[revelinhifi]

Hello ncore owners,

Has anyone tried the Ric Shultz WIMA capacitor modification, (replacing a single stock wima cap .68mf with two .33mf polypropylene)

http://www.audiocircle.com/index.php?topic=110782.0
The post reports better sound after swapping out the stock cap's "Way more ambience, clarity and reality"

I have been unable to establish contact with Ric, no replies to emails. Therefore I ask any forum members have they any personal experience or knowledge of anyone having done this?

Absolutely no offence intended to the wonderful designer Bruno or anyone else wishing to leave their NC400 stock standard. Purely a curiosity factor for myself and the possibility of trying the sound if it seems worthwhile. I'm not a tall troubled by replacing parts at this level of difficulty, however I'm not at all willing to experiment without previous knowledge that this is a worthwhile exercise.

I'm interested in the fact that the NC1200 uses polypropylene caps

My current system is 4XNC400. Each pair of NC400 is feed by a single SMPS600. A pair of amps feeds by AMPed Speaker

Appreciate your responses

Thanks
John
 

[jtwrace]

however I'm not at all willing to experiment without previous knowledge that this is a worthwhile exercise.

Since it would not be based on any science I think you would need to try for yourself.  Subjective is just that.

[revelinhifi]

Hello jtwrace,

yes loud and clear, and I would agree that's the way to go. Although I would appreciate any comments experiences of anyone having already done the same.

Regards
John

[barrows]

"I'm interested in the fact that the NC1200 uses polypropylene caps"

I do not believe the above is correct.  I know that Ric did speculate about this in his posts, but in every picture of the NC-1200 which I have seen the caps appear to be polyester Wimas.
The problem with the mod, is that the caps do not really fit, and considering the emphasis Bruno puts on keeping loop areas compact, I really doubt changing  the caps to ones with much greater loop areas are really going to have good results. 
What I do find interesting is that in the description of the Mola Mola amps (which, apparently do not use the stock OE NC-1200, but a variant of it, with a separate board for the input stage) the company mentions a special monolithic cap which is used...
On the one hand Bruno dismisses the idea that the NC-400's stock caps are at all a compromise (see DIYaudio thread for his opinion), but on the other hand he seems to feel it is worth it to produce and use a special cap for the Mola Mola...  Hmmm.
Wish Hypex had chosen to have a custom film PP cap made for the NC-400, to me polyester seems a little "ordinary" for the job at hand.  It seems a custom 2 uF cap is in order... to replace all three of the .68 uF Wima polyesters.

[Ric Schultz]

Wima polyester and polyprop caps look identical.  The pic of the 1200 modules on the six moons review of the Atsah shows 6 caps per channel.  That would make them .33uf each.  This is the largest value you can get the polyprop caps in.  So, are they .33 polyprops or polyesters?  In any case they have steel leads are are not marked for outside foil so they are not necessarily in the right direction.  Loop area?  The modded caps are mounted one third inch above the board.   You think you can hear that?  Changing one .68 polyester to two .33 modded and properly marked Wima polyprops on the NC400 simply sounded better.  I trust my ears.

It will be interesting to see if the Mola Molas sound better than the Atsah or Merrills.  Right now I am looking at the Job 225 amp (will be listening and modding one in the next week or so).  It may be this years giant killer.   What, Ncore already outdated?  We shall see.

[Folsom]

There's a man on here that modifies WIMA's. They use to be perhaps world class, then they stopped putting copper leads on them. They haven't been the same since, is the story. He opens their ends and replaces the leads with copper. The results are probably all you'd ever ask for, no need to "swap" to another cap for the situation. But I haven't used them. I think it'd explain the change from WIMA's prominence.

[gstew]

Am I the only one that finds it humorous that the previous post references "...a man on here that modifies WIMAs" and that the post just above it is by that man?



Then, spending a few minutes looking through the easily-findable web-photos along with the datasheet and manual for the NC1200, one sees that its output filter has positions for three sets of two caps... and in most pictures (including on Hypex's info), you see both positions filled. And if you zoom the pictures in Hypex's info, you see the caps are .33uf 100v, supporting the suggestions that they are polyprops. Only the pics in the 6Moons Industry Feature announcing the NCore amps shows one cap in each position. And if they are .68 as one might suppose, then I'd agree with Ric that in this instance, they were mylar as Wima does not list a .68uf polyprop.

This plus the comments above lead me to a few questions:


1. Why is there any question about the goodness of polyprop caps versus mylar? I thought this was all settled in the early '80s with the Jung/Marsh article on 'Picking Capacitors' in the now-defunct Audio magazine and tons of related articles since.


2. I get the importance of tight loops in digital circuits (tho I am not an EE). Why is Ric's stacking of two .33 polyprops significantly worse than using two side-by-side as in the published photos of the NC1200?


3. Has anyone considered or better-yet tried some of the SMD PPS capacitors available in place of the Wima's?  Panasonic lists a .22uf 100v PPS available at Mouser. I've read varying accounts of the 'goodness' of PPS versus polyprops and polystyrenes, but they sound to be in the same class. At  7.1mm L x 6.3mm W x 4.8mm H, they could be soldered together into three stacks of about the same footprint as the existing mylar Wimas, cobbled into the circuit with some add-on leads similar to what Ric uses to modify the Wimas, and provide about the same capacitance and loop area with a better dialectric. The only question I have is outer foil directionality... I believe these are stacked foil, not wound like the standard Wimas. Do stacked foil caps have an outer foil?


4. Barrows, my thoughts looking as the components around the output filter top and bottom are that the three .68uf caps are not in parallel to make a single 2uf cap, but that they are seperate legs of a more complex filter. Have you traced the circuit and confirmed that they are used as a single 2uf-value cap?


Greg in Mississippi


P.S. Reading the NC1200 manual and datasheet was very instructive. First, the little circuit boards sticking up that I assumed were add-on input circuits turn out to be the regulators. Second, it appears to come with an input circuit (buffer) built-in, but it can be easily bypassed with soldered-jumpers... that's likely how one adds a different front-end circuit.

Pity, they don't sell this one to us DIY'ers! It'd be a lot of fun!!!

And does anyone who read the super-long NCore threads here and on DIY Audio remember if there are comments about the NC400 having high-quality regulation built-in for the input stages? If the NC400 does not, this is likely another thing that helps explain the sonic differences reported between the NC400 and the NC1200
implementations.

[waver]

if "there is no question about the goodness of polyprop caps versus mylar" i wonder what keeps Bruno from using polyprops in the first place?

[barrows]

Wima polyester and polyprop caps look identical.  The pic of the 1200 modules on the six moons review of the Atsah shows 6 caps per channel.  That would make them .33uf each.  This is the largest value you can get the polyprop caps in.  So, are they .33 polyprops or polyesters?  In any case they have steel leads are are not marked for outside foil so they are not necessarily in the right direction.  Loop area?  The modded caps are mounted one third inch above the board.   You think you can hear that?  Changing one .68 polyester to two .33 modded and properly marked Wima polyprops on the NC400 simply sounded better.  I trust my ears.

It will be interesting to see if the Mola Molas sound better than the Atsah or Merrills.  Right now I am looking at the Job 225 amp (will be listening and modding one in the next week or so).  It may be this years giant killer.   What, Ncore already outdated?  We shall see.

Ric, I suggest you do a little more research.  Polyester Wimas have silver printing on them, and the polpropylene ones have black printing.  They do not have identical appearance.  They also have the model numbers marked.  Not that I am suggesting that polyester is good!  I am with you, I would much prefer a polypropylene cap, or better yet, polystyrene in the filter circuit.  But Bruno would probably suggest: "size matters".  RE: the loop area, Bruno has stated many times the critical nature of the loop areas, in MMs, on the Ncore modules.  I do not know if I could hear the difference or not, I have not tried, but it is just as likely that you hear differences caused by increased RF pick up as it is you hear a difference related to the cap change, or lead orientation.  Until you back up your mods with measurments on an AP 2, there is no confirmation that the amps are performing better, and the subjective change could, actually, be worse performance...  I am not saying that it is, just that there is no confirmation.

[barrows]

Greg:  Right on!  I think your suggestion regarding PPS caps is worth a try.  I have used some of those in mods to cheaper DACs, replacing SMD ceramic caps used in the analog filter with the PPS films with good results, but I have never compared the sound of PPS to polypros or polystyrenes, but they are compact and could be made to fit on the board very well.
I do not know that the 3 caps on the NC-400 are acting as one, tracking the traces on the board is not so easy, but next time I have one off of the heatsink I will take a look and see what I can determine.  I am just not sure how the LPF would be "more complex" than the simple LC I assume it is?  Do you have ideas about this?  Most class D amps where I have seen the schematic just show the LC, and the output of the NC-400 being single ended seems to support this, or maybe my thinking is all wrong here?  Please let me know what you think.
There are big technical differences in the Mola Mola versus the Atsash and Merrill.  Mola Mola uses a separate input board, with a discrete input stage, and the aforementioned monolithic caps in the output filter.  Merrill and Atsah just use a stock NC-1200 module with the OE IC opamp input stage.  I would expect differences in sonics for sure.  The Audience NC-1200 based amp should also sound different, with transformer coupling and their own discrete input stage.

[barrows]

Greg:

"2. I get the importance of tight loops in digital circuits (tho I am not an EE). Why is Ric's stacking of two .33 polyprops significantly worse than using two side-by-side as in the published photos of the NC1200?"

We are talking about the NC-400 here, Ric's mod with the stacked caps doubles the lead length of that cap.  Considering how much Bruno stresses the importance of PCB layout and lead length, critical to fractions of a mm, that additional lead length bothers me.  Without doing some precise measurements it is impossible to know if it causes a problem for the module.

RE differences between NC-400 and NC-1200.  Both have the discrete regulators (based on the Hypex HxR regs) for powering the input stage.  But the NC-400 uses a discrete input circuit as well, and the NC-1200 uses an IC opamp for the input circuit.  Hence the Mola Mola amp using a separate, discrete, input circuit, and Audience also using a discrete input circuit of their own design.  Considering how accurate the Ncore output stage is, I suspect these different input circuits would be responsible for much of the difference in sonics of these amps.  Even in well designed class A/B amps, the input gain stage is often what defines much of an amplifiers sonic character.

[gstew]

if "there is no question about the goodness of polyprop caps versus mylar" i wonder what keeps Bruno from using polyprops in the first place?

Pure speculation from me is the need to have a small, good-quality cap that is available and can be easily accomodated in the manufacturing process. Modified or stacked SMD caps just don't work when you are doing pick and place. And AFAIK, there are not any other comparible quality & sized caps available in the marketplace. My guess is the Wima mylar cap worked from these perspectives and that the sound quality hit is not sufficiently large to worry about when there are SO many other things that can impact system sound quality.

Ric, I suggest you do a little more research.  Polyester Wimas have silver printing on them, and the polpropylene ones have black printing.  They do not have identical appearance.  They also have the model numbers marked.  Not that I am suggesting that polyester is good!<SNIP>

Thanks, I've seen that, but really haven't noticed it. That means that the caps shown on the NC1200 in the datasheet and manual are .33 100v Wima polyester.
Why two polyester where two polyprops could fit, I have no clue!

Greg:  Right on!  I think your suggestion regarding PPS caps is worth a try.  I have used some of those in mods to cheaper DACs, replacing SMD ceramic caps used in the analog filter with the PPS films with good results, but I have never compared the sound of PPS to polypros or polystyrenes, but they are compact and could be made to fit on the board very well.
I do not know that the 3 caps on the NC-400 are acting as one, tracking the traces on the board is not so easy, but next time I have one off of the heatsink I will take a look and see what I can determine.  I am just not sure how the LPF would be "more complex" than the simple LC I assume it is?  Do you have ideas about this?  Most class D amps where I have seen the schematic just show the LC, and the output of the NC-400 being single ended seems to support this, or maybe my thinking is all wrong here?  Please let me know what you think.
<SNIP>

I've used SMD PPS caps to replace SMD ceramics too with good effect, along with use a local PS filters on the error amps of Dexa regulators I've modified.

I've also kept watch for others using them... some pictures I saw of an Ayre DAC seemed to show a number of SMD PPS caps around the digital circuits, as do the pictures I've seen of K&K Audio's RAKK DAC. AMR touts their use of SMD PPS caps in their literature. And in the DIY world, I currently use a small ES9022-based DAC card from EUVL on DIYAudio (that actually sounds much better than one would expect). One thing he does is to solder filtering caps right on the legs of the ES9022 chip. In his latest version, these are now SMD PPS.

On the output filter of the NC400, I had in mind diagrams I saw of the output filters used in B&O Icepower amps which show an LC and a Zobel. Download the datasheets for the 1000asp and 200asc to see what I  mean. Also, I was looking at a picture from SerengetiPlains' NCore mods thread awhile back. He posted a number of closeups of the backside of the NC400 board and one shows the components around the filter caps, of which there are a few R's and C's. I've posted it here again for reference. But I have not traced it and cannot say what the filter is.

<SNIP>
We are talking about the NC-400 here, Ric's mod with the stacked caps doubles the lead length of that cap.  Considering how much Bruno stresses the importance of PCB layout and lead length, critical to fractions of a mm, that additional lead length bothers me.  Without doing some precise measurements it is impossible to know if it causes a problem for the module.

RE differences between NC-400 and NC-1200.  Both have the discrete regulators (based on the Hypex HxR regs) for powering the input stage.  But the NC-400 uses a discrete input circuit as well, and the NC-1200 uses an IC opamp for the input circuit.  Hence the Mola Mola amp using a separate, discrete, input circuit, and Audience also using a discrete input circuit of their own design.  Considering how accurate the Ncore output stage is, I suspect these different input circuits would be responsible for much of the difference in sonics of these amps.  Even in well designed class A/B amps, the input gain stage is often what defines much of an amplifiers sonic character.

I too am curious if it causes problems and would love to see those tests, if they were ever done. OTOH, I have a lot of faith in Ric's abilities to identify and develop ways to build & modify EQ for better sound focusing on areas that seem un-important or counter-important to others, but in practice make significant differences. I have several pieces of gear build or modified by him and have never been disappointed... and when I've reverse-engineered his techniques and applied them to my own DIY efforts, I've also been happier than expected.

Good to know on the regulators on the NC400... I had a vague recollection they were there, but as you've noted in the past, it is hard to get information out of all of the data in the monster DIYAudio NCore thread.

And on the differences between the various incarnations of the NC1200, I really haven't paid much attention here as I'm never expecting to be buying one, given what they charge. I'm just happy I've gotten to the point of really liking the NC400's in my setup.

Greg in Mississippi

[revelinhifi]

Thanks to those that responded to the question I asked. 

I will be modifying a pair of NC400 amps with a stacked pair of polypropylene .33mf =.66mf
I will be replacing the three polyester .68mf capacitors for a more life-like tone to the Ear

My current setup with the NC400
The four NC400 are located in a single aluminium container. Each pier of NC400’s is powered by a single SMPS600. This is bi-AMP arrangement to power a single speaker.

This is bit perfect setup for me to prove this mod stock polyester to polypropylene, test instrument the Ear (physical cap surface area after change twice the area though same PCB board placement of stock cap) emi I will listen out for it

-Components-

PS Audio power plant 
PS Audio MK11 DAC with Ethernet Bridge
Software – Player - Jriver - driver bit perfect PS Audio WaveStream out on windows 8 CPU 3770k and 16 gigs ram CPU shows <1% utilisation streaming 24\96 bit album

Amplifier 4 X  NC400 bi-wire pair power speaker
Cables - silver
Cables - balanced
Speaker- Theophany Rhapsody

Thanks  

[revelinhifi]

If anyone is interested

Two of the four NC400 have been modified swapping out the .68mf polyester cap for .33mf staked polypropylene. One of the six caps pulled had steel leads.

They have not fully burned in yet, only two hours operation, from first switch on it was plainly evident the mod was successful sounding better to my ears.

I have lived with the 4 NC400 for about a year and very familiar with the tone. I wont go into a lot of detail other than the sound is greatly improved and well worth the effort of the mod. Big thanks to Ric Schultz at tweakaudio for the modified WIMA Caps.

The sound improvement was one I can reference to going from the mk1 digital to analogue board to the mk2 digital to analogue board in my PS Audio DAC those having done this will know exactly what I mean. In essence the NC400 is now more detailed and a greater sense of that's how that instrument sounds.

I will update as the caps burn in
First hour - power up and after some listening it was apparent to me that the mod sounded better than stock -
24 - hours in the sound is more open, WOW love the sound of the guitar and violin, there is an improved sound to the bass and vocal is more articulate
48 - many hours of listening now and these are my last words. The Polyester caps don't do the sound justice for the NC400. The polypropylene caps have resulted in a more articulate musical performance, the ambience is easy to connect with and I find myself tapping to the rhythm of the music. I do find it incredible there is more music to hear with this caps swap-out. If I ever get a chance I will get the amp measured to see if the specs have changed. My concerns the caps would have attracted noise was not to be the case the. the highs are significantly more detailed, vocals more etched and lifelike the bass has good weight but sounds more defined. 

Changing out the .68mf polyester caps for .33mf stacked polypropylene has a sound difference way more significant than changing cables in your system.
 
I have not finished the job, this was initially to check the mod was good. Well it passed without any doubt.  I plan to mod the remaining two, at that time I will rotate each NC400 180 degrees shorten the wiring significantly. Another idea I will take the power loom and put this to the outside beneath the case and pop it through the case at each point connecting to the NC400 and SMPS600 respectively. I will shield the cable on the outside beneath the case. I will partition off the PSU from the NC400's as well

You may see I removed the Speaker post and replaced those with Silver wire contacting all four point beneath the PCB

If you have an NC400 and have not done this mod there are sound on your music files you are not hearing. I am familiar with my music and I am hearing musical sound not previously heard.     

You could say what a horrible job with the wiring and that it is, therefore it should be quite the antenna and radiator, hmm I'm not getting any audible noise but beautiful music.

I recommend the mod over the stock unit, however there is a little skill and patience. I wonder if the wiring tidy up will benefit the musical performance. The high frequency I like very much, cant wait for my other caps to arrive. Big thanks to the other forum member who shared his experience and knowledge - go Mississippi

waiting now for my other caps to arrive so I can do the final pair and go back to a bi-amped configuration

Cheers
John



     

[revelinhifi]

The last six pairs of caps have arrived they have been installed into the NC400. An initial listening session was completed to confirm the installation was successful in which it was.

I have not done all my wiring reconfigurations as mentioned in previous posts here, but the capacitors are Installed and the Speaker Posts Have Been Replaced for Direct Silver Wired wires.

The SQ of these amplifiers are significantly improved by these two modifications. In a nutshell the highs are significantly more extended the mid range is improved and the base has tighter control. I cannot detect any elements to the music heard which is not improved over the stock unit so I'm very positive about this modification that the sound produced by the NC400 is better than the stock unit.  If anyone has any interest, I will be happy to provide further feedback. Drop me a message and I will reply as soon as able.

I will post new pictures sometime in the near future. This will show four NC400 modified.

Regards John

[revelinhifi]

All done now here is the 4 NC400 with the modified caps and the connecting power loom outside the case. At a latter time I may put in a wall inside the case and still to fashion a cover beneath the case to contain the power connecter cable. 

Oh yes, the SQ is way better than the stock unit, I'm confident to announce that fact here, there is no negatives heard the sound is better in all respect compared to the stock unit..



 

[jtwrace]

How they measure with the new caps?

[Julf]

How they measure with the new caps?

And, just out of curiosity, have you asked Bruno about his opinion on the mod?

[revelinhifi]

Jtrace said: How they measure with the new caps? This is a good question I used the human EAR for the test and don’t have recorded results - just the knowledge and experience.

The un-modified unit was compared to the modified unit before completing all four. All I had to do was swap speakers between the amps.
No deficiency heard compared to the stock caps. The improvement is not minor but significant audible improvement.
Julf: No I have not asked Bruno; to ask that question you would want to be supported by measurements of the modified amp. And I did not set out to improve measurements, how can you if the stock unit measures near perfect.
The available space on the NC400 does not allow for polypropylene X 6 CAPs and only DIY can implement as I’ve done. The manufacturing process does not allow this type of implementation, it is non-standard and restricted to space, only accommodating the polyester caps.
I have implemented same capacitance, they should be very close in measurements - just different capacitor material and slightly bigger physical foot print.  Consider these facts below

1. All caps foil side facing ground
2. No steel lead caps

Power connector routed outside the case minimising RF and cause induction to sensitive wiring and components.
I can comment as I have done, because I have lived with my NC400 for over a year and know the sound they produce and compared the stock unit to the modified unit. I have taken time doing this and can say the stock unit is significantly down in SQ in comparison, nothing lost only gained. There is significant untapped performance in the NC400 for the human Ear to hear.

Regards
John

[cab]

We don't really know if this is an improvement. All we know is that you prefer it. It is completely subjective.