[adeadcrab]

I wish Vesper made earpads without leather, just Alcantara.. maybe they'd make one custom for me.

 

[catscratch]

Makes sense. There isn't a lot more that can be done to improve the headphones.

I dunno about that. They could, for instance, add dust covers to the drivers so they wouldn't squeal after a few months, improve the earcups so that they feel like they cost more than $5, and create a new mechanism for attaching earcups to headband that doesn't break when you look at it sideways.

When you have a lifetime warranty, it probably makes financial sense to solve chronic problems that cause your customers to use said warranty over and over, and which have been known about for almost 3 decades now.

 

[Vesperaudio]

Those are the ones that cost $90.

Is is cheaper if ordered directly with custom parameters, but it takes time.

@adeadcrab
with ease

 

[barondla]

I dunno about that. They could, for instance, add dust covers to the drivers so they wouldn't squeal after a few months, improve the earcups so that they feel like they cost more than $5, and create a new mechanism for attaching earcups to headband that doesn't break when you look at it sideways.

When you have a lifetime warranty, it probably makes financial sense to solve chronic problems that cause your customers to use said warranty over and over, and which have been known about for almost 3 decades now.

Good points @catscratch . There are "structural" things Koss can improve upon. Michael Koss Jr. said they have been improving construction in various ways. While the headphones are light weight, they seem pretty durable. Haven't seen many reports of broken headbands or frames in this or the other thread.

Was referring to sound quality, where the most room for improvement seems to be. More sophisticated output circuitry, with a stronger and cleaner linear supply could do wonders. Believe it will happen in the future.

 

[JimL11]

Good points @catscratch . There are "structural" things Koss can improve upon. Michael Koss Jr. said they have been improving construction in various ways. While the headphones are light weight, they seem pretty durable. Haven't seen many reports of broken headbands or frames in this or the other thread.

Was referring to sound quality, where the most room for improvement seems to be. More sophisticated output circuitry, with a stronger and cleaner linear supply could do wonders. Believe it will happen in the future.

I haven't ever seen a schematic of the Koss ES950 amp, however according to Kevin Gilmore, it has a very high potential voltage swing with +/-600V supplies (the Stax SRM T2 only has +/-500V supplies), but very low standing current of 2mA per channel. By contrast the Stax SRM-T1/006, which is their lowest powered amp, has +/-320-350V supplies but 9.7mA standing current per channel. The KG Carbon has +/-450V power supplies and around 36-40 mA standing current per channel. along with BIG heatsinks (relatively speaking).

This means that the Koss amp will hit current clipping well before it hits voltage clipping. Now I understand why they use such a low current - They wanted a light weight portable, battery powered amp, and it's hard to make heatsinks out of plastic. But at least to me, a desktop amp with a more moderate voltage and higher standing current - which implies heatsinks, which implies quite a bit more weight, is a more satisfactory overall solution. A linear supply would also add weight from the transformer, so we are talking about a 10 lb amp instead of a 2 lb amp. Probably higher cost as well - heatsinks and bigger transformers cost money, and higher shipping weight is more expensive as well.

 

[kevin gilmore]

ugly but correct schematic, ac coupled with feedback around opamp.

 

[barondla]

ugly but correct schematic, ac coupled with feedback around opamp.

Thanks for the schematic @kevin gilmore. I've glanced at your site and see some beautiful amps for Stax. With the renewed interest in the Koss have you considered offering an amp with built in Koss connector? That would save us from buying an adapter cable. Thanks.

 

[timb5881]

I haven't ever seen a schematic of the Koss ES950 amp, however according to Kevin Gilmore, it has a very high potential voltage swing with +/-600V supplies (the Stax SRM T2 only has +/-500V supplies), but very low standing current of 2mA per channel. By contrast the Stax SRM-T1/006, which is their lowest powered amp, has +/-320-350V supplies but 9.7mA standing current per channel. The KG Carbon has +/-450V power supplies and around 36-40 mA standing current per channel. along with BIG heatsinks (relatively speaking).

I haven't ever seen a schematic of the Koss ES950 amp, however according to Kevin Gilmore, it has a very high potential voltage swing with +/-600V supplies (the Stax SRM T2 only has +/-500V supplies), but very low standing current of 2mA per channel. By contrast the Stax SRM-T1/006, which is their lowest powered amp, has +/-320-350V supplies but 9.7mA standing current per channel. The KG Carbon has +/-450V power supplies and around 36-40 mA standing current per channel. along with BIG heatsinks (relatively speaking).

This means that the Koss amp will hit current clipping well before it hits voltage clipping. Now I understand why they use such a low current - They wanted a light weight portable, battery powered amp, and it's hard to make heatsinks out of plastic. But at least to me, a desktop amp with a more moderate voltage and higher standing current - which implies heatsinks, which implies quite a bit more weight, is a more satisfactory overall solution. A linear supply would also add weight from the transformer, so we are talking about a 10 lb amp instead of a 2 lb amp. Probably higher cost as well - heatsinks and bigger transformers cost money, and higher shipping weight is more expensive as well.

I would think their best move would be to retain the E90 and offer an upgrade amp as an option or sell the newer better amp as a stand alone. Price of entry would remain lower, with the option for a better amp. Another option might be to license an amp to others so that they could do the upgrade with a sanction from Koss.

 

[fjf]

I know nothing about electronics, but there is one thing I never understood. Because the electrostatic drivers are essentially capacitors, once they are charged (a few seconds after plugging) they require very little current to change the voltage of the stators. Wouldn't the Koss voltage-driven design make more sense that the current-driven staxes?. The plastic E/90 looks like is worth very little, but it sounds surprisingly good, despite its limitations. I've never heard it clipping, BTW. It has power to drive volume to insane levels.

 

[JimL11]

I know nothing about electronics, but there is one thing I never understood. Because the electrostatic drivers are essentially capacitors, once they are charged (a few seconds after plugging) they require very little current to change the voltage of the stators. Wouldn't the Koss voltage-driven design make more sense that the current-driven staxes?. The plastic E/90 looks like is worth very little, but it sounds surprisingly good, despite its limitations. I've never heard it clipping, BTW. It has power to drive volume to insane levels.

Both the Koss and the Stax are voltage driven. That is, the amp has a relatively low output impedance and varies the voltage of the stators in opposite directions. Since the diaphragm is charged, the varying voltage causes the diaphragm to be attracted to one stator and repulsed from the other in a push-pull manner, which minimizes distortion. The stators represent a capacitor, so in order to change the voltage on the stator, some charge needs to flow from or to the stator from the amplifier. For a given voltage change, the amount of charge that is needed is a constant, however the speed with which the charge needs to be pushed onto, or pulled from, the stators, and hence the current = charge/time, needed rises linearly with frequency. The higher the voltage change, and the higher the frequency, the higher the amount of current the amp needs to generate. So for example, for a typical electrostatic headphone which measures about 100 pf capacitance, at 100VRMS and 1 kHz (usually around 100 dB, which is loud) the amp needs to deliver less than 0.1 mA, however, increase that to 20 kHz and the amount of current needed rises to about 1.8 mA. If we double the loudness to 110 db (very loud), the amount of current needed jumps to almost 18 mA. Note that live rock concert levels have been measured between 100-120 dB. The threshold of pain is around 120-130 dB.

What saves us is three things: 1) the fact that most music power is concentrated in the region below 1000 Hz or so - i.e. long before you get voltage clipping at 20 kHz, you will get voltage clipping at a much lower frequency; in fact, Peter Baxandall and Nelson Pass made measurements that showed that an amplifier that has a slew rate sufficient to amplifier a 6 kHz sine wave undistorted is enough for practically all music reproduction, because at the level that clipping occurs in the midrange, the 20 kHz level is at least 10 dB lower, 2) most people don't listen at peak levels even approaching 100 dB, and 3) most music only rises to peak levels at a few moments, so most of the time the demands on the amplifier are significantly less severe. A few milliseconds of clipping in a piece of music that lasts minutes to an hour or more may pass un-noticed. These reasons are why many people are satisfied with bookshelf speakers at the commonly rated 86 dB/watt/1 meter combined with 60 watts/channel amplifiers, which are capable of producing about 104 dB at 1 meter and less than that at the listener's seat. OTOH, Kevin Gilmore once noted @pabbi1 audibly clipping his DIY T2 (+/- 500V PS, close to 2000 V P-P and 28 mA standing current/channel) on SR007 headphones from the sounds transmitted from the back side of those open phones.

 

[fjf]

Thanks for the detailed response. That makes sense, and because the E/90 has a 2mA output, it can get to loud levels without clipping. Also, seems that a more powerful amp ouputting more current should make the headphones faster. The key point then would be: is this audible?. The only way of knowing is trying it...

 

[Allanmarcus]

Thanks for the detailed response. That makes sense, and because the E/90 has a 2mA output, it can get to loud levels without clipping. Also, seems that a more powerful amp ouputting more current should make the headphones faster. The key point then would be: is this audible?. The only way of knowing is trying it...

I've listened to my Koss headset on JimL's amps; some relatively inexpensive Stax ones that he modded, and his BHSE. The Koss sounds significantly better on the BHSE , and slightly better on the lesser amps. Lesser in this context means ~$1000. The 950 at $500 or less is a steal, and doesn't really require an upgraded amp unless you area really picky and can hear a nat's fart. Of course, "require" is a very relative terms in this "hobby".

 

[adeadcrab]

So, I picked up a linear power supply to replace the stock wall-wart on the ESP 95/X.

(not my pics)

The bad news is that it doesn't fix ground loop, which must be isolated to the energizer itself (edit -- or my mains circuits!). Whether it's only 220AC versions of the energizer or all energizers I don't know.
The good news is that it improves the sound, particularly in stronger bass + detailed treble (esp. cymbals), soundstage / imaging. I went back and forth with both wall wart and LPS a few times. The imaging of low end sound is bigger, wider. Treble has more sparkle. Drums sound more like real drums now, before they were a little lean.

Kinda bummed about that ground loop but there doesn't seem to be anything else that I can do for it.

 

[barondla]

@adeadcrab that is a cool looking power supply! What are the specs? 9 or 12v?

Bummer about the groundloop still existing. Does it hum while playing or only when you touch the E90X? Have you tried any different cable between source and E90X?

 

[adeadcrab]

It's 9V @ 2.5a -- I think 12V would make some nice blue smoke and a dead energizer.

I still need some more time with it (just got it today), but the very immediate impression was prominent low end, and other differences were picked up after more listening.

edit - ground loop might be from the RCA connectors. Still investigating.

final edit because I'm sick of it: my desk is motorized and touching the metal frame or the motor section is the cause of the hum. I've disconnected the desk's motor and there is now only a slight hum which I am chalking up to the energizer just being the energizer. This baseline hum is 10 times the order of magnitude less than before and barely audible. Problem solved.

In short, an LPS provides an improvement to the sound of the ESP 950.