The Stax SRM-001 mod thread
Jan 22, 2008 at 7:36 PM Post #32 of 440
I have two spare 4.5V wallwarts. The Sony, and a RadioShack 4.5v/6v switchable 550ma in the attached photo, which I have been happy with. The RadioShack has the same size tip as the sony, but can accept different sized tips which can flip for a polarity change.

I also have a RadioShack 3v/4.5v/6v/9v/12v switchable with 1000ma that runs my SRD-X nicely. I suppose I can try all three.
 
Jan 23, 2008 at 3:38 AM Post #33 of 440
Quote:

Originally Posted by sclee /img/forum/go_quote.gif
Thank you for AudioCat tutorial,it a very useful.
I just new entry to sr-001 for few week, the module still burning, as i like to listen the change of burning.

Previously, i only want to use this on my travel on night at hotel, but after use i thing this can be compare as to or better to my using K701 system.

The speed quick and not noticeable delay, when i come back to K701, the speed is so slow and sound push out some time. The bass is more better that entry level elec. phone such as 3030 and 2020

I do not want to mod the unit so much, i study the amp, good design.
I want to improve the speed and detail of sr-001.

I like to change the 10uf 50v cap to Blackgate and paellel as a PP, i buy two 10uf 50v Blackgate N cap, but can not find good PP fitted on the room, because two BG is too big, so that i just use two BG and no PP.
That mean the cap uf had been increase 10uf vs (10+0.15)/2 =5.075uf. Normal the speed become slow and bass increase

But after i install and run few hour after, the speed is so great and faster that before about 50%. The bass is not much deep and open before, i know that should be burn at least 300 hr, as our K701 head amp BG output cap already run one year but also can listen the improve.

As i use a DC-AC selectable transformer with noise reduce function, i can not listen the noise even to the higher volume, i have 400ma and 800ma, a 800ma transformer have better bass detail and more deep.



The first thing you should change is the opamp. If you don't want to mess with the high voltage section too much, change the C6 to a 0.1uf/630v X2 series polyprop, that will increase your high voltage reserve 10X and fits directly onto the board. It will open up the highs and extend the bass quite a bit.

10uf/50V N is too big, but the DC component there (assuming you are talking about the input caps) is only about 1.5V, so the 22uf/6.3v NX will fit nicely onto the board. with that you can still add a 0.15uf polyprop as by-pass cap.
If you are not a bass head, you really only need about 4-6uf as input caps. I have tested with a pair of FKP 6.8uf/150v polyprop (huge, about the size of battery compartment) and the sound was nice and smooth, good bass but not as much as I want though (after all, my first pair of good headphones are the 600 ohm version Beyer 990)

I personally don't think the stock 001 is as fine as a 701/good amp combo. I can use a pair of silver/teflon recabled iGrado (driving by a PPA type amp) to compete with the stock SRS-001 (on two NiMH batteries), the stock 001 will win some area while the silver-iGrado/ppa will win the others, kind of breaking even. Upon direct comparison though, the dynamic phones always sound hazy, as if looking though a layer of thin fiber....

I'd like to hear some comparison result of FC/SFC modded 001 vs. K701 driven by a $300 amp. Will be interesting.
 
Jan 23, 2008 at 3:32 PM Post #34 of 440
Quote:

Originally Posted by AudioCats /img/forum/go_quote.gif
The first thing you should change is the opamp. If you don't want to mess with the high voltage section too much, change the C6 to a 0.1uf/630v X2 series polyprop, that will increase your high voltage reserve 10X and fits directly onto the board. It will open up the highs and extend the bass quite a bit.

10uf/50V N is too big, but the DC component there (assuming you are talking about the input caps) is only about 1.5V, so the 22uf/6.3v NX will fit nicely onto the board. with that you can still add a 0.15uf polyprop as by-pass cap.
If you are not a bass head, you really only need about 4-6uf as input caps. I have tested with a pair of FKP 6.8uf/150v polyprop (huge, about the size of battery compartment) and the sound was nice and smooth, good bass but not as much as I want though (after all, my first pair of good headphones are the 600 ohm version Beyer 990)

I personally don't think the stock 001 is as fine as a 701/good amp combo. I can use a pair of silver/teflon recabled iGrado (driving by a PPA type amp) to compete with the stock SRS-001 (on two NiMH batteries), the stock 001 will win some area while the silver-iGrado/ppa will win the others, kind of breaking even. Upon direct comparison though, the dynamic phones always sound hazy, as if looking though a layer of thin fiber....

I'd like to hear some comparison result of FC/SFC modded 001 vs. K701 driven by a $300 amp. Will be interesting.



How the affect for 0.15uf by-pass cap when i use higher or lower value PP cap ?

I don't thing on two NiMH batteries SRS001 can be better that K701, at least use 4.5v power to SR001. K701 good on big and complex such as Symphony. But SR001 better on solo or Concerto, the brass,guitar, wind and volin. The detail of instrutment can be fully delivery as like real thing, the K701 more dynamic and deeper bass .

The cable between sound source and amp will affect the sound so much on SR001, so that should be use good 3.5mm plug
 
Jan 24, 2008 at 4:49 AM Post #35 of 440
Quote:

How many hours is "a few hours"?


two hours runtime on 1 Li-ion 14500
 
Jan 24, 2008 at 4:54 AM Post #36 of 440
the reason for 0.12uf (sorry, I got it wrong, should be 0.12 instead of 0.15) is that will be the largest value polyprop that can fit onto the board as a direct replacement of the original 0.15uf poly(ester?). of course you can use smaller value, I have a habit of using as large and as nice a cap as possible, so the 0.12uf/63v was the default choice for me.

On AC adaptor the stock 001 can beat the silver-iGrado/PPA, but I have no confidence with it competing with the 701 (on the other hand I have never compare the silver-iGrado with a 701 either).

Yes the baby stax is unforgiving (a lot more so after the mods), you can hear all kinds of problems in your system, wrong cable, bad recording files, bad source...... sorry about your wallet



only 2 hours on a 14500 in a stock 001? It only draw about 220mA, I though the 14500 should last at least three hours....
confused.gif
 
Jan 25, 2008 at 4:39 AM Post #37 of 440
Quote:

only 2 hours on a 14500 in a stock 001? It only draw about 220mA, I though the 14500 should last at least three hours.


Can also be three hours (I hate looking at the clock when I am lying in bed listening to music
biggrin.gif
) Also I try not to run them dry (I am using protected cells though)
 
Jan 26, 2008 at 5:34 AM Post #38 of 440
Quote:

Originally Posted by AudioCats /img/forum/go_quote.gif
the reason for 0.12uf (sorry, I got it wrong, should be 0.12 instead of 0.15) is that will be the largest value polyprop that can fit onto the board as a direct replacement of the original 0.15uf poly(ester?). of course you can use smaller value, I have a habit of using as large and as nice a cap as possible, so the 0.12uf/63v was the default choice for me.

On AC adaptor the stock 001 can beat the silver-iGrado/PPA, but I have no confidence with it competing with the 701 (on the other hand I have never compare the silver-iGrado with a 701 either).

Yes the baby stax is unforgiving (a lot more so after the mods), you can hear all kinds of problems in your system, wrong cable, bad recording files, bad source...... sorry about your wallet



only 2 hours on a 14500 in a stock 001? It only draw about 220mA, I though the 14500 should last at least three hours....
confused.gif




Finally I added the epcos 0.10uf 63v which can fitted to the 50v 10uf BG N Cap, i also change 5 0.01uf to WIMA RED MKP 0.01uf to improve the performance.

I alsp want to change the 3.5mm phone jack to a better one, but now i am seeking and can not found high grade jack.
 
Jan 27, 2008 at 7:22 AM Post #39 of 440
Finally the blackGates burn-in is completed. I connected them all in parallel and act as the output filter stage for a headphone jack, then I let them run for 24/7 for two weeks (more or less, I had to turn it down a bit when I go to sleep, and sometime I forgot to turn it on after CD change). Using the headphones I can hear how the burn-in changes the sound: at first it was muddy and bass heavy, after two days it started to clear up, then at around one week the sound became very sharp and metallic, then it slowly smooth out. Not sure if that was a 100% accurate impression since I was using two pairs of different headphones alternately…. But anyway, burn-in did change sound for the BlackGates. Picture below is the “board of blackGates”
smily_headphones1.gif


IMGP1515.jpg


Now the output caps burn-in: I connected the sonic I & II in parallel with the blackgates and run together, but apparently low voltage burn-in doesn’t work for high voltage caps, I could hear the sound change after installing them into the 001. Fortunately the sound settled down after about 3 hours. I am sure it will require more time to fully settle down but I think after 8 hours they are stable enough to get a impression, longer burn-in just improve what is already there. All three (Sonic I, II, and Auri) got 8-10 hours run time in the amp prior to the listening test.

This will be a long report. I have fund the 001 is so sensitive to input cap / output cap / IC combinations, that comparing just individual components doesn’t give conclusive results. The same component can have totally different effect in different configurations. I will separate the contents into four posts instead of one super long report. The detailed list of parts-to-use-and-things-to-do in different class mods will follow shortly after.
 
Jan 27, 2008 at 7:26 AM Post #40 of 440
Interconnects can cause major sonic difference here, it is more about using the “right” IC in the combination instead of using a “better” one, plugging in a more resolving/ more expensive IC doesn’t guarantee a better sound. So, check your IC inventory to see what you already have and what you plan to buy for your other audio gears, then pick a mod package. Using the wrong IC can cause unbearable sonic results, the baby stax will make sure you hear the mismatch …..The cost difference between different classes can be cheaper than the cost of another good cable.

The FatCat class uses non-boutique output caps with AD8599 as opamps. In general, this class prefers more bass-heavy IC’s, such as the tri-braid Teflon/silver and Esoteric Musica200 I used. (the M200 is a relatively inexpensive IC, $20-$40 price range). I have also tried quad-braid Teflon/silver, twisted pair Teflon/silver, quad-braid cotton/copper, Teflon/sivler in parallel PVC tubes, and the tri-braid turned out to be the “right one” (while in general the quad-braid Teflon/silver is considered as a better IC). This probably have something to do with the output cap selection, the winner is BC component MKP 0.011uf/630v, this cap is gives a lot of details but is a little overly bright (comparing to the boutique caps that it). By the way, this mod class prefers the tri-braid silver over quad-braid silver regardless of which non-boutique output cap was used (cap test results in post #20).
In lower price IC’s, my Musica200 sounds pretty good while the Monster Interlink250 sounds more detailed but is also too bright for long term listening, at least not for me.

The Super and Ultra class uses boutique output caps and you MUST use fast and cool/cold cables (I used twisted silver and quad silver). Install a tri-braid IC then the sound will get thick and chunky and chewy, all the tasty goodies are there but it won’t flow well, kind of like soup instead of wine (try to drink some chunky chicken noodle soup straight from the can then you will know what I mean). This is still true with AD825’s installed in the Ultra.
IC selection for these two classes is at least a little easier than in the FatCat class, at least it looks like better cables give better sound. However, my cotton/copper quad-braid also works pretty well which really surprised me since it sounds like crap in my dynamic system (the problem was bland sound and dry bass, resolution was good)…..Also, the Monster Interlink sounds much better than M200, the 200 sounds veiled and layback and there is actually less bass
confused.gif
……

For those of you that don’t believe expensive cables make better sound but still want to get a 001 and mod it, stick with the FatCat mods, it is a little bit easier to find less expensive cable to match it.
 
Jan 27, 2008 at 7:29 AM Post #41 of 440
Input caps (top left to right: BlackGate NX 22uf/6.3v, old tantalum 100uf/16v, old dip tantalum 22uf/10uf; bottom left to right: NX 47uf/6.3v, N 33uf/16v, NX 100uf/6.3v)

IMGP1530.jpg




Output caps (left to right: Auri 0.01uf/600v, Sonic I 0.022uf/600v, Sonic II 0.0091uf/600v, the blue ones are BC MKP 0.011uf/630v.)

IMGP1523.jpg




Cap test rig (two input caps on top, four output caps into the lower sockets)
IMGP1521.jpg
 
Jan 27, 2008 at 7:36 AM Post #42 of 440
The cable used was tri-braid silver, power was lithium batteries with about ¾ charge. Signal source was modified Sony X555ES cdp (with 8599 opamp upgrade and two Auri 12uf inside. This CDP has pretty good resolution and speed – it uses not two, but six dual-opamps in the output stage ).

The best output cap that I have tried in the FatCat class is BC component 0.011uf/630v MKP and that is what the 001 had during this test. A FatCat 001 with BC cap, in general, yield much stronger bass than the Super and Ultra 001’s, but with less details and more recognizable distortion. It is also sharper (what it shows really stands out) and more raw-sounding, not as smoothed-over like in the higher level mods. With the right input cap the bass texture really shows. It is great for listening to live performance, it just sounds, well, live! The sound fills up the space… It is more also impressive, if you want to get the “Ah” and “wow” from somebody, have him listen to a FatCat 001….

All the caps below were connected in parallel with a pair of BC component 0.12uf/63v polyprop by-pass caps when tested.

*47uf/6.3v NX BlackGate -- good everything, great bass impact, but not much texture showing.

*100uf/16v tantalum – sharp/metallic sounding but shows less details than the BlackGates. Probably treble emphasized; less bass than 4.7uf NX, and not much texture.

*22uf/10v dip tantalum – not as sharp as the 100uf tantalum. Great texture (really shows). The sound is more mid-range emphasized. Details wise, about the same as the 100uf tantalum. In general the BlackGates give better details than the non-boutique electrolytics.

*22uf/6.3v NX – very spacious sounding, ok bass impact, not much texture; fast and bright.

*33uf/16v N -- relatively mids emphasized, darker, and smoother than the ones above. Strong bass impact, but texture is not very good (somehow reminds me of bass from dynamic phones).

*100uf/6.3v NX -- very soft sounding, that is all I can say.

*the original 10uf/35v United-Chemicon (possibly a SME series) aluminum electrolytic –- very muddled sounding.
* various other 47-220uf aluminum electrolytics I have in the parts box) –-not much better than the 10uf UnitedChemicon. In general the higher ripple/lower ESR ones, like Jamico WG series, sounds a little more clear, but still not there with the BlackGates and tantalums.

The overall winner --- I know I am gonna get beat-up by the audiophiles for this --- is the 22uf/10v dip tantalum. Yes the tantalum isn’t very linear but relatively speaking, this one impressed me the most. I suppose if I listen to the 47uf NX more I could like it more but I am a kind of a bass head (still in the Nile) and the bass texture really impressed me. Don’t ask me what make/model/parts # those tantalums are, I truly don’t know, if I remembered right they were taken from an old computer power supply (therefore more burnt-in than any of the BlackGates). Maybe after more burn-in the 47uf NX will start to show good textures too, but not when I did this test.

Recommendations: buy some dip tantalums when you order the mod parts from digi-key and try them out and find the best one…. That way all mod parts will come from one source, you don’t have to place another order to partsconnexion or SoniCraft, this will save you like $10 on shipping alone…
 
Jan 27, 2008 at 7:52 AM Post #43 of 440
Test conditions:
IC-- quad braid silver/Teflon;
power -- lithium rechargeable at full to ¾ full;
signal source – modded X555ES.

Again, if you want to do this mod (or the Ultra), please make sure you have clear and fast cables. When I first started the boutique cap test I was using the tri-braid (which was the winner for the FatCat 001) and the sound was so thick and chunky, I was very disappointed and depressed—seriously—for a few hours. Until I figured “might as well try my other silvers”…….

In the SuperFatCat class the opamp’s are still the easy-to-install AD8599’s, but we can use as much boutique stuff as we want, as long as there is room for it….

Boutique output caps:
Generally speaking, the Sonic I sounds thinner and lighter than the II and Auri. It puts a lot of details right in front of you, has a wide sound stage, the most “visible” bass texture in the three (but not necessary more visible than the FatCat class), and is faster than the Sonic II. The Auri, on the other hand, is creamier (like what others have told me before), has as much detail as the Sonic I, but the details don’t stand out as much. Sonic II has the thickest and smoothest sound (which is not necessary a great thing), it doesn’t show as much detail as the Sonic I and Auri (but still more than the FatCat class). it is best with sharp sounding cables (with this cap installed, my twisted pair silver sounds better than quad-braid).
Comparing the three, I’d like to say Sonic I sounds like water, Auri like wine, while the Sonic II like broth…..,(or we can say the Sonic I = gin, Auri = whisky, Sonic II = cognag for all your drinkers, I hope this is a good analogy because I don’t even drink
wink.gif
). Overall, I like the Auri’s slightly more when using the AD825, but that will be the Ultra class mods. Auri is also difficult to squeeze into the case, per the spec it will barely fit but I didn’t realize the lead wires are so thick (adds to the over all length and requires more room), or I might not have order them.
Between the easy-to-install Sonic I and II, I wouldn’t say the Sonic I is always superior than the II, it also depends on what IC you already have and what source you will be using. For a I-mod with a short three wire cable the II might be a better choice.

Input cap results (all parallel with 0.12uf BC polyprops):

Sonic-I as output cap:
* 22uf/6.3v BlackGate NX – lots of details, light and fast. Slightly bright and thin. Wide sound stage. Bass ok.

* 33uf/16v N – rich, meaty, layback and slightly sluggish. Sound stage is not as wide as the 22uf NX.

* 100uf/6.3v NX – better bass than the two above. Not as fast as 22uf NX but faster than 33uf N. More up front sounding (more mids?).

* 47uf/6.3v NX – sharp, fast, spacious (wide sound stage), good bass, good texture.

* 100uf tantalum (might as well try them, right?
wink.gif
) – clean but more constricted; mids-emphasized, veiled and less details than any of the blackGates.

* 22uf dip tantalum – again mids-emphasized. Clean sound but slower and not as detailed as the blackgates.

--- The overall winner is 47uf NX, though the 100uf NX might work better for cold sounding sources.



Sonic II as output cap:

* 22uf/6.3v NX – good clarity, strong bass, natural sound stage (neither wide nor constricted), smoother than the 47uf NX below.

* 100uf/6.3v NX – bright, bass shy and drier
confused.gif
You will probably jump up and say “how the hell….” Well, that was what I said. Checked the connection, caps were plugged in all the way. I am only writing down what I heard, not why --- 1-27-08: I went back to this one again. It appeared that there is (or supposed to have) bass, I can hear the attemps, but somehow the bass doesn't come out right therefore not being recognized as bass..... kind of like swing drum sticks really hard but not really hitting the drum.

* 33uf/16v N – bright and cold but not as bad as the 100uf NX. Ok bass.

* 47uf/6.3v NX – similar to the 22uf NX but more spacious.

* 22uf/10v dip tantalum – clean, not as much detail as the blackGates, but somehow is more “impressive sounding” than the blackGates.

--- the overall winner is, again, the 47uf NX. If the source is a little “raw” "edgey", then the 22uf will smooth it out better.

Now some might ask what do I mean by “more impressive sound” like in the FatCat class… To me it means it grabs your attention right away and impress you right there, you don’t have to listen to it for a long time before realizing how good it actually is. How can this be? Well during the Seattle meet (last May), I was really more impressed by the AT AD2000 than the W5000, while the general consent is the W5000 is a “finer and better” phone. It is probably not just a personal taste thing, the finer stuff takes longer to enjoy, and you can enjoy it for longer period of time. I guess I can also say the “impressive” FatCat class is like hard liquor that gets you tipsy right away but you are always kind of standing (not fully convinced), while the SuperFatCat and the Ultra are like something smooth and by the time you realize it you are already on the floor, 100% taken over…. Again I hope this is a good analogy
wink.gif
Liquor experts please provide the proper name of the wine so I can refine my writing here
biggrin.gif
 
Jan 27, 2008 at 7:55 AM Post #44 of 440
...... (deleted some unnecessary lines)
....Also, I need to do some test without those 0.12uf BC polyprops and see how it sounds. The BlackGates are indeed so fast they might not need by-pass at all…..…..-- update 1-27-08: using the polyprop bypass caps is a must. It adds ambience, sparkle, and sharpness to the sound. The blackgates are fast sounding for being electrolytics, and that is about it. Without the polyprop bypass, the claps of hands can sound like slamming forarms together in some combinations, missing so much details that the sound was almost not recognizable. (I only knew they are claps because I am very familiar with that track
mad.gif
) .


1-27-08 Update:
OK, enough rant, now the Ultra class results:
--test conditions the same as above two classes.

--Notes on music used in these tests (all classes): the baby Stax excels at acustic music and vocal, so that is the kind of music I am tweaking it for. I use only three songs to evaluate each cap combination: Cascade from Jesse Cook's Montreal album ( acustic guitar, live performence to test ambience), Pool of Dreams from Oracle's Pool of Dreams album (bass presentation, details, and vocal naturalness), and Call Off The Search from a Katie Melua CD sampler disc ( intimacy of voice). I think these three tracks covered most of the test points pretty well.

In the ultra class there is basically no limitatin on what to use as long as it can be put in the case. Opamps are AD825. All input caps are by-passed with 0.12uf BC polyprops. You might notice (yes you will) I also use 220uf/6.3v NX here, the reason why they were not listed in the above two classes was due to the size, they are simply too big for direct installation. But on the other hand the 220uf NX's don't sound good with AD8599 anyway (I actually did the tests, just didn't bother to include in the reports).

Auri as output:
* 47uf NX -- too bright and thin.

* 22uf NX -- all I can say is mediocre sounding.

* 33uf/16v N -- good everything but not enough bass…..

* 220uf NX – best ambience, best detail, good texture. I can listen to this one all day, and I have noticed I didn’t tend to turn the volume knob high in order to get into the sound. But the sound is not as sharp as I want.
Note: the 220uf NX’s weren’t burnt-in with the other blackGates. There was only about 4 hour of burn in done to them. I am sure the sharpness will improve after a full burn-in, lets hope the bass doesn’t get weak by then.

*100uf NX – thicker, darker sound and more layback, strong bass but not much texture.

--- the overall winner here is the 220uf NX. I hope it will only get better with more run time. The second place is 33uf N (not everybody is a basshead after all).
--- In general the Auri gives more/finer details than the Sonics (both I and II).


Sonic generation I as output:

* 100uf NX – sharp and upfront, the bass didn’t impress me. Otherwise great.

*47uf NX – more bass than the 100uf, good texture, but a bit too bright.

*33uf/16v N – warm, the sound has a lot of weight. Lots of fine details, but not sharp/crispy enough. Not enough bass.

*22uf NX – very sharp, treble/upper mids emphasized, gives a sense of great details. When listening to vocal I thought I heard the sticky lips closing/opening when the singer sings, and it shows the raspy-ness of voice very well, so well that I wonder if it is actually really like that…... Not enough bass.

*220uf NX – the most impressive one of the group. Balanced, spacious, real, lots of weight, and good bass and texture.

--- The 220uf NX is the winner, again. Some might also like the 22uf’s hyper details, again it is all system dependent.



Sonic generation II as output:

*220uf NX --- good everything, nothing bad. Gives a slightly better feel than the Auri/220uf NX combo, but is not as detailed and not as much ambience.

*47uf NX --- bright and cold.

*33uf N – layback, distant, and very “fine” sounding. Doesn’t give me much enjoyment. Bass is good.

*100uf NX – Sharper than the 220. not as much weight and ambience as the 220; more refined and polished sounding (not as "raw", probably due to missing minute details). Balanced and enjoyable.

--- yes you have guessed it, 220uf NX won again. The 100uf NX might work better for a more “raw” source though.



Over all, I like the Auri/220uf NX combo the most. There is no absolute winner in the ultra class, at this level it is more about system matching and personal preference.

Now I think I understand why there aren’t too many people using electrostatics, it is so picky! “Better” components don’t necessary improve the final sound, everything is down to system matching once you have reached a certain height. Some of you might want to stick with dynamics where you can at least predict the results ….. (but you won’t get the crystal clear sound of the electrostatics
wink.gif
)
 
Jan 27, 2008 at 10:51 AM Post #45 of 440
Thanks for these updates, audiocats.

You've been doing a LOT of work (and writing) and I just wanted to say it is genuinely appreciated out here.

I look forward to the further details when those additional options are also done.

Spare SRM-001 and soldering iron ready here...
biggrin.gif
 

Users who are viewing this thread

Back
Top