• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

ESI [email protected] XTe Sound Card

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
1,889
Likes
1,950
Location
Pacific Northwest
#1
This was a popular card 10-15 years ago. I've had mine about that long. It's unique in that it has a daughter board that you can unplug & flip around to switch between single ended (RCA) & balanced (1/4" TRS) outputs and inputs. It uses the AK4358 for DA and AK5385A for AD. It supports PCM up to 24-bit 192 kHz, both input and output. For outputs, it also two digital: coax and toslink.

Summary: it was a great card for its time, still decent but nowhere near state of the art.

I tested it by playing REW frequency sweeps through a loopback connector. I'll show some unexpected results and ask questions.

Here's distortion & noise, SE, at 48 kHz, sweep at -0.5 dB digital:
julia-THDN-48-se.png

Nothing special or surprising above.

Below is the same thing in balanced mode:
julia-THDN-48-bal.png

This is surprising: higher noise, about the same distortion. If you squint you can see the 2nd harmonic (light orange) is lower than SE, as expected with balanced.

Question 1: why is noise higher in balanced mode? Balanced should have lower noise!

Here's distortion in SE mode, with sweeps at 0, -1 and -3 dB.
julia-dist-48-se-0-1-3.png

Nothing surprising above. Distortion is slightly lower as you reduce the digital signal level below max.

Here's the same thing in balanced mode:
julia-dist-48-bal-0-1-3.png

Well that is different: distortion is relatively constant at 0, -1 and -3 dB.

Question 2: why does distortion vs. level react differently in SE vs. balanced mode?

To help with this question, here are direct comparisons of distortion (not noise, couldn't get that to work in REW) between SE & balanced, at different signal levels. In each graph, SE is black and balanced is blue:

Sweep at 0 dB:
julia-dist-48-se-bal-0.png

Balanced has slightly lower distortion up to about 1 kHz, above which it rises and SE is slightly lower.

Sweep at -1 dB:
julia-dist-48-se-bal-1.png

Above we see the same pattern, though SE distortion dropped a bit while balanced remained the same.

Sweep at -3 dB:
julia-dist-48-se-bal-3.png

Here, balanced has lost its advantage and SE has lower distortion across the board.

Sweep at -30 dB:
julia-dist-48-se-bal-30.png

This is almost identical to the situation at -3 dB, but of course the levels are higher overall due to the lower level signal.
 

Old Listener

Senior Member
Joined
Apr 28, 2016
Messages
382
Likes
382
Location
SF Bay Area, California
#2
I used this card for years. First with analog outputs and then using coax and Toslink digital outputs. The driver software was more functional than that supplied with some other PCI sound cards.
 
OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
1,889
Likes
1,950
Location
Pacific Northwest
Thread Starter #3
I'm using it on a Linux (Ubuntu 18) machine. No driver or other software needed. Plug and play with the standard ALSA/Pulseaudio Linux sound system.
 

Matias

Major Contributor
Forum Donor
Joined
Jan 1, 2019
Messages
2,518
Likes
4,052
Location
São Paulo, Brazil
#4
OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
1,889
Likes
1,950
Location
Pacific Northwest
Thread Starter #5
So does anyone know why balanced mode has slightly higher noise & distortion? That's the opposite of what I expected.
 

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
1,654
Likes
1,526
Location
.de
#6
Has to be down to specific implementation details. I have a feeling that 0 dBFS levels in and out may be different for one, do they line up? Both are supposed to be +20 dBu (balanced) / +6 dBV (unbalanced) though, so not sure. I won't be too surprised if this modular concept (as genius a design it may be) is a bit prone to ground-related (and possibly contact) issues... I can't see it being particularly AES48 compliant either. There is no good reason for balanced showing much more in terms of power supply harmonics otherwise.

A bit of reverse-engineering might shed some light on the matter. Hmm, so just where did have have that old [email protected] floating around again...?
 
OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
1,889
Likes
1,950
Location
Pacific Northwest
Thread Starter #7
Yes, that noise looks like power supply harmonics. But... wouldn't power supply harmonics be equal on + and - (e.g. a ripple on the DC), so balanced mode differential signalling should cancel it?
 

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
1,654
Likes
1,526
Location
.de
#8
That's what one would think. I suspect there is something fishy going on whenever you make a balanced loopback connection in particular. I would attempt to reverse-engineer the balanced receiver circuit, and if nothing obvious is found there, look at the grounding if possible. Since the circuit has both balanced and unbalanced inputs it is also possible that part of the signal path is actually unbalanced. I can't imagine that routing this analog board would have been much fun to begin with, there may well be some surprises lurking within.
 

bennetng

Major Contributor
Joined
Nov 15, 2017
Messages
1,321
Likes
1,267
#9
This was a popular card 10-15 years ago. I've had mine about that long. It's unique in that it has a daughter board that you can unplug & flip around to switch between single ended (RCA) & balanced (1/4" TRS) outputs and inputs. It uses the AK4358 for DA and AK5385A for AD. It supports PCM up to 24-bit 192 kHz, both input and output. For outputs, it also two digital: coax and toslink.

Summary: it was a great card for its time, still decent but nowhere near state of the art.

I tested it by playing REW frequency sweeps through a loopback connector. I'll show some unexpected results and ask questions.

Here's distortion & noise, SE, at 48 kHz, sweep at -0.5 dB digital:
View attachment 130764
Nothing special or surprising above.

Below is the same thing in balanced mode:
View attachment 130765
This is surprising: higher noise, about the same distortion. If you squint you can see the 2nd harmonic (light orange) is lower than SE, as expected with balanced.

Question 1: why is noise higher in balanced mode? Balanced should have lower noise!

Here's distortion in SE mode, with sweeps at 0, -1 and -3 dB.
View attachment 130766
Nothing surprising above. Distortion is slightly lower as you reduce the digital signal level below max.

Here's the same thing in balanced mode:
View attachment 130767
Well that is different: distortion is relatively constant at 0, -1 and -3 dB.

Question 2: why does distortion vs. level react differently in SE vs. balanced mode?

To help with this question, here are direct comparisons of distortion (not noise, couldn't get that to work in REW) between SE & balanced, at different signal levels. In each graph, SE is black and balanced is blue:

Sweep at 0 dB:
View attachment 130769
Balanced has slightly lower distortion up to about 1 kHz, above which it rises and SE is slightly lower.

Sweep at -1 dB:
View attachment 130770
Above we see the same pattern, though SE distortion dropped a bit while balanced remained the same.

Sweep at -3 dB:
View attachment 130771
Here, balanced has lost its advantage and SE has lower distortion across the board.

Sweep at -30 dB:
View attachment 130772
This is almost identical to the situation at -3 dB, but of course the levels are higher overall due to the lower level signal.
You recently discovered this, or it has been like this since new? I wonder if it is caused by ageing or not.
 
OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
1,889
Likes
1,950
Location
Pacific Northwest
Thread Starter #10
The former. I've had the card for a long time, but never before compared SE vs. Balanced measurements.
 

bennetng

Major Contributor
Joined
Nov 15, 2017
Messages
1,321
Likes
1,267
#11
The former. I've had the card for a long time, but never before compared SE vs. Balanced measurements.
Try this:
I had a problem that several USB 3 thumb drives were only able to run at USB 2 speed on my PC (both front panel on the case and sockets on the motherboard) after years of use. I was not sure if it is caused by the thumb drives themselves or the motherboard, or driver issues. At the end I figured out it was caused by invisible dust and dirt on the sockets. Applied contact cleaning fluid on either the socket or the plug, re-plug several times and problem solved LOL. Not that sensitive with USB 1 or 2 devices.
 

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
1,654
Likes
1,526
Location
.de
#12
Can confirm. The front panel connector in my PC has rather loose-fitting contacts and the USB (2.0) is starting to act up again. I also revived a USB 3.0 thumb drive that would only work in 2.0 (and not at all in 3.0 ports) with contact cleaner once.

Still haven't come across the stupid card btw, I know it's around here somewhere and there aren't actually too many places it could be.
 

bennetng

Major Contributor
Joined
Nov 15, 2017
Messages
1,321
Likes
1,267
#13
Can confirm. The front panel connector in my PC has rather loose-fitting contacts and the USB (2.0) is starting to act up again. I also revived a USB 3.0 thumb drive that would only work in 2.0 (and not at all in 3.0 ports) with contact cleaner once.

Still haven't come across the stupid card btw, I know it's around here somewhere and there aren't actually too many places it could be.
Perhaps you are right about the previous replies. The test on iXBT shows the exact issue, and yes, the PCIE version.

https://prosound.ixbt.com/hardsoft/esi/[email protected]/2444.shtml
https://prosound.ixbt.com/hardsoft/esi-julia-xte.shtml

My X-Fi PCIE card (of course, unbalanced) does not have these things show up.
https://www.audiosciencereview.com/...nts-of-evga-nu-audio-pc-card.9137/post-235666

Full RMAA report:
https://www.audiosciencereview.com/forum/index.php?attachments/x-fi-titanium-hd-zip.29512/
 

preload

Addicted to Fun and Learning
Forum Donor
Joined
May 19, 2020
Messages
942
Likes
1,080
Location
California
#14
Back in the day, I knew this card was special because the pcb was white instead of green.
 

Similar threads

Top Bottom