), and as a lot of guys were still asking me some questions on the ART, I decided to go back to, err, work.

Among the newcomers is Carlos Martinez, a Brazilian audiophile. Most of the credits of what comes next goes to him, as he was the one who really pushed me further into tweaking, who gave me at last the complete ART schemes (allowing me now to look at the complete PS), and who helped me understanding the awful voltage multiplier, as well as giving me some good advices. So, thanks a lot Carlos! The fact that we might go different routes in order to improve our units will just enable all of us to improve things further.

But let’s start with an easy tweak…

FINAL OUTPUT FILTER REMOVAL (+3)

I believe for 10 years now that most of the digital sound of the CD is due to the digital and analog filter. I guess that everybody agrees nowadays with that. There is little to improve with the digital filter inside the AKM. I do neither believe that a cap upgrade around the active op-amp analog filter will really improve things in a sensitive way, nor a resistor upgrade nowadays (parts are already very good usually)…

But I do believe that altering the filter values could change a lot. Therefore, I decided to look first at the 72kHz final output filter (C59, C60, R7, R9). Looking at the active filter, and considering that I use the 88kHz modus, I decided to get rid of that supplementary filter. Despite the 45kHz limit of my hybrid loudspeakers, as I don’t have batears, I have to say that I am unable to hear sounds above 23kHz (test 5 years ago), neither a single resistor into the signal path. Therefore, I left R7 and R9 in place, and simply removed C59 and C60 to get rid of one of the filter pole.

Only hearing sounds from 16Hz to 23kHz does not mean you can’t perceive lower and higher frequencies (bones resonance and co), but I believe mostly that even people with a bad ear’s bandwidth are sensitive to frequency coherence, impulse response, and phase shift. So trying to listen without this high frequency filter makes indeed sense, and, as there is still plenty of active high frequency filtering, you don’t need to worry a lot about your amp.

To put it straight, I preferred the sound without filter. The sound is a lot more open, dynamics are better, the soundstage is much deeper, and, mostly, there is a lot of analog air between the instruments now, erasing the usual cramped / closed digital sound. The treble seems completely set free and very natural, but the bass corpus does improve as well. In fact, everything sounds much more transparent and natural.

Are they any drawbacks? Not really. Well, that is, the sound changes a bit (brighter?), and you might need to get used to it. Also, as with the output caps removal, you now listen to music directly, so, if your unit is not perfectly tweaked, you might prefer less transparency to mask the defaults.

Last but not least, you might prefer a more compact and digital sound, as well as your system may not react in a positive matter to this tweak. But, as we speak about an easy cap removal, it would be a shame not to try it…

My sound is now close to the bright side, so that I don’t whish to improve transparency further, especially with the current record quality. Therefore, I will keep the existing active filter without altering it, the sound being analog and open enough.

On the other side, as I can hear a lot more now, I whish to improve somewhat the corpus in the middles, the purity of the treble, as well as the bass and the dynamics, while retaining the excellent transparency and the outstanding musicality and balance. Even if there is currently no way to complain really, I will therefore look to the very promising and last weak part of my ART.

THE POWER SUPPLY INSIDE THE UNIT (to be finished)

Some of your guys know that I am working on the caps for a couple of months. I believe that even a friend of mine posted a question on that subject last month earlier. Now that I can see the entire schemes, I have to say the PS inside the unit is a mess. No wonder it reacted so sensitive to all the previous tweaks related to the PS…

A lot of things could be done, as building an entirely new external PS : some did it. I didn’t wish to go that route, I like it easy, things being not that bad, but I hope that Carlos will go that route and compare it to mine.

In an ideal world, it would be good to replace the Voltage Regulators with better ones : I was told the + and – VRs (7815/7816) could be upgraded by LM340-T15 / LM320-T15, and the +5V LM2940-CT-5 VR by a LM340T5. Also, if you don’t care about the ADC capacity of the unit, it would be good to get rid of the parts around Q1, as well as of the parts around VR2. But I personally care about reversibility and functionality, and I like it reasonable.

I believe 2/3 of the remaining potential in that area can be obtained by working easily on the caps I listed bellow, and, as we might well speak about a +4 tweak, that would mean that the supplementary hassle, cost, risk and restrictions could well just be a further +2 worth… But I might be wrong (???)…

At last, using the low power consuming AD825, with their very good line noise rejection, is especially in the ART’s case of advantage (no place for big PS and caps), and results may vary if you use another output amp.

So I decided to work on what I believed to be the most sensitive parts, ie some selected caps. My experience (and works at the French Audiophile in the early 90s) is that an average / good cap can be lifted to the highest level, higher then the most expensive audio grade wonder caps, provided you bypass it cleverly by several chosen caps of increasing quality. I guess you all know cap’s bypassing and dimensioning, but not necessary the requested steps in the values, nor the number of requested bypasses.

IMHO, the caps to consider are, in order of importance:
- C44 and C36
- C19
- C9, C7 and C6
- C3, C4 and C5
- The remaining parts should then not be really sensitive IMHO

In an ideal world, I would increase the values of the above by quite a lot (using good audio quality Philips electrolytic caps for example), bypass the biggest of them with a middle size cap (Philips), and bypass all of them (so bypass again for some) with the excellent silver square Siemens MKT goodies, my favorite ones (say 0,47 micF).

But place is a big issue inside the ART, as well as the part’s supply (no Panasonic available here, so sadly no 3300 mF and 680mF for me that could fit instead of the existing 2200mF and 470mF). Also, the double sided board is very sensitive and not really resistant: I didn’t want to solder and desolder caps several times on the main board, and ending up ruining my unit finally! On the other side, there was nearly as much space under the board as on it, so I preferred to solder the requested supplementary caps in parallel under the board, as place allows.

First of all, I wished to bypass all the mentioned caps with the silver Siemens, so using a total of 9 of them. But, I am not sure then to be able to close the unit safely, so I might well have to switch to smaller caps, maybe Wimas, for the definitive final report on a closed and well ran in ART.

C44 and C36 just need to be rated at 16V, so I still hope to increase their value to around 100mF (instead of 22mF/25V).

C19 just need to be rated at 5V, so I will increase value here as much as possible (100mF as well instead of 22mF?).

And that’s all, as I believe we then will get the most of it!

I hope to report soon, and remain open to any suggestion (please use my direct mail) before this ultimate tweak (?)

Enjoy Music

Claude G