X-24K DAC Mods

[Deleted]

Matt
I am glad it all came together.

The 3 most knowledgeable members regarding the technical aspects of DACs, appear to be Allan,Leo and Rick. Although there may be others just as knowledgeable in this area.

Alex

[dejanm]

Matt,

I would have a question about tantallium caps. You said that you replaced them in PSU. Did you replace them also in the output stage where they are used as filters for OPA power supply ? Did you get any improvement there ? Which caps you used instead of tants ? Did you use the same values as for tantals or did you go for larger ones ?

[matt7941]

Good afternoon dejanm.

After reading various threads (on other forums also) I did replace all the tantalum caps, both in the power supply and output. On the lower power supply board I replaced them with equivalent Elna Silmics, on the upper digital board I used Elna Cerafines and in the output stage the choice was Rubycon ZLs. This did seem to alter the sound for the better, especially Cerafines and ZLs.

The tantalums were originally replaced with some of the Panasonic AM series in the output stage and Oscons elsewhere. However, whilst sounding smoother and less harsh than before they lost some of the top end sparkle and sounded a little closed in.

I changed the Panasonics in the output stage for the Rubycon ZLs and the difference was immediate. Some of the top end sparkle had returned but without any of the metallic tantalum harshness, the soundstage was also improved. I then replaced the Oscons with the Cerafines and this really opened things up. More sweetness, more open, more detail and definitely wider soundstage. It was almost as if the Oscons had somehow limited the digital supply section (at least in this application).

All the values have remained the same wherever possible. I really do like the Rubycon ZLs, I know some people are using these instead of the hard to get Blackgates. The Cerafines also bring what appears to be a quite neutral sound to the presentation, very satisfying.

All told, I'd agree with Leo's and Alex's opinions on the tantalums, avoid wherever possible especially in large numbers where they do appear to dominate the sound.

Hope this helps.

Cheers

Matt

[dejanm]

Matt,

Thanks for this ...

[matt7941]

For anyone who is interested I have finished tinkering with the X-24K and can tell you it is a vast improvement over the original "stock" sound.

The latest bits to be swapped around have been the opamps. They were opa2604s followed by LM4562s. However, I have recently had a little time spare and spent a day or so rolling various combinations and settled on the LM4562s moved to the first position with op275gps on the output. The op275gps were quite a surprise, I just took a gamble on them as my local supplier had just got them in and they were cheap. The combination (Alex, I'll qualify this by stating "in this application at least") has all the power, drive, detail and subtlety of the LM4562s with the wide open soundstage of the op275gps and a slightly more organic sound to boot (again Alex your theory of combining opamps from different manufacturers works a treat, I did try op275gps at all positions but things became muddy and lacking dynamics).

I have also removed the original crystal and associated Philips chip and replaced it with a high accuracy TCXO module with it's own regulated power supply available here:

cgi.ebay.co.uk/Low-Jitter-clock-to-upgrade-Cd-player-Usb-to-spdif-dac_W0QQitemZ220282761555QQcmdZViewItemQQptZOther_MP3_Players?hash=item220282761555

As per my previous post I was unsure which to undertake first, the DAC or the transport. I'm glad I did the DAC first as it is a very noticeable improvement. The soundstage is immediately more immersive and separation is noticeably improved. The leading edges of notes are razor sharp and decay of cymbals and the like is soooo good. I was a little unsure whether the benefits would be worthwhile but they certainly are.

For an outlay of just over £30 (incl postage) it's a steal. The bonus being the unit fits perfectly inside the DAC which is amazing considering the lack of space in the X-24K. It arrived within a week and appears to be all genuine components. The board is well made and even has a suspended section for the clock to try to reduce vibration.

I'll post a couple of pics over the next few days when I get a moment so anyone who's interested can compare the final tweaked version to the original as posted by canjunkie some time ago.

I do have three op275gps spare if anyone would like them to try. I was quite impressed with their sound. If you drop me a PM with your address you can have them free, but I do only have three.

Cheers

Matt

[insomniac]

Hi Matt

Sounds like your X DAC has had a fair bit of surgery performed

Glad to hear the Clock replacement has made a nice improvement. What size is that Clock PCB you fitted, it looks quite small?

I will have a batch of Low-jitter Clocks available soon in five popular clock frequencies: 8.46MHZ /11.28MHZ /16.93MHZ /33.86MHZ and 27.00MHZ for DVD/Multi players. PM me if you would like one sent over for comparison.



Low jitter Clock circuit with Parallel Stabilized Voltage Supply.
Power Requirement: 10 -30 DC... Power Draw: 60mA

Quality Components:
DALE Resistors, NoVer Capacitors, EVOX capacitors,
Extra thick gold plated PCB, Silver solder
Lower Error: 1.5 ps
PCB Size (70x45mm)

Price £20

[Deleted]

Wot ! No 24.576 ?

[insomniac]

Feb 24, 2009 1:21:39 GMT @sandyk said:

Wot ! No 24.576 ?

24.576 probably available as a special order

[Deleted]

insomniac
The MF X-DAC V3 already uses an oscillator with reasonable specs. I doubt that there would be much benefit over that ,combined with a JLH PSU feeding it. For people like XTRProf, it could be worthwhile.
Alex

[insomniac]

Alex
I haven't seen the X-DAC V3 oscillator but the X-24K uses an el-cheapo 2 pin one IIRC. Maybe that element of doubt would be dispelled if one of these clock's mysteriously found itself on your doorstep one morning

I think the Prof would agree...as we know he is Pro-shielding

[Deleted]

Insomniac
The X-DAC V3 uses a 5V DC TTL/HCMOS Xtal Oscillator with a stability of +-25ppm (MC-1500A 24.576MHZ ), so when supplied via a JLH, it has quite good stability.
SandyK

[matt7941]

Good afternoon all.

Insomniac, the new clock module measures just just 58mm x 58mm across the PCB.

Hopefully, the below link will take you to my photos I've just taken. Let me know if there's any problems with the link. The only bits you can't just about see are the full recap of the ceramics on the bottom layer with WIMA or Evox-Rifa polyesters, the opamps hiding under the clock board on their little sockets with associated full recap with polyprops where applicable and the bypasses for the opamps under the main board.

















I've also included a couple of shots of the power supply I've cobbled together based on Mike's DIY PSU for the original X-series components. The enclosure itself is allegedly EMI/RFI shielded with a special coating.

One last question, probably more for Alex I think. Would it be that worthwhile replacing the 7815 and 7915 regulators that supply the analogue section including the opamps with some better regulators?

Cheers

Matt

[insomniac]

Insomniac, the new clock module measures just just 58mm x 58mm across the PCB.Hopefully, the below link will take you to my photos I've just taken. Let me know if there's any problems with the link. The only bits you can't just about see are the full recap of the ceramics on the bottom layer with WIMA or Evox-Rifa polyesters, the opamps hiding under the clock board on their little sockets with associated full recap with polyprops where applicable and the bypasses for the opamps under the main board.

Cool...similar surface area to my clocks but square rather than rectangle. It looks like it was made to fit in there. I find it sometimes difficult to work out the size of something from a photo, until I see it in relation to a known object.

Your X-24 is looking well in those photos, better crack the lid back on and do some listening

[matt7941]

Hello all,

Just a quick question.

I'm using a 30va toroidal transformer in the X-24K's power supply. Initially this was using both secondaries in parallel to provide a 2.5A supply at 12v. (Similar to Mike's Little Pinkie). However, as an interim measure I have split the secondaries as one now feeds the clock module I installed, thus both outputting 1.25A.

However, I have noticed the bass response, and dynamics, has dipped slightly, very similar, but obviously inverse, to the response I encountered when I replaced the stock wall wart with the 30va. I thus assume the X-24K prefers a "meatier" supply.

The question is would the X-24K stand the use of an 80va? With each secondary supplying 3.33A would this "fry" either the clock module or the X-24K itself? If I used a 50va each secondary would supply 2.08A and thus be an increase on the 1.25A it is being supplied with at the moment. But would this be enough to make the X-24K sing again with respect to the bottom end?

The other option is of course to feed the clock with a mini-toroidal of approx 7VA and feed it 600ma and once again couple the secondaries of the 30va in parallel to provide the X-24K with the 2.5A. However, if I was to replace the 30va I'd be taking the opportunity to use an encapsulated toroid as its replacement to try further to eliminate hum etc.

Any thoughts from the members would be most appreciated.

Cheers

Matt

[Deleted]

The question is would the X-24K stand the use of an 80va? With each secondary supplying 3.33A would this "fry" either the clock module or the X-24K itself?

matt7941
It comes down to the quality of the transformer. A good quality transformer's output voltage shouldn't be much above it's designated output when lightly loaded. (regulation)
This means that with a transformer like the 50VA, the output voltage is likely to be only a smidgin higher than with a 30VA that isn't overloaded. The circuit will draw similar current (perhaps a tiny bit higher with the larger model) , irrespective of the VA rating of the transformer . Why not try the 50VA model which gives only a modest increase in output capacity ?

SandyK

[matt7941]

Cheers Alex, I'll give the 50VA a try. Until it arrives (I've just ordered an encapsulated version) I've paralleled the 30va secondaries again and I'm using the original wall wart from the X-24K to power the clock module. Voila, the bass response, dynamics and separation are back. It really is amazing what some extra oomph in the power department can do. The only thing being it means I have to unplug components if I'm using the DAC, I'm a little short on sockets thus by using a larger transformer, or indeed two with the idea of the separate mini-toroidal, I wouldn't be using any extra plugs than before.

Thanks again

Matt

[dejanm]

Hi,

What is the chip set inside 24K ? i mean: which receiver, digital filter, dac chip ... ?

What are the plower supply requirements ? Can it be driven with MF X-PSU ?

[matt7941]

Evening all.

Just a quick update to the X-24K project as it's finally finished. Following on from my last post, the 30va power supply is powering the DAC main boards and I'm using the original wallwart to supply the clock module as it has its own on board regulation. This has worked out very well.

Alex, you were so right in what you said about power supplies being key. I've just replaced the 7815/7915 which supply the analogue stages and what a MASSIVE difference. I'm using the Audiocom Super-regulator 2's and have heavily modified the case with drilled holes for ventilation and the heatsinks for the regulators on the top of the case. The DAC looks as though it's developed a mohican! But what a difference in every single aspect of the sound, its now so far removed from the original stock sound it is unbelievable. If anyone is thinking of modding any of the MF DACs I'd have to suggest using better regulators or the JLHs if possible has got to be the first port of call. You were also right about the quality of the LM4562s and that they do appreciate clean power, it really makes them sing. Everything is so organic, detailed and totally natural. I've definitely finished modding the DAC now as I don't want to risk spoiling anything. I'm now wondering if I can do the same in my A3 amp to supply the quad version of the LM4562 and see if it has the same effect. Now that would be something!

I'll post some pictures when I get a minute or two, I'm just a touch busy at the moment, as pictures seem to go down quite well on here.

Cheers, Matt.

[matt7941]

Finally finished the DAC and got the camera out to take the final shots. The new regulators are bolted directly beneathe the heatsinks on the top and I've had to ventilate the case a little too. I've also added a captive digital lead with 75ohm BNC plug. The new regulators run quite warm, much more so than the original 7815/7915. I think it looks kinda mean, but then again I would









Thanks again Alex, Leo and everyone else for advising on this project.

Cheers

Matt

[Deleted]

May 22, 2009 22:57:31 GMT matt7941 said:

Finally finished the DAC and got the camera out to take the final shots. The new regulators are bolted directly beneathe the heatsinks on the top and I've had to ventilate the case a little too. I've also added a captive digital lead with 75ohm BNC plug. The new regulators run quite warm, much more so than the original 7815/7915. I think it looks kinda mean, but then again I would









Thanks again Alex, Leo and everyone else for advising on this project.

Cheers

Matt

Matt
Yes, it does look kinda mean!
What were the major areas of performance gains?
Alex

[clausdk]

It is indeed mean looking.
At the beginning my first thougth was that it should be painted with some black spray on the top, where you drilled the holes, but the longer I look the less I think so..

[matt7941]

Good morning Alex. I'm not sure whether you're referring to gains from the entire project or from the voltage regualators so I'll address both, starting with the latter in isolation.

The new voltage regulators have just completely opened up the sound. I thought the previous tweaks had made a difference but these were frankly amazing. The bottom end became much tighter, the air and space around instruments increased and the dynamics were taken to another level. It dredges up more detail than ever now, but the biggest improvement is the overall "cohesion" to the sound. I used to listen to individual parts of the music in isolation (if that makes sense), but now everything is where it should be and nothing is forced making it so easy to sit back and just listen to the bigger picture. Everything is just played out in front of me in the most natural and satisfying fashion. It did take quite a bit of messing around with various heatsinks and a Dremel-type mini drill to make the holes for the regulators in the top of the case (by which time I was grey from being covered in aluminium dust) but it was certainly worthwhile. At £34 for two they weren't cheap, but looking around the internet at regulators from other sources they were by comparison, but worth every penny. Every area of performance was improved by fitting these. As I alluded to in my previous post it would, I presume, be a similar effect to that being realised by members by the use of the JLH modules. I would suggest the LM4562s I have in the filter/buffer stage of the output, as you stated initially, love the clean power they provide. They are really singing now!

Overall, the sound is just completely removed from that of the stock X-24K. Initially, after "downgrading" from a Cyrus CD7Q and PSX-R, the sound was lightweight and veiled. After all the tweaks the sound is open, detailed, deep, melodic, natural and sweet. Dynamics are superb and timing is excellent, giving any music a rhythmic, fluid quality. Detail retrieval is stupendous, the amount of little touches in the production I've picked up on which I never knew were there is amazing. In EVERY department the X-24K now betters what I previously had. The captive digital lead with the 75ohm BNC doesn't, in truth, add that much. If anyone else has one of these I'd stick to the standard arrangement with regard to the digital inputs. I am so happy with the sound I will be putting the soldering iron down for a while and just enjoying the sound for what I hope will be years to come. With a better transport, possibly one of the TEAC VRDS equipped machines, I suspect there may well be more improvements to be heard.

If I were to do this over again I'd do the following in this order:

1) Recap the horrible large Jamicons in the power supply side with Panasonic FMs and their smaller accompanying Jamicons with Elna Silmics (10uf 16v if I remember correctly). Big improvement with this alone.

2) Replace the 7815 and 7915 regulators with something better. The Audiocom Super Regulator 2s I used are excellent but do need heatsinking and thus a bit of surgery to the casework. This is pobably the single biggest, if most awkward to achieve, upgrade you can do to this little beastie. Every area of the sound is improved and lifts the performance into another arena entirely. I presume the JLH modules would do a splendid job here but a very clever arrangement would be necessary to house them, as space is at a premium.

3) Replace the horrible NP output caps with something better. Mine were replaced with the 220uf 16v Nichicon Muse ES BP caps and bypasssed with 1uf Mudorf Mcaps. Sweet. Bass improves and so does air, space and naturalness in and around instruments. Like lifting a veil. Simple and worthwhile.

4) Change the opamps (after fitting sockets) to LM4562s at the first position and OP275GPs at the output. I know people will criticise the OP275s but in this application they are excellent, adding depth and breadth to the soundstage which just improves the sound eminating from the LM4562s. With the sockets however you are free to experiment.

Any other tweaks really are a matter of try for yourself, as the individual improvements aren't massive but the cumulative effect is worthwhile.

The only other definite tweak is a better power supply, build one yourself or get one of Mike's Little Pinkies which will do the job nicely. If you don't you won't realise the improvements from the above tweaks, or at least maximise their effects.

Cheers Claus, I was thinking of painting the sides of the holes where I've drilled for ventilation but, like you, thought it would be better to leave them. It highlights them and actually makes them a feature. If viewed from the front it almost looks like the barrel of an old fashioned machine gun. The bonus is the red leds on the regulators give off a subtle light which reflects off the bare aluminium at night, it almost makes it look like its alive. Cool.

BTW, Alex would you suggest that I could use the JLHs or similar in the A3 amp? It too has its quad opamp supplied by 78xx/79xx regulators so I presume would be happier with something better. Do better regulators give the same improvements in amps as they do in source components? If I could get equally good results from this I'd put the soldering iron away permanently and be content for the rest of my days to listen to the sytem I've got now. Probably (!).

Cheers,

Matt.

[insomniac]

If viewed from the front it almost looks like the barrel of an old fashioned machine gun.

I thought of the Sterling SMG when I first saw it:



Cool modifications

[matt7941]

Cheers Insomniac, that's exactly what I was thinking of!

Question: did you just happen to have that lying around? Are there any banks safe from you in Ireland?

[robertkd]

May 23, 2009 11:51:22 GMT insomniac said:

If viewed from the front it almost looks like the barrel of an old fashioned machine gun.

I thought of the Sterling SMG when I first saw it:



Cool modifications

and that looks like an F2, 9mm open breach firing SMG