One of the channels needs the signal path rerouted to clean it up. As is, signal for that channels runs to close to some PS circuitry. I think it was the left... A bit of SMD soldering is needed, but I am not a regular and was able to do it. I'd have to look up this info, but it's on these forums somewhere.
I should also say these mods were for the USB interface, so I'm not sure if PCIe is the same or not.
After the mods, the results are outstanding (noise floor goes in the whereabouts of about -140dB).
Last edited:
There are two kinds of EMU 0404, one is USB the other PCI. The USB has had more close review for measurements. There is a patch for the EMU pci stuff to work in Win 10 and that has worked for me on the 1616m. E-MU 1616m, 1212m, 1820m, 0404 Audio Interface - Complete Windows 10 (1903 & Above) Installation Guide - KVR Audio
For someone starting with nothing the Focusrite USB stuff seems to work pretty well.
For someone starting with nothing the Focusrite USB stuff seems to work pretty well.
The latest version beta software for USB EMU0404 works fine with Win10 for me.
There has a growing level of awareness of Focusrite USB products like the Scarlett Solo over the last few years, due to their low entry cost and 192kHz range - plenty of performance tests on various forums, and the performance seems superb for the cost. If one aim is to try and use a 10:1 or 100:1 scope probe to provide a simple attenuator (which is what I do with the EMU0404) then be careful to find a device with a 1 Meg unbalanced input spec (or close match as the actual attenuation magnitude is not a concern, but the frequency response is), as that makes it dead easy to connect with, and what I use for nearly all my bench probing.
There has a growing level of awareness of Focusrite USB products like the Scarlett Solo over the last few years, due to their low entry cost and 192kHz range - plenty of performance tests on various forums, and the performance seems superb for the cost. If one aim is to try and use a 10:1 or 100:1 scope probe to provide a simple attenuator (which is what I do with the EMU0404) then be careful to find a device with a 1 Meg unbalanced input spec (or close match as the actual attenuation magnitude is not a concern, but the frequency response is), as that makes it dead easy to connect with, and what I use for nearly all my bench probing.
Last edited:
Is this "EmuPMX_PCDrv_US_2_30_00_BETA.exe" (2018.11.28)?
I also use 0404 USB, but my standalone "measuring station" laptop has -14 year old- MS Vista, which is now like as a dinosaur.... and at least so slow.
The Lexicon Alpha is one to avoid. The reason is that the input amplifiers can't take enough level. Example of the fail is a high output radio broadcast mixer buss at line level that will clip this box deeply. The Focusrite Scarlett works fine in this place but I still have my reservations about it. I'd love to open it up and see the works but they have come up with a package that you can't get into without breaking something.
A quick google showed there have been some teardowns - eg. of the first gen
Khron's Cave: #55 Focusrite Scarlett 2i2 (1st gen) teardown & modding
Thanks for that link! The ones I use are for my employer's radio stations and I usually don't have time to solve the puzzle of opening them up. Love it with screws to open and will always have a look. For my own stuff I love to mod for upgrade but most of that has been with the D to A devices or amplifiers. I use to mod lots of the broadcast air chain devices but these days nobody seems to appreciate or care about the chain of audio quality. Also that much of the newer devices have no area for better voltage regulators or film capacitors.
Panel size, it not too late
They are the same size as the plastic inserts that come with the box from the BOM.
171x33x1.5mm from what I can tell
They are the same size as the plastic inserts that come with the box from the BOM.
171x33x1.5mm from what I can tell
I don't know if this has been discussed earlier in this thread but I made an interesting finding regarding the THD of Pete's soundcard interface.
I used Pete's soundcard interface to measure a low distortion SS amp that I had designed. To my huge disappointment I noticed that the distortion at 20kHz (or actually above 10kHz) was highish (0.02%). First I thought it was my SS amp but some additional measurements confirmed that the soundcard interface was the cause. I tried some low distortion opamps (LM4562, LME49860) but no change in THD.
Then finally after some additional measurements I found the culprit. The over-voltage clamping diodes (D5-D8) were the cause. Removing them dropped the THD to 0.0003% at 20kHz!! That is nearly 2 decades better than originally.
This issue has also been discussed here. I assume 1N4148s that are used in Pete's soundcard interface have too high leakage. Next I will try the FDH300s that Jan Didden is using in his AutoRanger.
I used Pete's soundcard interface to measure a low distortion SS amp that I had designed. To my huge disappointment I noticed that the distortion at 20kHz (or actually above 10kHz) was highish (0.02%). First I thought it was my SS amp but some additional measurements confirmed that the soundcard interface was the cause. I tried some low distortion opamps (LM4562, LME49860) but no change in THD.
Then finally after some additional measurements I found the culprit. The over-voltage clamping diodes (D5-D8) were the cause. Removing them dropped the THD to 0.0003% at 20kHz!! That is nearly 2 decades better than originally.
This issue has also been discussed here. I assume 1N4148s that are used in Pete's soundcard interface have too high leakage. Next I will try the FDH300s that Jan Didden is using in his AutoRanger.
Are you aware of anyone putting your Silent switcher board inside the Bud case that Peter Millett selected for his sound card interface? I am considering where to mount it for that Bud case Peter selected (which is pretty small) and USB connector hole is already placed.
D
Deleted member 148505
D
Deleted member 148505
If it is a leakage issue, the Nexperia BAS45A diode has very low leakage (1nA), but one may have an issue with excess junction capacitance (which is voltage dependent). The BAS45A diode has lower junction capacitance than the FDH300 diode.
