KGSS Mod(s)

[bobkatz]

It seems to me adjusting the range on the attenuator to match what need makes more sense than changing the gain on an amp that isn't readily set up for that.

Let's just say you're right, that's more convenient, but making your own 1 dB/step attenuator out of a 24-step switch is more rewarding. You sing your mantra as you work through the weekend sorting and installing Dale resistors. I generally wouldn't change the gain structure or feedback on the amp, but either attenuating the source (with a control on the DAC) or adding an attenuator at the input terminals is legitimate. ).

BK

[luvdunhill]

Let's say your right. However, that's not nearly as rewarding as say, building your own ES amp

[mypasswordis]

Yup he's not changing the gain of the amp, he's attenuating the input signal.

Bob you may still want to put the thing under an oscilloscope as adding a pole may have caused further oscillations (listening for it is not a good way of doing it).

[bobkatz]

Yup he's not changing the gain of the amp, he's attenuating the input signal.

Bob you may still want to put the thing under an oscilloscope as adding a pole may have caused further oscillations (listening for it is not a good way of doing it).

Right, I'll take a further look at it! But I'd first like to hear from others who have the latest iteration of the KGSS and can measure its response to confirm this is not an anomaly.

Bob

[spritzer]

Let's face it, Spritzer, you like to listen loud. ;-). We could get into measurements of input level versus SPL, etc., but it's not important. I strongly feel the KGSS has excessive gain with U.S. or Professional level outputs. Either way we can agree to disagree on that. And clearly, for those who wish to try the Katz "eq mod", you have to attenuate to create the RC network. As for whether the problem with the Stax is due to the Stax or the KGSS (or the particular KGSS that I have) why not say that it can be due to both! If the amp does not measure flat, that is a problem and as far as I'm concerned the first place to tackle any perception of HF loss. We all need one or two of you to please carefully measure and either confirm or refute my measurements before I/we do further investigation. As an experienced audio engineer I know how to measure frequency response and I would not have posted if I was not 100% sure of my measurements. The only thing I did not do was measure frequency response at the input terminals to see if there is a loss in the cabling or a capacitive rolloff due to the input impedances involved.

I would have to say the amp is pretty much flat, as much as it can be into the SR-007 load. You won't ever get the bass performance of the Blue Hawaii but it is a far cry from clearly colored amps like the Rudistor Egmont and the Exstata.

The bass issue intrigues me! I did read about the port mod but concluded that people were saying the port was to fix the "Stax Fart" and it looked to me that sealing the port was not to alleviate any sonic disorders. I'll certainly investigate and try the port mod when I get a chance. I'm quite worried about the procedure to put the earpads back due to that spring, never having tried opening these phones. Does anyone have a link to the instructions for earpad takeapart? Will Mortite work to seal this port? http://www.amazon.co...e/dp/B000LNODSQ

As with any planar transducer, baffle leak will alter response quite a bit. The port is just there to eliminate the noises the drivers make because they are virtually airtight. Ditto for the increased earpad height as some users had ears that protruded too much.

The spring isn't really a problem. You just set it in place (there are flanges inside the earpads where the spring sits) and you plug it into the rubber grommet in the center of the protective plate. Then you just grab the edge of the pads and feed it into the slot it sits in, while keeping one end firmly in place. Once you are done you rotate the pads a few times to make sure they sit properly.

That Moritite might work but I haven't tried it. I like Blutac for this as it leaves no residue and you can find it anywhere.

[kevin gilmore]

forgot about that, KG's ancient pictures are too fuzzy to see if he snuck in any compensation caps

also, the 2SA1156 was changed to 2SA1486 because the 1156 was running too close to its 400V limit, though that was really more of an issue on the Blue Hawaii.

i'll give you ancient and fuzzy pictures... I don't even know what i had for a camera back then, but

it had to be grim by today's standards. Here are some fresh ones.

http://gilmore.chem.northwestern.edu/kgssorig1.jpg

http://gilmore.chem.northwestern.edu/kgssorig2.jpg

http://gilmore.chem.northwestern.edu/kgssorig3.jpg

no compensation caps on the top. And none on the bottom either, but its hard to take

that off because the power suppy is screwed to the bottom.

Will measure again, probably tomorrow or wednesday.

Edited January 24, 2011 by kevin gilmore

[bobkatz]

i'll give you ancient and fuzzy pictures... I don't even know what i had for a camera back then, but

it had to be grim by today's standards. Here are some fresh ones.

http://gilmore.chem.northwestern.edu/kgssorig1.jpg

http://gilmore.chem.northwestern.edu/kgssorig2.jpg

http://gilmore.chem.northwestern.edu/kgssorig3.jpg

no compensation caps on the top. And none on the bottom either, but its hard to take

that off because the power suppy is screwed to the bottom.

Will measure again, probably tomorrow or wednesday.

Well, if Kevin's prototype measures flat (into an open circuit, that's how I measured mine), we have to wait till someone else measures a current issue KGSS with its current PC board layout to know 100% for sure. Does the headphone load produce a resonant circuit? I have a friend in Belgium who just had a KGSS built by Justin and bought an O2 at my recommendations and and who has the laboratory gear to make accurate measurements (he works at Galaxy Studios) so when he gets the amp he'll measure it for us. I know that Robin is reading this thread, so since he has the "standard" (DACT?) input attenuators, I'll ask him to measure the response (balanced) at the input terminals as well since there is a difference there, should not affect response but we have to be scientific about it.

BK

[bobkatz]

I should really move this question over to the port mod thread.

As with any planar transducer, baffle leak will alter response quite a bit. The port is just there to eliminate the noises the drivers make because they are virtually airtight. Ditto for the increased earpad height as some users had ears that protruded too much.

Spritzer, are you saying that the perceived "porty quality" of the phones, which I perceive as extra energy below about 60 Hz (probably lower, more like 40) will be tamed a bit by sealing the port?

[spritzer]

I should really move this question over to the port mod thread.

Spritzer, are you saying that the perceived "porty quality" of the phones, which I perceive as extra energy below about 60 Hz (probably lower, more like 40) will be tamed a bit by sealing the port?

Yup, changing the angle of the arc on the headphones and finding the correct earpad angle also helps but the port is the biggest offender here.

[bobkatz]

Yup, changing the angle of the arc on the headphones and finding the correct earpad angle also helps but the port is the biggest offender here.

Right you are. I just sealed the ports with Blu Tack today. It tightens the bottom end very nicely. Shame on Stax for having porting the Mk2 and thank you Spritzer for discovering the anomaly. I imagine they had gotten a lot of complaints from unsophisticated users about the "fart". In fact, the seal on my head is so tight that I get quite an annoying snappy fart when placing the cans on my head and/or pushing down even slightly on the sides of the phones. But I'll live with it for the more accurate bass that the sealed port produces.

As a result of the tighter bass I'm considering taking down the 20k about 1/2 dB by changing the 270 pf cap to 180 pf. With some sources the high end on the O2's is a bit "high fi" but pleasantly so with 270 pf, which gives an amp boost of 1.3 dB at 20K. I think perhaps an amp boost 1/2 dB less (0.8 dB boost at 20K) would complement these phones very well.

I wonder if my SR5-Golds have a port? They really don't fart much at all.

[spritzer]

When I get my hands on another Mk2/A I plan to see if there is a middle ground where air can escape but not utterly fuck up the sound at the same time.

No port on the SR-5N in this sense but the drivers aren't properly secured to the baffle and the earpads do not seal completely. My SR-1 farts like crazy when fitted with SR-007 Mk1 earpads but the drivers are also glued to the baffle.

[Craig Sawyers]

When I get my hands on another Mk2/A I plan to see if there is a middle ground where air can escape but not utterly fuck up the sound at the same time.

No port on the SR-5N in this sense but the drivers aren't properly secured to the baffle and the earpads do not seal completely. My SR-1 farts like crazy when fitted with SR-007 Mk1 earpads but the drivers are also glued to the baffle.

Original (low bias) SR-lambdas were the same - the baffles and electrostatic units were a single glued assembly.

[spritzer]

Yup, all of the Lambdas until this latest series had the drivers glues to the baffle.

[kevin gilmore]

So i got around to measuring the prototype.

It should be mentioned that there is no attenuator of any kind

in the prototype, and any standard potentiometer setup will

induce an automatic RC filter at the input due to the capacitance

of the input fet.

So at 400 volts peak to peak, stator to ground, at 20khz

the left channel is .12 db down, and the right channel is .14 db

down, and both are down about 3db at about 45khz.

Measured with a signal generator with a 50 ohm output impedance.

Reference 1khz

Edited January 30, 2011 by kevin gilmore

[luvdunhill]

very nice!

[mypasswordis]

Good news, we can carry on blaming roll-off on the headphones now! I take it this was measured with the load?

[mypasswordis]

wut dbl post

Edited January 29, 2011 by mypasswordis

[kevin gilmore]

140 pf load. Which is the 120 pf of the O2's and 20pf for the

cable.

[bobkatz]

So i got around to measuring the prototype.

It should be mentioned that there is no attenuator of any kind

in the prototype, and any standard potentiometer setup will

induce an automatic RC filter at the input due to the capacitance

of the input fet.

Reference 1khz

Dear Kevin:

Right. Well, there is always some form of RC filter, often due to the input wiring capacitance and any medium-to-high impedance input attenuator. I assumed my input wiring capacitance is negligible being of very short length. Do you think my measurements of (if I recall correctly) 1.8 dB down at 20 kHz can be explained by the input attenuator/potentiometer circuit I also posted? If so, then what you are saying is that automatically ANY mid-to-high-impedance input level control will introduce considerable HF rolloff in the KGSS. If so, then a slight advance HF compensation is not out of order. I just settled on four hand-matched 200 pf caps as sounding correct in my front end circuit with the O2 Mk2/Spritzer port modded!

What is the input capacitance? It might be possible to design a 600 ohm (2 kohm if possible) input variable input attenuator circuit that would not be significantly affected by the FET's capacitance. This would be the best way to go about designing an attenuator so not to need an eq circuit.

BK

Edited January 30, 2011 by bobkatz

[bobkatz]

140 pf load. Which is the 120 pf of the O2's and 20pf for the

cable.

How significant is the load capacitance to the measured frequency response of the KGSS?

[luvdunhill]

Depends on the amplifier the amplifier is made to drive a target load though, just like a dynamic amplifier has given design constraints and goals. In the case of an ES amp, the load is the primary determiner of FR.

Edited January 30, 2011 by luvdunhill

[Craig Sawyers]

What is the input capacitance? It might be possible to design a 600 ohm (2 kohm if possible) input variable input attenuator circuit that would not be significantly affected by the FET's capacitance. This would be the best way to go about designing an attenuator so not to need an eq circuit.

Well, the K389 has a Ciss of about 30pF. Logically that is about right, since it is a low noise small signal JFET. So to roll off at a 3dB point of 25kHz would need a series resistance of 200k.

[bobkatz]

Good news, we can carry on blaming roll-off on the headphones now!

ONLY if you are not using a passive attenuator at the input to the amp of too high an impedance. (.

If you are, I think the safest thing for you is to measure the output frequency response with your attenuator at various positions. Here are my current conclusions after my most recent round of mods: 1/2 dB of loss at 20 kHz with modified O2s (port sealed) sounds just right to my ears. I've listened to a wide variety of material that I'm extremely familiar with and it varies equally on either side of neutral when the 20K is -1/2 (no load). To my ears, the 1.8 dB of loss that I initially measured is probably excessive. Is this loss due to the approximately 10K of source resistance I'm presenting right at the input terminals? I'd have to test, or if Kevin would be kind enough to calculate.

Kevin probably nailed the reasons for potential rolloff to be the input capacitance of the FET front end. In that case, it should be made clear to buyers that the KGSS is susceptible to such problems and either an attenuator of such and such impedance be used, or that it be driven by a low impedance source and a buffered volume control should be used. While some engineers I know are fond of using 2K impedance attenuators (volume controls), I've found that it seems to "choke" all the sources I've tried. Even discrete transistor circuits like to breathe, to my ears; they sound better with bridging loads, even if they are spedced to drive low impedance loads. I learned it the hard way, by actual experiment, that 600 ohm load is definitely too low, that 2K load is too low sonically, and that somewhere above that (10K is safe) is the minimum load for all the various driving circuits I've tried. This includes 5532s as well as esoteric discrete opamps with dc servos, such as Forssell and Dave Hill's excellent circuits.

This leaves me with an interesting choice. I already designed this calibrated attenuator into my KGSS. Perhaps the simplest thing is for me to add a buffer stage if it is really true that the KGSS's input FET has a fairly high capacitance. Kevin?

BK

[kevin gilmore]

About 25pf for the fet is right. Add in wire and connector

and its 30 pf for sure. There is likely nothing left available

in solid state that is going to do any better and still have

an extended frequency response.

I run a 50 ohm system with 50 ohm correctly terminated cables.

The krell stuff is designed for this. So is the Ayre stuff.

Anything with significantly lower capacitance has to be tubes.

[Craig Sawyers]

I run a 50 ohm system with 50 ohm correctly terminated cables.

The krell stuff is designed for this. So is the Ayre stuff.

That is exactly the way to go - and for sure is the way to remove interconnect cable dependence from sound quality.

Kevin, my man - while waiting with bated breath for your KG-Blowtorch, I'm embarking on a Borbely pre with K216/J79 outputs that drive 50-ohm matched impedance. For precisely that reason.