The Stax Thread III
May 22, 2023 at 11:46 PM Post #23,641 of 25,611
those power resistors and caps in question hardly amount to say $25 - why not replace those when having the thing open anyway since the fault is well known?
did you get any explanation as to why the static was left channel only and caused by a fault in the power supply common for both channels?
The amp didn't ship with the issue a year ago so I assume the caps and resistors were fine at the time. The issue likely isn't so common that those parts should always be replaced when modifying the amps listed.
He didn't explain why it was only in the left channel, but by the time I sent it out to him the amp stopped working entirely.
Any diagnosis by me is mostly guesswork; I'm a mechanical engineer and circuits are voodoo to my people.
 
May 23, 2023 at 1:37 AM Post #23,642 of 25,611
Apparently, the "cartridges" are to keep folks from selecting the wrong bias. My question is what happens if you have, say, an x9000 and you select 600 or 620 bias, as is easy to do on the original IESL. Will the higher bias damage the phones?
Nothing would happen at that sort of bias difference other than a little louder I imagine. Different story if it was grossly over the rating like feeding pro bias to normal bias phones.
 
May 23, 2023 at 10:27 AM Post #23,645 of 25,611
This reminds me of Koetsu cartridges old versus new. 4 decades of production with lots of well known long-running models, plus mysterious versions / variants that are not well documented. Then you factor in deterioration and wear. And the "lore" of the late old man Sugano Sr. being magical somehow (at least, more so than his sons who took over the business). So when somebody claims their vintage Onyx or Rosewood sounds "way better" than modern production, there is no way to guess what that actually means without going over to hear it. I had a vintage Onyx that seemed to have some "magic" beyond modern Onyx (which I also had) but it was impossible to isolate this "magic" from possible colorations caused by its (obviously) heavy stylus wear and suspension degradation. In the end I had to get it rebuilt because the inner groove distortion was getting too high. My guess/conclusion is that the motors do have a different voicing -- older motors are voiced a bit "sweeter", while new ones are cleaner and more linear -- the new production has introduced new magnet (Platinum) and body materials (exotic stones, Urushis) to partially compensate for the sweetness. They nailed it on the Blue Lace model :)
 
Last edited:
May 23, 2023 at 1:04 PM Post #23,646 of 25,611
Anyone here messed around with collecting multiple LNS and heard a difference between different serial numbers the way there is with the 007?
Yes. I do have one with a suspicious Airbow-like SN (yes, on an original headband) that seems to be better than others (more clarity and 'liveliness' without loosing anything in smoothness) even though I have not seen them operating before the 404, so not sure if that was a custom one-off or what's up with that.

Otherwise, differences can also be attributed to how well the process of re-gluing the drivers went.
Because pretty much every single pair I have or had suffered from this problem, and if dust gets in during the process...it's not good.
I have started to see this appear on some 202/303/404 sets as well.
 
Last edited:
May 24, 2023 at 5:40 PM Post #23,647 of 25,611
srm212,252 etc are 6 volts 1 amp. So 6 watts
koss is 9 volts 1 amp. So 9 watts.
topping is 15 volts 1 amp. So 15 watts
.............
carbon cc is 125 watts
T2 is 225 watts
megatronxxl is 325 watts

no way an opamp driving transformers is going to be able to do .0008% thd

amir (incorrectly) measured koss at .01%
i (correctly) measured koss at .005% which is the limit of my equipment.
(i'm waiting on the new quantasylum qa404)

i do have a problem with the quoted noise figures. Likely they are measured with the attenuator in place.
That is clearly not correct.
What does input power have to do with output power? Especially if a big chunk of that is used just for heating the tubes and the regulators, wish you could quote output power instead because that’s what matters
 
Last edited:
May 24, 2023 at 6:15 PM Post #23,648 of 25,611
Anything with tubes will waste a crapload of power on filament heaters and such. Anything class A will also waste a ton of power as heat. You cannot use input power as any sign of performance. If anything you can draw conclusions that anything with tubes will have more distortion, more susceptible to microphonics and a higher noise floor. Thats not to say they don’t sound good though…
 
May 24, 2023 at 6:33 PM Post #23,649 of 25,611
If an electrostatic headphone has an impedance of about 150 Kilo ohms like they typically do, even if you feed it 1000 Volts rms Which I doubt any headphone amp in the world is doing, except maybe an SRD7 setup, it will be sucking only 7 Milli amps which calculates to 7watts, VA to be precise.
 
Last edited:
May 24, 2023 at 6:51 PM Post #23,650 of 25,611
Enjoy your 10 Watt *total* power consumption amps -- I'm sure they're just fine :sweat_smile:
The difficult part of high quality electrostatic amplification is not in pushing raw power into the headphones - they don't use that much. It's in keeping the output devices biased into class A at rail voltages of 400V+, even regardless of tube heaters. That's why so much heat is burned up relative to the raw output power. And that's why people keep get tempted to use step-up transformers to avoid that issue.
 
Last edited:
May 24, 2023 at 7:22 PM Post #23,651 of 25,611
Enjoy your 10 Watt *total* power consumption amps -- I'm sure they're just fine :sweat_smile:
The difficult part of high quality electrostatic amplification is not in pushing raw power into the headphones - they don't use that much. It's in keeping the output devices biased into class A at rail voltages of 400V+, even regardless of tube heaters. That's why so much heat is burned up relative to the raw output power. And that's why people keep get tempted to use step-up transformers to avoid that issue.
quoting input power as a ranking system though is one heck of a stretch, all a manufacturer has to do is build in a short circuit and say mine is the best.
 
Last edited:
May 24, 2023 at 7:25 PM Post #23,652 of 25,611
quoting input power as a ranking system though is one heck of a stretch, all a manufacturer has to do is build in a short circuit and say mine is the best.
He didn’t advertise it as a black and white ranking system, but it does convey useful
information given the application and the disparities involved.
 
May 25, 2023 at 12:01 AM Post #23,655 of 25,611
On the topic, I can finally share some data from testing my Lundahl LL1630 setup and limited testing with all the gear I have available... It's fascinating to see it all plotted out and compare STAX gear to transformer setup. The STAX SRM-1 surprised me with its low end performance but also how quickly it falls off, also interesting was its behaviour changing quite dramatically based on the load used. Even a 2 wire reading from my oscilloscope (1mOhm) would drag down its output significantly...

In contrast the Lundahl LL1630 just doesn't give a crap really, it will deliver consistent results regardless of what load I put on it and is basically flawless across the FR (voltage output), its only limitation seems to be its fall off at 20hz and lower (270Vrms limit when the core saturates) it may be possible to work around this with a different winding configuration or possibly series them???

Out of the amplifiers used, the Topping LA90 was the most linear, clean and consistent from what I saw and 2 of them fit nicely on a desk with balanced inputs used. Also threw in some random testing with different amplifiers to see how it impacts the response and also a 1960s Encel tube amplifier for the LOLs of "tube sound"

Tried various loads from 120pF capacitor to 150kOhm resistor with capacitor in series / parallel and ended up settling on the 120pF capacitor alone with a single wire reading off my scope and multimeter (10mOhm) to give the SRM-1 a fair chance. I chose 150Vrms as the reference based on my listening levels.

What was used:

STAX SRM-1 mk2 (Fully restored and calibrated)
STAX SRD-7 modified with 4.7ohm resistors and switching removed (performed the best compared to my other SRDs)
Lundahl LL1630 - Driven by Topping LA90 Monoblocks (Roughly 90Wpc)
Lundahl LL1630 - Driven by Sony TA-N9000ES (Roughly 200Wpc)
Lundahl LL1630 - Driven by Luxman LV103 Tube / MOSFET Hybrid (Roughly 75Wpc)
Lundahl LL1630 - Driven by Encel CSM-40 OT Tube (Roughly 18Wpc)

chart1.jpg
chart2.jpg
chart3.jpg
chart4.jpg
 
Last edited:

Users who are viewing this thread

Back
Top