Musical Fidelity nuVista M3 Mods/Upgrades
Hey everyone,
So, I've been horsing around/having fun with "upgrading"/"rebuilding" my Musical Fidelity nu-Vista m3 Integrated amp. Figured some of you here might be interested in seeing the guts. I know some of what follows it going to be "controversial," in terms of whether or not any audible difference can or is perceived/heard. Believe me, I'm an E. Eng, and I am skeptical myself! Anyhow...it is what it is. Take it with a grain of salt, or feel free to object (within forum policies/behavioural expectations of course)! I'm open to opinions/suggestions as I go along.
A little background, this is a dual-mono amp with integrated preamp. The three toroidals--2 x for each amp channel, 1 x for the preamp--are actually contained in an outboard unit making it a "not quite so integrated," integrated amp, but anyways... Only 500 units were built because of the "supposed rarity" of the 4 x RCA 6WC4 nuVistors utilized in the pre-amp section. (Actually, I'm not sure if the sockets were the claimed rare thing, or the nuVistors themselves!) Here is an online link to LiQUiD AUDiO who has some nice photos of what it looks like stock. (I didn't really take stock photos...more component photos for my own reference.) I picked this one up used for what I consider, a "really good price" considering shipping the darn thing (80+ lbs for both pieces) must have cost the seller a lot!
Round 1 Problems
Anyways, left channel had some "intermittent issues" that turned out to be the output-muting relay and not the ALPS volume potentiometer that I first suspected. (Seller did disclose this was an issue and also thought it was the volume pot.) I tried to clean, then replace it, to no avail. Then I noticed that when I hit mute (which I never really used) it would sometimes "fix" the "dead" left channel which led me to look at the Omron G2R relay. Sure enough, with the top off, I noticed the contact pads on the Omron relay on the left side channel looked crooked--the centre part of the round pad was not making good contact.
These Omron relays seem to be decent parts actually. And it would appear that they are supposed to be self-cleaning. Every time they engage (well, when they are centred nicely and "flip" to the NO or NC side), they don't move perpendicular to the contact surfaces, but actually rub slightly, parallel to the contact pads, which themselves are ever so slightly convex. This "micro-rubbing" is supposed to clean away any tiny amounts of oxidation every time it makes contact.
Of course, this doesn't happen optimally if the contact nub is crooked, resulting in only a tiny portion of the pad makes contact and due to arcing, ends up "oxidizing quickly." So instead of spending $10 to get a new relay (which I probably should have done) I took it apart, cleaned it all with Deoxit D100L, recoated with G100L, bent the stalk so it is better centred and tada...no more intermittent channel.
The Mod Bug Bites
Thanks to @Krunok and @balbuze and others who developed DRC for volumio and gave pointers to n00bs like me, the first mod bug was to implement DRC. I am running an rPi 3B+ with Allo Katana board into the nuVista m3. Of course, the DRC made a huge difference, but then I started to want to poke around in the amp, not because I wanted a "better" or "different" sound, but I suppose, because of curiosity and for "fun."
So, knowing the amp was 17-18 years old, I obtained a copy of the schematics from the Canadian MF dealer which turns out to be Focal (really nice guys that actually offered a lot of suggestions/help!) and noted all the electrolytics in use. Then, (since I had the top off the amp a few times by then) I figured, why not change the yellow chicklet film caps to "something better" as well. I'm guessing the (L) labelled film caps were probably nothing to write home about given their use of Jamicon electrolytics. (For those of you already crying foul about capacitors not affecting the sound sure, maybe correct, but I was re-capping because I noticed some of the electrolytics were not "flat" on the top anymore...i.e. they were very, very slightly bulging. Were they in any danger of blowing up soon? Probably not, but hey, this is a hobby.)
After a few big orders through Mouser, I had in my hands a pile of Elna Silmic II's, WIMA MKP10's and Nichicon Electrolytics mainly for the power supply filtering as their were lots of size constraints there. I decided to start with the PSU and mod component by component, with listening (actually more testing to make sure I didn't break anything) in between. I still have not gotten to the Power Amp modules as I am waiting for a few final components still before I tackle that.
The Outboard PSU
Here's the "stock" PSU courtesy of a pic online since I guess I forgot to take a photo when I opened it up.
PSU.jpg
Here's the "more or less" final result of my tinkering.
Redone PSU.jpg
Like the braiding? I know, tightly twisting the wires would have resulted in better field cancellation, blah blah blah, IF I was actually worried about that. The braiding was just so fun and it made it look so pretty though! Also, it is hard to tell from the overhead shots, but I've actually tried to bend the braided cables in 3D space so that they are "maximally perpendicular" to other wires where they are "running parallel" in a 2D slice. I was going to put those PET cable braid sheathing on it but decided not to cause of the pretty braids.
You'll also see that I replaced the IEC input connector. MF actually put in an inductor for filtering purposes which is nice and probably already a step up from many other manufacturers. But why not drop in a nice filtered Schurter power entry module since it was open anyways right? I totally acknowledge this might have been thoroughly excessive, but I would imagine the Schurter is far better than the single inductor component. Some audio-snobs might claim that it makes the sound "compressed," but I haven't noticed any negative effects thus far. I also replaced the Teapo electrolytic cap in the little green PCB with an Elna Silmic II. Why? "Just cause." I was ordering a tonne of caps anyways and it was only an "extra dollar."
Oh, I also replaced the 2x Powercon panel connectors and the 1x5 pin XLR as well. Now, what "audio effect," did this all have? Well the capacitor that was replaced on the little green PCB is filtering for a) power LED light, b) 9VAC that gets converted to 12VDC to heat the nuVistors and 5VDC for the control ICs (i.e. motor/remote controls) which aren't in the audio path anyways. So, that cap probably had next to no audible effect on the sound. I suppose it is possible that the cleaner 12VDC supply that heat the nuVistors has some audible effect...but *shrug.*
What did noticeably change was the noise floor! I was rather surprised about this. You know, the faint faint "hsssssss" that comes from the speakers when amp is turned up with no input signal. I could turn the volume to about 50% before mods, before I could hear the soft "hsssss" coming from the speakers. With the PSU changes, the knob has to be about 75%.
Last change related to PSU thus far is the new pre-amp umbilical cable (the 5-pin XLR). After taking apart the OEM one I noticed the shield is not even connected! Internally in the PSU, there is a cap and resistor between the 0V line and shield of the XLR connector--chassis ground and "signal ground" basically. The new cable I made out of "tow-line cable" which is nice and flexible. It has 5x19AWG and a foil + braided shield with the shield connected on the PSU side only.
Audible difference? Noise floor dropped again. I can now turn up the volume to 100% essentially and still not notice any "hsssssss" coming from my speakers now. All this without having yet touched the amplifier umbilicals (the ones to the powercons).
The Pre-Amp Section
Here's a "stock" photo of the pre-amp, again courtesy of a photo online since the "before" photos I took were really more focused on particular components.
img_6811.jpg
And here is the more or less final result of the pre-amp changes
Redone PSU.jpg
Pre-Amp Power Supply board (the floating top board) has been recapped with Nichicons, KG Super-Throughs(!) for the section that takes 9VAC to 12VDC, and 5VDC. Excessive? Yes, most probably, especially since none of it is really in the "audio path." Remember, the 12V is for heating the nuVistors and the 5V is for the IC's relating to remote control functions. I originally was going to throw some Elna Silmics II in there too but, they were too tall. :-( The Elna's definitely cost less than the Nichicons!
The larger 1000uF, 100V are also Nichicons but are KS series caps simply due to clearance issues. Pretty much, they were the only "audio grade" caps I could find that were not too tall, nor too wide. Though I don't buy into the hype around "audio-grade" caps, in this case, Nichicon's size + specs were better than equivalent parts that might have fit. And they were cheaper too! There are a few smaller filter caps that were replaced with Elna Silmics II, and the yellow film caps were replaced with WIMA MKP10 equivalents. This section could theoretically have some impact on sound quality as this supplies the pre-amp with the +90V and -90V rails. How much? Well, I didn't test it extensively on its own before I changed the caps in the pre-amp board below.
Onto the pre-amp itself (the larger, lower board). Same idea here. All electrolytics were replaced with Elna Silmics II and all films swapped out for WIMA MKP10s. This was quite the challenge as the lead spacing supplied on those caps are actually too wide for the holes on the PCB (at least for the bigger values). I had to carefully bend those and then when soldering, ensure the little nub of a lead that barely poked through, was fully attached to "through pad" part of the solder pad. It's definitely not "ideal" but it is making "good contact" at least for now! There are also some output coupling caps, 10uF, 63V that are bipolar electrolytics which can't be seen in the photos. I wanted to put in some film caps (i.e. the red WIMAs) but they are WAY too massive. I've put in Elna BP capacitors for now. I also have some Nichicon Muse ES capacitors that I may try out later.
I've also replaced the short RCA interconnects with some silver-plated MPS M-880 Coaxial Cable to which I added a translucent/gold-wrapped, PET braided sleeves. Connectors are CMC-1036-WF, (they are pure copper base metal with gold plating on top, RCA plugs) which are slightly too large for the cable, but with the braid, it fits "nicely enough." I am thinking I should have gone with the CMC-8036-WF instead. In any case, they hold well enough (soldered) that I chose not to use the little locking screw that normally crushes the coaxial cable. No point in messing up the characteristic impedance of the cable if one doesn't have to. Of course, the cable is 6" to 8" long so...yeah, probably makes zero difference too!
Further planned upgrades for this section include: a) replacing three of the rear panel RCA's, directly wiring one end of the RCA interconnects to the back panel pre-amp out RCA's, b) replacing the ribbon cable with mini-75 ohm coaxial cable, but that seems to be major pain so we'll see, and c) possibly re-capping the phono section, but I don't use it anyways so, maybe not.
OK, subjective listening differences now. Feel free to skip this part if you feel it is too "controversial." I've definitely noticed the sound-stage has been expanded in width and depth. Along with this, it is far more "holographic," and by that I mean, you can definitely feel like you can "walk-around" the instruments playing in an orchestral recording. It "seems" like there is a little more "sparkle" in the high-end--not fatiguing brightness. There's a bump clarity/definition throughout the range, but definitely noticeable in the highs and on metallic percussive instruments like cymbals or chimes. Perhaps the best way I can describe it is that in a great recording, you can pick out now when a drummer slightly drags his sticks across the top of a hi-hat or cymbals that have concentric machining marks on it (like the cheap ones our school used to have in the band room).
Mid-bass may be a touch reduced, but I think that is because clarity has improved/muddiness removed. Low-bass seems to have also benefitted, not just "generically increasing" but it is far easier to pick out the specific notes in recordings now.
The Power-Amp
You'll see from the previous photo, I haven't gotten to this section yet. But, basically, all the yellow film caps will disappear, replaced with WIMA MKP10. The various coupling/bypass capacitors are also all slated to be replaced with Elna Silmic II, with the few bipolar caps, to be replaced with Elna BP.
As for the giant 6800uF, 100V Jamicon caps, they are going to be replaced with Nichicon LLS capacitors. The stock capacitors are 35mmx60mm so it is quite "tight." Ideally, I would replace them with some "industrial" capacitors rated at 8000 or 9000 hours of use at 105C, but those won't fit. The 8000-9000 hour 85C caps from Cornell Dubilier are 40mmx63mm which may fit, but are $25/piece and I would need 8 of them. You might be wondering why I would prefer (but am "too cheap") to use the higher rated hour and temperature caps. This is simply because it is way harder to get to the Power Amp module. Ideally, I'd like to replace them once and not have to worry about them for a LONG time. That said, I did use a FLIR camera to check the temperatures around the capacitors there (hot = bad for lifespan!) and they actually run quite cool. If I remember correctly for electronic components, typically for every 10C drop in operating temperature, lifespan doubles. So even the Nichicon -LLS which are 3000 hours at 85C, if they operate in a 45C environment (they actually run about 37C-38C), should last 48,000 hours or so, give or take!
The heat-sinks will all be "re-pasted," and possibly sanded/lapped to ensure optimal contact. I'm currently waiting for some Keratherm Kerafol 86/26 thermal pads for the Sanken Darlington output transistors. I was going to use some 0.6mm Aluminum Oxide ceramic thermal pads, but having to put thermal paste on both sides of the pads makes it really messy to deal with.
Finally, I'll change out the rear panel Powercons, the 5 pin XLR, and 1 set of speaker terminals on each side as well when I take out the Power Amp Modules. And of course, braid the wires and "maybe" sleeve them, depending on how everything looks. I'm also going to put spade connectors on the pre-amp section when I strip apart the whole unit again, just to make it easier for myself to assemble/disassemble everything. Of course, after re-capping the Power Amp modules, I'll re-bias as necessary. Oh, and make new umbilical cords from fancy pancy, silver-plated, OCC copper 16 conductorx15AWG (sourced from China for cheap) because...why not. The cable is cheaper than the Neutrik PowerCons I'll be replacing anyways.
That should be it! Feel free to leave comments/suggestions/insults!
Thanks!
verkion
Reply With Quote
Originally Posted by verkion
I guess they figure that their products are cheap enough that they don't want to "splurge" on better components. 