[SolarCetacean]

They clearly don't see the need to do a full voltage swing at 20 kHz, because, well, it's pointless lol

Do you apply those same relaxed standards to regular headphone amps? Does it matter if a normal headphone amp rolls off in the high frequencies? If a regular headphone amp measurably rolled off the upper treble by several dB, ASR and all of the amp measurement reviewers would be throwing a fit. Headless panthers and all. Yet with estat amps, somehow we get into this discussion on whether amps really need to have linearity throughout the entire frequency range.

An ideal amp is a wire with gain. That means at minimum full bandwidth from <20 Hz to >20KHz. And it's not like we're dealing with excessive amounts of headroom either. Bob Katz measured the Audeze CRBN acoustically on the Mjolnir Carbon and that amp, one of the most powerful amps on the market, only went up to 113 dB SPL before clipping. This isn't normal headphone amps that have practically unlimited headroom on regular headphones.

 

[tabness]

K, just tried it. Yeah looks like I have perfect hearing.
I was able to hear 20khz, tell the difference between 0.1db, can tell 2c difference in pitch ( I got 1c correct but not sure) and could tell 1ms difference in sync, etc etc

You are super blessed! I can only do the minor 1C pitch changes consistently (I didn't play an instrument in band cause who wants to play like violin or trumpet but I did briefly play guitar in high school which helps). Other than that I can only consistently get .5 dB volume changes (.2 and .1 are too hit and miss), and consistently only hit 15 kHz hearing

(for an example of consistent, I can still get decent scores on 20 kHz, but not perfect, so I can't consider myself to have 20 kHz hearing)

Your subjective opinions on the amps are cool, but they're not the point of the discussion right? I'm not really interested in the supposed benefits and differences between mafia amps to begin with so I don't bother engage in those conversations. You were bringing up the output stage current and making a point of why it matters, I think that makes no sense and I explained why in fairly great detail.

Do you apply those same relaxed standards to regular headphone amps? Does it matter if a normal headphone amp rolls off in the high frequencies? If a regular headphone amp measurably rolled off the upper treble by several dB, ASR and all of the amp measurement reviewers would be throwing a fit. Headless panthers and all. Yet with estat amps, somehow we get into this discussion on whether amps really need to have linearity throughout the entire frequency range.

An ideal amp is a wire with gain. That means at minimum full bandwidth from <20 Hz to >20KHz. And it's not like we're dealing with excessive amounts of headroom either. Bob Katz measured the Audeze CRBN acoustically on the Mjolnir Carbon and that amp, one of the most powerful amps on the market, only went up to 113 dB SPL before clipping. This isn't normal headphone amps that have practically unlimited headroom on regular headphones.

Please explain why your standard even is a full voltage swing at 20 kHz? What purpose is there for it? What are you getting from this particular metric? Because if this is your standard, then guess which amp meets the bar and is not current limited? The little 252S lol

Reading your post again, I think I understand the issue. What we were discussing is an amp being able to drive particular frequencies at that amp's maximum output voltage. You seem to be confusing this with it meaning the amp is rolled off in general (simplifying because STAX obviously has non linear amps by design like the T1S which purposely roll off). Rest assured it's not that at all output voltages.

I've been around the STAX online group a bit and have heard the "wire with gain" philosophies and blah blah blah, it doesn't seem to mean anything in this discussion. The linearity of the amps isn't in question.

 

[Chefguru]

Your subjective opinions on the amps are cool, but they're not the point of the discussion right? I'm not really interested in the supposed benefits and differences between mafia amps to begin with so I don't bother engage in those conversations. You were bringing up the output stage current and making a point of why it matters, I think that makes no sense and I explained why in fairly great detail.

There is a concrete reasoning for why the amps are different and I do believe that those differences matter. My comment on owning/having owned all the equipment in question is to point out this isn't theoretical on my end.
If there isn't a difference between the amplifiers, then why does stax even produce different models? Just marketing? If that's the case - then why believe anything they say...

I just dug through a closet and found my 252a. Still has bass distortion above 85db on the 007 lol.

 

[tabness]

There is a concrete reasoning for why the amps are different and I do believe that those differences matter. My comment on owning/having owned all the equipment in question, it isn't theoretical on my end. If there isn't a difference between the amplifiers, then why does stax even produce different models? Just marketing? If that's the case - then why believe anything they say...

I just dug through a closet and found my 252a. Still has bass distortion above 85db on the 007 lol.

But the output stage current limitations of the majority of the STAX amp lineups do not seem to be the concrete reasoning for why the amps are supposedly "different" for the vast majority of cases (and was the actual concrete reasoning in doubt? differences in power and differences in circuit topologies as STAX clearly tunes various amps not to be linear are going to be why the STAX amps sound a bit different and there are different versions)

Every STAX amp I know of from the Xh and descendants to the T2 will clip on quite a few of the phones before the phones themselves reach their max SPL. I'm not sure this is a big gotcha or anything we didn't know. Because the output stage current of the 252 doesn't matter, it's voltage limited lol (further you mentioned bass, and well, slew rate limiting doesn't happen at the lower frequencies, it happens at the higher frequencies, so it's clearly not a "concrete reasoning" for this example)

(my man you are seriously asking about why various audio products exist? and that on a place like head-fi? lol)

@Svperstar I don't exactly know, was going by the T1 line marketing material where STAX basically said as much. I don't know of any T1S measurement, and I haven't run the circuit through SPICE or something to simulate its frequency. The T1S will also be "current limited" at its max voltage swing for high frequencies, so that could affect the frequency response as well, but we're talking real loud here for that to show up in itself.

 

[Svperstar]

(simplifying because STAX obviously has non linear amps by design like the T1S which purposely roll off).

Where does the T1S roll off at?

 

[zolkis]

New larger pads for the 007 are here, I'm afraid to take the old ones off and deal with the metal bar......such a bizarre design putting a bar in the pad where it can puncture the driver if installed incorrectly

The metal ring is to help keep the pads sealed by tensioning, since the flaps are behind the recess and the ring is lifting them a bit. But they are not strictly necessary, especially if you don't use the original pads. Keep utmost attention to the middle rubber gasket so that it doesn't break.

I found my 007 Mk1 sound better without the metal rings, it was more open, and as a matter of fact, also without the dust screens (no problems with dirt or particles - old Stax designs also didn't use a dust cloth, but YMMV). Also, I modded the original pads by removing (cutting off within the glue substrate) the white foam disk (used for reinforcement) from the open cell foam insert, making the pads thinner - which is the opposite you seem to want. Anyway, you could add another layer of similar foam if you wanted to make it thicker. But the lack of the white foam layer was not subtle for bass resolution, relaxed sound etc. Then I found the 009 pads (modded the same way) sounded even better on the 007 Mk1 and also Mk2. That's what I've been using for many years. But on my 009s the original pads were the best, without any mods. I suspect the X9K also has pretty optimal pads (but there the problem might be the assembly pins).

Replacement pads (not originals) rarely sound good with any headphone (I have tested 50+ pairs of ear pads of various sizes), but the 007 and 009 original pads are exceptions, as most headphones sound better with them than with their own pads. Especially after the mods. I use them on a lot of headphones, and they last a long time (10 years old Stax 009 pads still in very good condition) - therefore I consider them among the cheapest, even for total cost of ownership alone, but they also sound better than most other ear pads.

I fully agree with you the best solution would be to have a glue ring for attaching the pads. That ensures the best seal and the best sound IME. You can try that with your setup, simpler and better IMHO.

 

[Svperstar]

I fully agree with you the best solution would be to have a glue ring for attaching the pads. That ensures the best seal and the best sound IME. You can try that with your setup, simpler and better IMHO.

I've been debating trying to track down a 105 mm glue ring or seeing if I can order one, or maybe gluing the pads to the 007.

The reason I wanted bigger pads is mainly comfort but also my experience with the LCD-2. I had my LCD-2 refurbished in 2023, I noticed the fazor dug into my ear because the leather pads were so thin. I preferred my LCD-2 Classic to the LCD-2........then I ordered the latest non-leather pads from Audeze and OMG, forms a perfect seal, better comfort, better bass. The LCD-2 with those pads really have to be heard.

If the driver is pressed up against your ear they don't have enough room to really form a seal and have proper low end.

 

[zolkis]

If the driver is pressed up against your ear they don't have enough room to really form a seal and have proper low end.

I used double sided adhesive stripes cut thin and forming a polygon. It doesn't have to be completely closed, as when on your head, the flaps create a seal.

A critical thing with the 007 Mk1 is to shape the metallic head bands so that the drivers are in the best angle to follow your head geometry. In my case, it looked like more flat up in the middle, angling in more, and the front and rear headbands had slightly different arc. The clamp force was a bit less than middle, that ensured the best deep bass thump and rumble and space. Experiment with that. The optimum is rather narrow - but when you find it, it's like a different headphone, everything is suddenly far better.

You can also do this with the original pads settings, and see how that works out vs. the replacement pads.

 

[number1sixerfan]

The problem is that too many people are running around saying that the minimum spend for an electrostatic amp is $5k and this idea is ridiculous.
I think your stance on the 007 needing serious juice makes sense but it has trickled down in a really unhelpful way; it doesn't help that nearly every non-KG amp gets slammed as hot garbage and a fire hazard immediately.
It's the worst part of the hobby when the gear hierarchy gets set in stone and interferes with people's enjoyment of music (that's not what you are doing).

Emphasis added for the parts I agree with

I think a decent general rule of thumb is to take any extreme takes with a grain of salt. The truth on most things almost always lies somewhere in the middle of extremes, in this hobby or any other.

I had the SRM-1/mk2 for a VERY long time until moving last year, and the T1s currently. And definitely both are enjoyable with a lot of stats (have heard others as well, but these are the two I've owned for a while), and you can clearly get an idea of what any stat can do with them. For me, they're just a bit softer with dynamics compared to the more notable amps, but certainly not "trash" and are very solid overall.

One issue that makes things a bit more complicated with this topic is that dynamics, punch and bass response imo are already major challenges with stats generally, so for people that are coming from planar/dynamic headphones that place a lot of emphasis on those aspects, it can be magnified quite a bit with budget amps compared to higher quality amps. The more notable amps are definitely worth at least exploring if you haven't.

The hardest to drive stats (the 007, SGL Sr. and a select few others) do improve considerably when optimally powered, but they're also few and far between and shouldn't really skew the entire conversation, as others have hinted at.

 

[kevin gilmore]

here is a synthesis of srm-t1s with 1vrms input, 120pf load.
feel free to do your own synthesis with ltspice, or your choice of spice spinner.
600v peak to peak at 1khz stator to ground with 320v power supplies. 1 vrms input
360v peak to peak at 20khz stator to ground with 320v power supplies. 1vrms input
6fq7 tube with 66k plate resistor load.

 

[Kiats]

Starting the week here near the Equator in Asia with an early version of the Sigma, NB. Before Stax moved to the same thinner flat cables that the Lambda NB used as well. Being driven by one of Mjolnir’s Octave v4, which @spritzer kindly did one with a Normal Bias output.

I am tempted to get a tube based Stax energizer with a normal bias output as well.

 

[Svperstar]

here is a synthesis of srm-t1s with 1vrms input, 120pf load.
feel free to do your own synthesis with ltspice, or your choice of spice spinner.
600v peak to peak at 1khz stator to ground with 320v power supplies. 1 vrms input
360v peak to peak at 20khz stator to ground with 320v power supplies. 1vrms input
6fq7 tube with 66k plate resistor load.

Can you do the same with the Stax 717?

 

[KDS315]

here is a synthesis of srm-t1s with 1vrms input, 120pf load.
feel free to do your own synthesis with ltspice, or your choice of spice spinner.
600v peak to peak at 1khz stator to ground with 320v power supplies. 1 vrms input
360v peak to peak at 20khz stator to ground with 320v power supplies. 1vrms input
6fq7 tube with 66k plate resistor load.

Hmm, STAX specifies the SRM-T1S to deliver 60dB amplification (1000x) i.e. 100mVrms = 0.1Vrms leading to 100Vrms output - how does that relate to your simulation results with 1Vrms input at 1kHz? I see 600Vpp / 2 * sqrt(2) = 212Vrms output i.e. 212x amplification (?). Was your simulation the complete T1S schematic with three amplification stages or only a partial simulation?

 

[kevin gilmore]

with virtual pot adjusted to 5% under clipping. spec on t1s is 320v power supplies, except in the real world its 350.
full simulation. too much work, others need to do this. back to your regularly scheduled program. anyone with a scope and
a signal generator can do the same thing with the real item. the last time i did this, the result was much worse than simulation.

 

[tabness]

I just wish we could move on from this artificial 20 kHz full power bandwidth at max output voltage metric had and go and have some simulated/measured slew rate distortion at output voltages required for like 75-85 dB with maybe 5-20 dB peaks at something around 6 kHz (also at 3 kHz)...

The guy behind the Nectar headphones did some measurements at the usual 1 kHz and 20 kHz, but at least he also include 10 kHz, and rightfully called out that "[f]or music, typically the higher frequency components of musical content are at a much lower amplitude than the lower frequency content therefore this is more of a stress test than a realistic scenario."

https://nectarsound.net/wp-content/uploads/2020/09/Simple-Measurement-Analysis-Rev001.pdf