advertisement


Mods to Naim CDS3

Today I was talking about doing exactly that.... the best way to upgrade the tant cap is to *replace* that tant cap. The 4.7 uF Wima may be a perfect candidate, although technically having a lower coupling value may result in early LF rolloff if the CDP is used into a preamp with a lowish input impedence.

Then simply parallel two 4.7uF Wima's as your base capacitor.

Another idea;

Given that you'd prefer to use all Cardas, but don't have enough space within the CDS3, why not simply bypass the Tants and fit your Cardas cap's in a suitably sized external 'coupling cap' box. In fact, an external box would mean you can very easily experiment with various combinations of capacitor without the PITA it must be to delve inside the CDS3 for every change. You might hit on something that sounds great and yet is small enough to be housed back in the CDS3.

Mr Tibbs
 
Ron - I'd suggest something else to look into:

The PCM1704 dac, which the CDS2/2 is based on , has a bipolar current output which is symmetric about 0v (+/-1.2mA)

Unless Naim's I/V + discrete output stage has a base Vbe offset w.r.t 0v (which I doubt) you should the find the output just before the muting relay is within a handful of mV of 0v when playing digital silence.

If so - i.e. if the offset is under, say 20mV DC - simply remove any coupling caps and bridge the gap with a piece of wire. It's quite safe, and the absolute best cap is NO cap. You may not belive the difference this can make...

(NB does *not* apply to CDS1 / CD1/CD2/CD3, because 0dbFS with the TDA1541 is at -2mA, which leads to a +3.4v offset... which I also have a way-around)
 
I will have Daryl look into that. I brought up this very topic a couple of days ago and he told me that there may be some DC offset AFTER the relays (which I use to toggle between the RCA and the DIN ouputs, although now I am commited to a non-Naim preamp, the DIN outputs are superfluous) which would make it next to impossible to get rid of the coupling cap. BUT, if as you suggest I took the signal out BEFORE the relays, not only would I lose the coloration of the coupling cap, but also that of the relay. Food for thought that needs to be consumed, methinks.

Ron - I'd suggest something else to look into:


Unless Naim's I/V + discrete output stage has a base Vbe offset w.r.t 0v (which I doubt) you should the find the output just before the muting relay is within a handful of mV of 0v when playing digital silence.
 
DC offset is only really significant if you fin more than a couple of tens of mV after the relay, whne the output is loaded with (say)10K. A DMM alone will show all kinds of nonsnense.

If you do find a consistent DC offset, it can be easily/sufficiently nulled by injecting a DC curent from a CCS added upstream. It will still sound a lot better than any cap. I can provide suitable schematic suggestions if reqd....
 
(NB does *not* apply to CDS1 / CD1/CD2/CD3, because 0dbFS with the TDA1541 is at -2mA, which leads to a +3.4v offset... which I also have a way-around)

Hi Martin,

With the CDS1 it is not possible to be removed the output coupling caps. There are no output coupling cap! Why does it function nevertheless?

Ronald

cds1.jpg

cds1u.jpg

cds1b.jpg
 
Now It played a track with digital silence on my CDI (not the CDS from above). And the voltages on the CDI output coupling cap are ~3.3V. Why however does the CDS without coupling caps get along?

Ronald
 
Apparently the problem with hotwiring around the output caps and the relays could be spurious DC surges when the machine is put in pause, or there are scratches on the disk. I will have some testing done and see if this is even possible.

Daryl has already listened to the 4.7uF Wima comparatively, and finds them soft and mushy, much preferring to use the physically larger Solens. These though have a pretty bad resonance circa 3k, which would be greatly palliated by bypassing them. He suggested squeezing in a pair of the 10uF Solens bypassed with the Cardas ones, and taking the tants completely out of the circuit. This project may have to wait some time, as the results I am currently getting are sonically so spectacular, I would hate to change a thing. Maybe good enough is good enough? Or not. We shall see.
 
Herat - thanks, I've not taken a CDS apart... I'll have a close look at those later.

Ron - if the CDS3 uses the same muting relay arrangement as earlier players, under muting conditions or detected DC/extreme LF output, the muting relay disconnects the output to prevent plops , thumps and 'DC' being sent out. Try it if you can.
 
TheOtherOne will be getting his redone CDS3 returned tomorrow (Nov 2nd- it would have been there 3 days earlier, except it got hung up in customs), and I am sure he will provide some follow up here about the sonic benefits of beefing up the amount of bypass uFs on the output caps. For the time being we decided to leave the tants in place, instead of replacing them with some 10uF Solens. For now.
 
Has anyone who's heard one of these few modded CDS3s been able to compare it to an Avondale CD3? Or a 555?
Would be interested to hear opinions....
 
I would like to comment re: the Naim sound, in particular PRAT which seems the signature for Naim. Having in the past five years gone from a CD5i, CD5x, CDX2, CDS3 and now a modded CDS3....I have to say my findings were that the Naim sound has been extremely variable between their line of CDPs (replay being through a 552/500...though not with the CD5x).

The greatest difference in sound between CDPs I found to be between the CDX2 and the CDS3....I am sure that if a neutral observer was to do a blind test, I expect they would be excused for thinking they were players from two different companies. The CDX2/XPS2 had PRAT galore though it also had a digital edge. The CDS3/XPS2 had a totally different presentation and IMO was not big on PRAT and had a languid / laid back presentation which certainly resolved any digital edge though at times I felt it needed a 'kick up the backside'....and I am not the first to come to that conclusion.

Until my CDS3 was modded, I had never experienced what I now know to be 'grip' and 'bass slam', it is when I hear these things I know I have found replay that meets my expectations of hi fi, it is a listening experience that is both physical and cerebral and is a whole new experience and a highly rewarding one at that. There is far more immediacy detail and balance to the music. Time and time again I will replay a recording as it is so involving and pleasurable. Without reservation I will be going ahead with the extra mods described by Ron, though my predicament is letting go of my machine for the mods though I still have my CD5i as back up.

I think it important that I point out that I know nothing about the technical stuff...absolutely zilch...all I am interested in is achieving the best sound possible and feel fortunate to to be able to benefit from Ron's many year of experience in hi fi and in particular with Naim products. It is also worth noting that Ron has been a supporter of Naim and its products pretty much from the get go......or at least when it was more of a cottage industry under the helm of the late founder Julien V.

As the author of the PRAT article and technical investigation I have followed much of this thread to se the amazing things many of you get up to.

NAIM gear is costly enough without ripping it up! This is a serious persuit of the audiophile hobby .

I will not comment on the merit of modifications I have not heard but will offer a small experience for a post olive NAP250.

Like many of you I derided tants and put a costly audiophile PP cap at the 250 input .

I was certain it would get better, based on independent caps tests of my own. It did but only in part, yes , clearer, bit sweeter, more focus, more depth but it lost slam , rhythm and dynamic contrast. Too large a loss and I reverted to the original.
I then learned that not all tants are equal, the bias and running in is important and NAIM's are selected for the particular circuit purpose: the whole shebang is carefully balanced and voiced. The PRAT thing is fragile elusive and can in stages be tuned out of gear almost without you noticing it.

While not wholly audiophile perfect, run in biased tants actually are the rhythm kings in audio electronics. NO windings , little self resonance , fast transit time.

Good luck

Martin C
 
Martin makes a very good point here:
"The PRAT thing is fragile elusive and can in stages be tuned out of gear almost without you noticing it."

Worth repeating as it exactly matches my experience with biulding power supplies for Naim gear 12 years ago and more using McBride and my own opinions. It is very easy to not see the wood (musial enjoyment) for the trees (technical perfection).


regards
Jason
 
MC, thanks for the response. As I mentioned in a few of the earlier posts, the tants were not replaced, merely bypassed, which I believe allows some of their flavor to remain intact, while removing many of their weaknesses. I now have had the opportunity to side-by-side compare my hotrodded CDS3/XPS2 with:
1. Stock CDS3/Stock XPS2
2. Stock CDS3/modded XPS2
3. Modded CDS3/Stock XPS2
4. Stock CDS3/Stock 555PS
5. Modded CDS3/Modded 555PS
6. CDX/XPS1
7. CDX/modded XPS2

I have no doubt that the 'PRAT' in the modded gear not only is preserved, but also greatly enhanced. In every instance, the choke-regulated PS was a very significant step ahead of the standard one with respect to timing, clarity, articulation and preservation of harmonics. Even the humble CDX took a giant step forward when equipped with a quad-choked XPS2. The modded vs standard head unit comparisons (and I have heard three different CDS3s against mine) showed the unmodded units to be tonally, spatially and dynamically flat. Including in an all Naim system.

I would comfortably put my CDS3 against any 555 out there, and know it would handily outdistance it. Any takers??
 
MC, thanks for the response. As I mentioned in a few of the earlier posts, the tants were not replaced, merely bypassed, which I believe allows some of their flavor to remain intact, while removing many of their weaknesses. I now have had the opportunity to side-by-side compare my hotrodded CDS3/XPS2 with:
1. Stock CDS3/Stock XPS2
2. Stock CDS3/modded XPS2
3. Modded CDS3/Stock XPS2
4. Stock CDS3/Stock 555PS
5. Modded CDS3/Modded 555PS
6. CDX/XPS1
7. CDX/modded XPS2

I have no doubt that the 'PRAT' in the modded gear not only is preserved, but also greatly enhanced. In every instance, the choke-regulated PS was a very significant step ahead of the standard one with respect to timing, clarity, articulation and preservation of harmonics. Even the humble CDX took a giant step forward when equipped with a quad-choked XPS2. The modded vs standard head unit comparisons (and I have heard three different CDS3s against mine) showed the unmodded units to be tonally, spatially and dynamically flat. Including in an all Naim system.

I would comfortably put my CDS3 against any 555 out there, and know it would handily outdistance it. Any takers??

Sounds impressive but I am still a bit cautious in my response.

Aspects of the 555 I respect but I actually preferred the CDS3 in review and did not really value even the 555 supply used on a CDS3.

The 555 scored a bit higher overall but not greatly so.

Re mods, those output relays / connections were made to stop interference coupling between the connected cables and or equipment.

Ground lifting is an important technique and by deleting this aspect the carefully arranged and the total designed independence of the RCA and DIN outputs will be compromised to the potential detriment of both.

MartinC
 
I relented and went for the second round of mods sooner than later, as based on my track record with Ron, I knew there was something very special to be had and oh boy am I delighted. Ron, as usual, generously looked after the return shipping and ensured all went smoothly i.e. he ensured the DIN output was modded not the RCA output as I use a Superline. I did not send the unit down until the capacitor bank had been built. I sent it on Monday and I would have had it back by Friday if it were not for Canada customs, no surprise there. Regardless, I received my machine back within one week of shipping it. How’s that for service.

I now have more of what I liked in the original mods. Presentation is flawless, regardless of what I throw at it and I am continually being gob smacked. A good analogy for what I am experiencing is the music is presented like 'an iron fist in a velvet glove' more presence, weight, grip, detail and excellent timing and balance. It makes for a very involving and exciting listen. Beginnings and endings of tracks are so much more detailed and relevant. With each disc I play there are subtle, and at times not so subtle, discoveries and I now have a whole new CD collection to re-mine.

I feel most fourtunate to have purchased Naim equipment and on the upgrade path to have become acquainted with Ron. I took a leap of faith with Ron and it has been all positives and zero negatives. Another great job by Ron and Daryl, I am truly delighted with the end results.
 
Ground lifting is an important technique and by deleting this aspect the carefully arranged and the total designed independence of the RCA and DIN outputs will be compromised to the potential detriment of both.

If the relays had absolutely no sonic footprint, then I would have to agree. But however they do, and bypassing the ground switching relay while preserving the RCA/DIN switching relay seems to offer genuine improvements and without any loss of functionality or sonic degradement. It maybe that the benefits are larger than the sacrifices. And besides, I have a very strong feeling that the DIN and the RCA grounds are tied in to exactly the same place anyway. I havent looked inside the 552 head unit, so I am not sure if there are independant relays for both ground and signal for the input selection, but IF there are not, then why have them on the CDS3?

And of course, I did the relay mod initially before any of the other ones, and approved of the end result. Had i not, it would have quickly been reversed and not incorporated into mods done on any other CDS3 head units.

The final mod that Theotherone is talking about involves four extra 0.22uF capacitors. These currently go for almost $50US each. And there are also a quartet of 0.1uF ones (almost $30 each) and eight .0068uF ($12 each). How much would this have added onto the list price of the CDS3 headunit if these had been part of the factory build? I am guessing $2000-2500. This is probably one of the many reasons why these were never part of the original design.

BTW, the prototype version of the CDS(1) player used very, very expensive components which would have easily tripled the retail price. The production model used far more mundane components and still ended up at a retail price of 5-10 times many of the audiophile players at the time. And the prototype player is still in use and is allegedly far better than any subsequent offering. So even Naim at some point realized the benefits (but not commercial viability) of ne plus ultra componentry.
 
Hi Ron,

I've been practicing with lower power chokes on preamps and done a few calculations on your design on the XPS2.

With a 40Henry choke with a 0.25 ohm resistance and 18800uF capacitors your looking at a very high Q filter Q = 188. Normally you would want to have the Q below 0.58 here to avoid ringing. Interestingly the ringing, if you can call it that, will be at a very low frequency of about 0.18Hz. Maybe that is OK as the pulses of feed current are of much higher frequency so the voltage never gets a chance to wander.

Perhaps a bit more worrying is the effect using the choke may be having on your voltage headroom for the regulators further on. The output voltage of the choke/capacitor filter will be 63% of the voltage you had with just the capacitor filter. That may be OK but my guess is that it is too low for the regulators that follow. They normally need 3 volts or so of headroom and it would be unusual for Naim to design with much much more voltage headroom than they need here.

If I were you I would check the voltages across the regulators in your XPS2.

By the way, how much do the chokes cost?

John
 
A question on one of the above comments: why would a (tantalum) capacitor of a given rating have faster transients than a.n.other type?
 


advertisement


Back
Top