|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
184.59.99.28
Thought I share some recent experience with my XM-44.
When I built my XM-44 kit I used Nichicon FG and Elna Silmac II caps in place of the electrolytic supplied with the kit. After being dissatisfied with the supplied components for the frequency modules I built new ones using WIMA FKP2 caps that I hand matched to within 1% along with Holco resistors which greatly improved the cleanliness of the sound.
Next I then tried OPA 1642 opamp which TI recommends as a direct upgrade the OPA2132's that come with the XM-44. At first they sounded good but after a while they became fatiguing with the high end just not sounding that clean, more detail but a little harsh so I took the 1642's out and went back to the 2132's.
Next I purchased Russian Teflon to replace caps in another set of frequency modules that were previously built with Vishay metal film resistors and the caps supplied by Marchand. As the Russian Teflon caps were a 10% tolerance the values varied widely and I was only able to match the Teflon caps to within 2%. Once I installed the Teflon cap frequency modules the improvement was immediately noticeable. The high end detail was much improved, very smooth and extremely listenable in the long term.
After listening to the Teflon's for a few weeks it became apparent that the upper high end had some fall-off. At that point I decided to try the OPA 1642's again and they proved to provide amazing results in concert with the Teflon's. The high end fall-off disappeared and the sound up and down the frequency range just got smoother, more detailed and most importantly more listenable in the long term.
There was just one last minor problem; with the 24 position attenuators I just couldn't match the levels of the tweeters to the mids to my satisfaction. The 24 position DACT just didn't give me a fine enough adjustment so I got a pair of 47 position attenuators from the Hi-Fi collective in the UK and that did the trick. The system is now well balanced from top to bottom and very finely detailed.
Anyone have any similar experiences or other XM-44 mods they want to share?
Follow Ups:
Their tube-based pre-amp and electronic crossover is miles ahead of the opamp based one.
The tube base one is far simpler and sound better. I have built one similar. They are great.
Never modded the XM-44 but one observation. Marchand uses fairly large resistors (> 100K) and small caps (low nf). Large value resistors are noisier and small caps are affected by parasitic capacitance. It seems to me that using resistors in the 10K region and caps in the 10 nf range would be a better design. Also rather than spend the money (oops too late) on 48 position attenuators, I would have used pots to get the right balance and then by-pass the attenuators all together and put in fixed resistors. Fortunately I got good balance with the 12 position attenuators so I never bothered to by-pass them. Currently I have MG-3.7i which use a series XO and are not suited to biamping without a major overhaul and redesign of the crossover so the Marchand is gathering dust.
You have two modules, one with WIMA FKP and the other with Russian teflons? Same xover frequency? If so, do you have a preference? The FKP's are the best cap WIMA sells, made with stacked film and non-magnetic leads. How did you fit huge teflon caps in the same space?
Cap upgrades in line level applications, whether electrolytic or film, are almost always audible and worthwhile. What you experienced with the teflon cap's highs "falling off" after a couple weeks of playing was probably break-in. Most caps will change character, some significantly, after passing signal for a while. Some take a few hours to settle completely, others can take hundreds of hours.
Polystyrene is another alternative to consider for small values typically found in active filters.
The Vishay metal film resistors are decent, but their bulk foils are even better, and the Vishay naked Z-foil is the most neutral resistor available. I've read good and bad things about Holco's. I've never tried them, but I have tried MANY others.
Interesting finding on the opamp/passives combinations. Great that you discovered the right one.
I have learned that it's best to use a variety of differently voiced components throughout a circuit to avoid any one part making its signature too prominent.
Peace,
Tom E
My comparisons were with all modules Wima and all modules Teflon.I had Wima FPK2's on all the frequency modules (sub, low-mid, high) and when I built modules with the Teflon's I build all the modules (sub, low-mid, high). The way I got them to fit on the card was to alternate the placement on the modules. One from the front of the card then one from the rear, one from the front and so-on as shown.
My point to all this is that the Wimas made the stock unit better but their shortcomings became apparent with the 1642's. With the use of the Teflon's the 1642's were able to shine. Prior to experimenting with the Wimas I built frequency modules with Rel-Cap polystyrene caps from Germany and was extremely disappointed with them. The Wimas are far superior to the Rel-Caps.
Prior to the XM-44 I used a Marchand XM-9 3-way that I modified complete with Russian Teflon caps and was very pleased with the sound. As with most of us audiophile addicts I was always looking for something better and when I got the chance to pick up a not yet assembled XM-44 kit from an overwhelmed purchaser cheap I jumped at the chance. Once assembled though I was disappointed, the XM-9 sounded better. Having sunk a bit of money into the XM-44 but not half as much as a full price assembled version I started upgrading to point I'm at now and I can easily say the XM-44 is now head and shoulders over the XM-9.
As far as the Visay naked Z resistors go, I would love to use them, however, the modules would need a total of 40 resistors which would total over $600 but the real show stopper is that the highest value I found on Partsconnection is 100k while several of the modules use values greater than 100k on up to and including 1meg. I suppose one could put resistors in series to get required values but that would increase cost and space requirements drastically. Willing to try other brands that encompass the full normal range of values.
Edits: 09/21/15
You can order ANY value of TX2575s from Texas Components via phone or their online web store. I used their on-line order form twice and just e-mailed it back to them. They don't make the 1% tolerance any more just .1% of course you could get .05%.... You can even order just one (???!).The higher values cost more but they are still cheaper than Parts Connexion (I really like them, but not for this) or putting values in series. I think they gave me a 15% discount, but then I bought $800 worth when I redid my Pass Labs Xono phono preamp.
Their service is excellent. Turnaround time is 1-2 weeks max with good personal communication. They even sent me an email a month later to ask how my project turned out.
Here's the link:
http://webdirect.texascomponents.com/
Edits: 09/22/15
Very nice construction.
If you are interested, and I believe you might be because you've already gone to great lengths, you could try the Z-foil resistors in a few critical paths such as feedback only in the high pass filters. No need for them in low pass areas. If odd values are required, you might have difficulty getting them from a commercial supplier, but perhaps Texas Components could help you.
I've read that those Relcap styrenes (RTE?) can be a little sharp. I've never tried them. I bought some styrene caps on eBay. The ones with magnetic leads didn't sound very good, but another batch with non-mag leads and larger physical size are very nice.
Peace,
Tom E
Thanks for the suggestions. I purchased the RelCaps on eBay and they were stated to be RTEs and picture that went along with the listing showed them marked 'RTE'. The caps I received, however, were marked RelCap but not RTE. When I contacted the seller he naturally claimed that they were indeed RTEs. I was anxious to try them and didn't want the bother of returning them to the seller overseas and there was no doubt that they were RelCap polystyrenes so I kept them to try. The result was disappointing as they didn't sound as good as the stock Marchand provided polypropylene caps which led me to the purchase of the Wimas and ultimately to the Russian Teflons. If they were not RTE's that may explain the disappointment but in theory standard RelCap polystrenes shouldn't sound worse than the polypropylenes but they did.
You say they were not marked "RTE". In fact, all the REL polystyrenes I have ever seen were marked "RT", but I sure can't say they never used RTE. That said, if you don't see RT or RTE, they ought to have some 2- or 3-letter designation on them, besides the Relcap label. REL uses a code to tell the user whether the cap is polystyrene or polypropylene, metallized or film and foil. By the way, you may know this, but REL capacitors are made in Southern California; no need to get them from Germany. The RTs are pretty nice sounding caps, so I was surprised you found them wanting or that the WIMAs could be so much better. (I'm guessing the WIMAs are metallized polypropylene types or maybe film and foil.)
I doubled checked where I got the caps and it was Hong Kong ... I got some other components from Germany and I them mixed up. The caps are in an epoxy case and are impressed with the marking Rel-Cap but don't say RTE. The listing says they are RTEs and the picture with the listing were stamped RTE along with the value and voltage rating. Partsconnection also carries them at a much higher price and the picture they have of them are stamped with RTE as well. I might have been duped but according to Rel-Cap specs the difference between RTEs and non-RTEs is that the RTEs are tin foil polystyrene while the non-RTEs are aluminum foil which is why I decided to try them. The link below shows the seller's similar offerings.
So we can conclude that the "E" designation stands for tin foil. I've never heard one of those but I would expect it to outperform the RT types, based on the fact that the REL MIT PPFX (polypropylene film and aluminum foil multicaps) do not sound as good as PPFX-S (the tin foil version of the same cap).
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: