[AudioCats]

I was browsing on allelectronics and saw these mica caps:
0.03uf/600v working voltage, mica, 1.75"x1.32"x0.42"

it looks to be new-old-stock, only $2 each, and is thin enough to be fitted into the case. (probably need to drop the top board, add 1/2" stand-off's and add a third "attic" board to hold the mica caps. Will be a tight fit but likely do-able.

update, 3-33-2021:
bought one piece to get a feel of it, the 0.42" thickness is just the main body, the threaded brass posts increase thickness to 0.46", then add wires to the brass posts, then add insulation layer over.... will likely be too thick.


so far the "two 0.01uF/600v Theta in parallel" route appears to be the easiest way to go.

 

[AudioCats]

impressions of opa627/637 as opamp
-- 8x opa627 in bottom board (they need to be 627 since some are working as buffers)
-- 4x opa637 in top board
-- all opamps have at least 150 hr operating time, should have settled down enough
-- signal caps are RTX 0.022uf/600v
-- opamp voltage at +/-12v (provided by add-on DC-DC converter)
-- listening using ESP950
Sound is clean and liquid, but lean and on the bright side. Showing emotion well, maybe by simply provide good details, but not as involving, to me, as the "AD8397 + opa637 +RTX" combination.

Impression, AD8620
-- AD8620 x4 (bottom board)
-- opa637 x4 (top board)
-- AD8620 have about 20 hr at this point
-- Signal caps are RTX 0.022uf/600v
-- opamp voltage at +/-12v
-- listening using ESP950
AD8629 runs cool, the chips don't even get warm, no heat dissipation issue here. Sound is not as liquid as above. also not as bright, and not as detailed. Not as warm as I was hoping for. Showing emotion ok, but not great. Good details, though the sound is a little thick and "chewy" (some might call it "dense", as in "all the details are there as long as you look for it, but that is IF you look for it").

At this point I'd say 8620 is an upgrade from the stock opamps but doesn't have the "transformation" effect like the 8397.
"8397+RTX" does everything ok enough for me. Sound is softer but I can listen to this combo for a long time and not really feeling anything lacking (except the bass impact, but there is no way to overcome that, the output of E90 can only do so much current). On the other hand it is probably just my personal preference, I care about the vocal/emotion part the most, and therefore maybe I tend to prefer the darker sound signature of 8397. And the soft/cushyness of 8397 complements the sharpness of RTX.

so far the ad8620 seems to be the most practical option for a straight-forward opamp upgrade, overall, considering heat dissipation, cost, and sound signature.

 

[ADUHF]

Howdy, AudioCats.

Since you seem pretty knowledgeable on this unit (and things electrical, in general), I wondered if you might have some ideas or possible solutions to this ESP950/E90 owner's humming or ground looping issues...

https://www.head-fi.org/threads/help-i-need-a-scientist.957174/

I'm not that smart on this stuff, but my theory is that his room is too dry, which is causing a buildup of static electricity on the surface of his laminated desk. And this is in some way inducing a hum in the headphones when they are disconnected.

It seems like the amp or maybe it's power supply could be the cause of the static buildup, and/or a residual charge in the headphones which could be making it more sensitive to other electric fields in the room. Does any of this make any sense?

I've experienced similar issues with other headphones (ie not e-stats) in the winter, when humidity has been very low in my apartment.

 

[AudioCats]

what he described doesn't appear to be related to the E90 at all, the problem was there when not connected to an amp:
"...I unplug my headphones and they continue to hum when the cord is in close proximity to the amp/DAC. Touching the amp/DAC makes the hum go away..."

maybe that particular location has strong electrostatic build up or electromagnetic field (radar / satellite stations etc)

 

[miketlse]

Reminds me of issues like this.

He should try earthing his amp or dac, to see if that removes the stray electromagnetic field, that is being picked up by his headphone cord.

 

[ADUHF]

what he described doesn't appear to be related to the E90 at all, the problem was there when not connected to an amp:
"...I unplug my headphones and they continue to hum when the cord is in close proximity to the amp/DAC. Touching the amp/DAC makes the hum go away..."

maybe that particular location has strong electrostatic build up or electromagnetic field (radar / satellite stations etc)

Thank you for the reply, AudioCats. Fwiw, someone also posted this to the thread late last night:

https://www.head-fi.org/threads/help-i-need-a-scientist.957174/page-4#post-16290116

Apart from the issue of the hum also being transmitted when the headphones are unplugged, this sounds like the same problem that I have at my place. Long story short: grounding issue (bad power supply and grounding in my entire home, as it also seems to be the case in yours, as the Koss amp (and your entire audio chain) doesn't seem to be grounded properly). Solution: manual grounding.

To quote the explanation/solution provided by the pro/studio gear manufacturer:
Unfortunately, the customer has a local potential and earthing problem, i.e. the electrical wiring of the house mains. The supplied, unearthed switched-mode power supplies are usually much better than earthed ones because they avoid hum loops. In this case, that is reversed. The secondary side of the power supplies delivers a high-impedance AC voltage that is discharged to earth via the customer's body. This minimal current (of course completely harmless) is then heard in the headphones or in-ear. It really helps to ground the DAC to the current ground potential (heating, water pipe, earthed socket, etc.). Or another earthed device. Screw a cable to the housing and connect it to earth.

Unfortunately I cannot find any photos of your Koss amp's back or rear portion, but connecting a blank wire to one of its screws (if there are any) and the other end of it to a nearby (unpainted) radiator, water line or the likes would most certainly resolve this issue.

This sounds very similar to the type of problems I would sometimes encounter with poorly-grounded gear during the drier winter months in my home.

If the listener is acting as a ground for the headphones (in lieu of proper grounding on the amp), then perhaps that could also allow them to form a weak circuit when the cable comes into contact with other objects in the room that are partially electrified with static charge by the dry air.

I think he's getting the hum though with both the headphones plugged in, and not plugged in. So grounding the amp would seem to be the first step in resolving the issue. If the rest of the room is not well-grounded, then he could still pick up some stray noise here and there from other objects. But at least he could listen to music without the hum occurring, which is probably the important thing.

I wonder if simply putting the amp or it's PSU on an antistatic mat (with a grounding cable) would be enough to ground it. If not, then maybe running a small piece of copper wire from the amp's metal chassis to the mat would allow it to ground via the mat. Maybe there's a simpler way of doing this though. Like a wrist strap that would be large enough to encircle the amp. Unfortunately I just don't know enough about this kind of thing to be much help in this case. (And it's been quite awhile since I put together my last computer.)

He also mentioned that shorting the ends of the headphone cable stops the humming. Which probably means they are retaining some residual charge after being disconnected from the amp. Perhaps this is also due to the amp being not properly grounded?

And maybe a humidifier would also help to reduce some of the electrical interactions with other objects in the room.

Replacing the power supply is also something that I believe has been suggested in various topics on this e-stat combo. Both in the ESP950 threads here on HeadFi, and on Drop, where they sell the ESP95x.

 

[ADUHF]

Reminds me of issues like this.

He should try earthing his amp or dac, to see if that removes the stray electromagnetic field, that is being picked up by his headphone cord.

Thank you also for your reply, miketlse. Grounding does seem to be at least one of the issues. So it seems like we're all pretty much on the same page on that. If you have any further thoughts on any the above, I'd welcome those as well.

Thank you both again for your insights and input on this.

 

[PointyFox]

He also mentioned that shorting the ends of the headphone cable stops the humming. Which probably means they are retaining some residual charge after being disconnected from the amp. Perhaps this is also due to the amp being not properly grounded?

It's just because they're electrostatic headphones.

Replacing the power supply is also something that I believe has been suggested in various topics on this e-stat combo. Both in the ESP950 threads here on HeadFi, and on Drop, where they sell the ESP95x.

I've tried 3 different power supplies and there is no change.

 

[AudioCats]

you are using an all battery powered audio system. Ground one of the components in the chain and see what happens.

again, it doesn't seem to be an E90 related issue.

 

[AudioCats]

E90 with teflon stax socket, from Mjolnir Audio

hmm, Spritzer has some E90 fitted with teflon stax socket, $130 is a good price.... shipping can't be that much if it is just the amp without the heavy transformer.

 

[ADUHF]

Thank you again for trying to help out a bit on this, AudioCats. I was hopin there'd be a rather quick and easy solution. But I guess that's not gonna be the case. So I apologize for taking your thead off its original topic for this.

 

[AudioCats]

that was an interesting problem, I still think it might be related to his particular location.

A non-grounded fully floating system can get back ground noise, my SRM-1 (non mk2) doesn't ground through the power cord, when I use a portable signal source with it there is very low hum in the back ground, grounding the amp or the source gets rid of the hum. Interestingly, the same signal source connected to the E90 doesn't have any hum even when not grounded (full system floating, fully battery powered).

 

[AudioCats]

wrapping up the E90 experiments.

** all OPA627 in signal chain, after >150hr of running: same impression as before this extended burn in.

*** mini heatsinks arrived and installed to AD8397 modules. The whole assembly is now pretty heavy, in real-life use it should be soldered to the PCB. IC sockets will not hold the pins secure enough, the module can get knocked out of socket if the amp is dropped.

And this extra-complex add-on is absolutely necessary, even with the mini heatsinks the module is hot, measured 75C at the heatsink (I would guess the opamp core is probably more like 100C ), and this is with open top. The upper-board opa627 opamp (the top chip) measured around 60C. Room temp around 28C.

 

[AudioCats]

FT-1 teflon 2200pf 600v // Miflex copper 0.022uf 600v

Got the FT-1 mini teflon caps and did an AC burn-in for a few days.

2200pf FT-1 parallel to 0.022uf Miflex copper cap, by-pass ratio 1:10.

Impression, comparing to RTX:
--- lower board using AD8397 x4 , upper board using opa627x4; driving ESP950

RTX: more bite, more excitment. Sound is more open with better "live" feel, but more sibilence tendency. Sound is not as rich, it is more light/liquid.

Miflex//FT-1: warmer and thicker, fuller bass, fuller sound overall, not as liquid and live as RTX, basically no sibilance, Narrower stage, more layback. might be a little dull.


The RTX still has more magic, the AD8397+ opa627+RTX combo is still the most impressive so far, it is quite a transformation from the sterile/emotionless sound of a stock E90. For short listening (3~5 songs at a time) this combo is hard to beat.

For longer sessions, the miflex//FT-1 wins, it is fuller, more easy-listening, and the "Miflex//FT-1" sound is more in agreement with my larger amps, especially the 007t.

This is the end of my E90 experiments spree. Much time and $$ later, the end result is good but not special, I'd say. It easily beats transportable stax amps such as my srm-212, in some aspects it can beat smaller desktop amps such as my modded srm-1 (which I rate higher than a SRM-1Mk2), but is it really worth the trouble to mod a E90? After all its only real advantage (over a proper desktop amp) is the power requirement, making it more suitable for a RV/off-grid setup.

But these little tests scratched my "need to know" itch. Hopefully the info in this thread can help scratch some of the curiosity itch you might have, as well.


Have fun

 

[AlexmodNew]

Hmm. It looks like oscilating occurs at AD8397 in dip to soic sockets.. They must not to get warm without signficiant load.
Better than equip it with heatsinks you must to solder some ceramic bypass capasitors 0.1uf just at supply pins of AD8397 on soic to dip boards and connect it to ground with short cables. If it doesnt help at all the chips you can try to remove any capacitors from feedback of that OA. (seen some small value caps on a schematic above)