That definitely doesn't look right. I think you have some outside interference. 2nd and 3rd harmoic are present. But the higher stuff is not a harmonic of 1Khz.
The sloping noise floor is also dubious. I'd expect some rise at low frequencies, but overall I would expect it all to be below -120db and closer to -140db.
edit: oh I just realised this is measuring something else so the noise floor comments above probably don't apply. could you post an SE loopback measurement for comparison?
Tony.
The sloping noise floor is also dubious. I'd expect some rise at low frequencies, but overall I would expect it all to be below -120db and closer to -140db.
edit: oh I just realised this is measuring something else so the noise floor comments above probably don't apply. could you post an SE loopback measurement for comparison?
Tony.
Last edited:
This is SE loopback test, looks pretty normal, Full Scale is about 835mVrms, if I turn the output volume smaller, the 2nd and 3rd will be pretty tiny.
The nasty measurement with high frequency interference, don't know where does that come from. But it's in every measurement I attempt to do. puzzling.
Okay, I did the loopback -40dBFS, some high frequency stuff showed up.
Attachments
Last edited:
Loopbacks look fine. How does it go with ARTA just showing the RTA without the 1Khz input signal? I'm assuming the high frequency spikes will still be there.
I found that the best thing to do for input of SE to the focusrite 2i2 was to only use tip and ring. hot to tip ground to ring don't short sleeve to ring. Also if you are doing single ended output do the same (ie no sleeve). There does seem to be some sort of ground loop problem in SE mode, though in my case it just manifests as higher 2nd and 3rd harmonic. Better still is to use an SE to Balanced converter.
Tony.
I found that the best thing to do for input of SE to the focusrite 2i2 was to only use tip and ring. hot to tip ground to ring don't short sleeve to ring. Also if you are doing single ended output do the same (ie no sleeve). There does seem to be some sort of ground loop problem in SE mode, though in my case it just manifests as higher 2nd and 3rd harmonic. Better still is to use an SE to Balanced converter.
Tony.
Last edited:
It looks super, but the interference is there, you can see the 7k , 16k peak, looks like the nasty stuff is just hiding.
Thank you for the nice tip, do you use BAL to SE on signal output as well? or only use the hot signal?
needtubes said:
My loopback measurement with RMAA shows the nasty stuff.
Attachments
Last edited:
Hmm hard to tell if that 7K and 17K is inherent to your focusrite or is being picked up by the leads. Do you get the same with balanced loopback? Regardless of whether it is coming from the card or not, there are quite a few other noise spikes in the high frequency range that are not there in the loopback.
I think I posted a circuit earlier in the thread for what I used for balanced to single ended on the output. Basically a simple circuit using an LM4562 opamp, though experimentation showed it was more a problem with the input than the output for the performance. Most specifically not allowing earth signal from either in or out to go back to ring on the input (from memory).
link to post with balance to se circuit. http://www.diyaudio.com/forums/equi...2-2nd-gen-measurements-whats.html#post4963930
Tony.
I think I posted a circuit earlier in the thread for what I used for balanced to single ended on the output. Basically a simple circuit using an LM4562 opamp, though experimentation showed it was more a problem with the input than the output for the performance. Most specifically not allowing earth signal from either in or out to go back to ring on the input (from memory).
link to post with balance to se circuit. http://www.diyaudio.com/forums/equi...2-2nd-gen-measurements-whats.html#post4963930
Tony.
Last edited:
Hello:
Just noticed this thread and thought some AP measurements might be of interest on the Focusrite Scarlett 2i2 (and RME Babyface Pro).
The sound reinforcement industry where I come from uses the inexpensive 2i2 quite a bit for acoustic tuning. OlegSh is dead right about sweet spots on the gain with these things and I don't have the patience to deal with stuff like that - I want to make measurements without having to think about hardware variables. The RME fulfills this requirement nicely, but at almost 5x the cost. But you also get a much longer useful life and a whole lot less frustration out of RME products. In the long run, I wouldn't be surprised if the RME is cheaper.
The Focusrite Scarlett 2i2 measurements.
The RME Babyface Pro measurements.
Just noticed this thread and thought some AP measurements might be of interest on the Focusrite Scarlett 2i2 (and RME Babyface Pro).
The sound reinforcement industry where I come from uses the inexpensive 2i2 quite a bit for acoustic tuning. OlegSh is dead right about sweet spots on the gain with these things and I don't have the patience to deal with stuff like that - I want to make measurements without having to think about hardware variables. The RME fulfills this requirement nicely, but at almost 5x the cost. But you also get a much longer useful life and a whole lot less frustration out of RME products. In the long run, I wouldn't be surprised if the RME is cheaper.
The Focusrite Scarlett 2i2 measurements.
The RME Babyface Pro measurements.
Why mono?
"Scarlett Solo’s inputs are on the front panel: Input 1 uses a standard 3-pin XLR socket, and is configured to work with microphones of most types; you will probably have a mating XLR male connector on the end of your mic cable. Input 2 uses a ¼” (6.35 mm) jack socket (2-pole when in instrument mode and 3-pole when used as a line input), and is intended to accept signals from an electric or electroacoustic guitar or bass."
I have the Solo 2gen and, i cant see on impulse resopnse mesurements L chanel :/ Only right is showing some scale (sound card mesurements).
Apologies for all the screenshots.
I also have a Scarlett 2i2 (snd gen) and here are some loopback test results for comparison, in case it helps anybody.
All tests were done on Windows 7, sample rate was 192kHz.
Arta settings: FFT: 65536 Window: Kaiser5 Avg: EXP
The Bode vs Phase was at 3dB and was taken with Holm Impulse
All THD measurements were performed at -3dB level.
I used various different programs, and in my experience, I have found that the THD measurement of Arta can sometimes be a bit misleading (the measurements in Arta Steps seem ok though).
I also have a Scarlett 2i2 (snd gen) and here are some loopback test results for comparison, in case it helps anybody.
All tests were done on Windows 7, sample rate was 192kHz.
Arta settings: FFT: 65536 Window: Kaiser5 Avg: EXP
The Bode vs Phase was at 3dB and was taken with Holm Impulse
All THD measurements were performed at -3dB level.
I used various different programs, and in my experience, I have found that the THD measurement of Arta can sometimes be a bit misleading (the measurements in Arta Steps seem ok though).
Attachments
Last edited:
Distortion and S/N measurements at various levels
Further back in this thread (I don't recall which post) it was mentioned that the measurement values with the Focusrite vary according to input voltage.
I did a spreadsheet with input voltages spanning 400mV to 2.8V RMS, verified with a 'scope.
These measurements were taken with RMAA, levels were set to -1dB, sample rate was 96kHz (even though mine supports 192kHz, I find that noise and THD seem optimal at 96kHz).
Source was an Ebay DAC kit (LJM AK4396) via my isolation test box (gain of x2)
I have attached a .zip version of the Excel spreadsheet, which I have tried to reproduce below (without too much success!).
Basically, for THD measurements, it looks as though 775mV is optimal, and for SNR measurements, 2.8V is optimal - any higher and it becomes difficult to adjust the input levels. I also notice that when the input level controls are around the 11 o'clock position, setting the levels is quite difficult, as they jump about with the slightest touch of the controls.
A quick comparison with Arta shows the THD measurements agree with RMAA.
I tried to graph my results, but I'm absolutely hopeless with Excel graphs
Finally, I guess the measurements are limited by my source, however the intention here was to demonstrate a trend.
Further back in this thread (I don't recall which post) it was mentioned that the measurement values with the Focusrite vary according to input voltage.
I did a spreadsheet with input voltages spanning 400mV to 2.8V RMS, verified with a 'scope.
These measurements were taken with RMAA, levels were set to -1dB, sample rate was 96kHz (even though mine supports 192kHz, I find that noise and THD seem optimal at 96kHz).
Source was an Ebay DAC kit (LJM AK4396) via my isolation test box (gain of x2)
I have attached a .zip version of the Excel spreadsheet, which I have tried to reproduce below (without too much success!).
Basically, for THD measurements, it looks as though 775mV is optimal, and for SNR measurements, 2.8V is optimal - any higher and it becomes difficult to adjust the input levels. I also notice that when the input level controls are around the 11 o'clock position, setting the levels is quite difficult, as they jump about with the slightest touch of the controls.
A quick comparison with Arta shows the THD measurements agree with RMAA.
I tried to graph my results, but I'm absolutely hopeless with Excel graphs
Finally, I guess the measurements are limited by my source, however the intention here was to demonstrate a trend.
Code:
I/P vRMS Noise Dyn. Range THD (%) IMD+N (%) Crosstalk IMD+N (swept freqs %)
0.400 -97.9 97.9 0.001 0.004 -90.7 0.0049
0.500 -99.50 99.50 0.0009 0.0035 -88.70 0.0044
0.600 -100.80 100.80 0.0010 0.0032 -88.90 0.0042
0.625 -100.90 100.90 0.0010 0.0032 -88.90 0.0041
0.650 -101.30 101.30 0.0010 0.0031 -88.30 0.0041
0.675 -101.50 101.40 0.0010 0.0032 -87.80 0.0040
0.700 -101.70 101.70 0.0010 0.0030 -88.00 0.0040
0.725 -101.90 101.90 0.0010 0.0030 -88.30 0.0040
0.750 -102.10 102.00 0.0010 0.0030 -88.10 0.0041
0.775 -102.30 102.30 0.0010 0.0029 -87.80 0.0040
0.800 -102.50 102.50 0.0011 0.0029 -87.90 0.0040
0.825 -102.70 102.60 0.0011 0.0029 -87.50 0.0040
0.875 -103.00 102.90 0.0011 0.0029 -87.80 0.0040
1.000 -103.60 103.70 0.0013 0.0028 -87.30 0.0041
1.200 -104.40 104.50 0.0015 0.0028 -87.30 0.0043
1.400 -105.00 105.00 0.0016 0.0028 -87.70 0.0045
1.600 -105.40 105.50 0.0019 0.0029 -88.10 0.0048
1.800 -105.80 105.80 0.0021 0.0031 -89.90 0.0052
2.000 -106.20 106.10 0.0023 0.0033 -89.90 0.0056
2.800 -106.60 106.60 0.0041 0.0057 -104.30 0.0083My unit is pretty noisy, under oscilloscope measureing hot and Gnd, noise Vp-p is nearly 20mV. Switch to 50ohm range, it's much better, about 2mv with lots of spikes. I don't know if that's normal.
In the 2i2 unit, there's a common mode choke type filter on power line, it failed to filter noise, but then I check my PC, USB Shield & Gnd were shorted with 0 ohm resistor. Maybe I'll try linear power supply on it. Just don't know if the noise is from the unit or computer.
In the 2i2 unit, there's a common mode choke type filter on power line, it failed to filter noise, but then I check my PC, USB Shield & Gnd were shorted with 0 ohm resistor. Maybe I'll try linear power supply on it. Just don't know if the noise is from the unit or computer.
Mine is also noisy when I connect a scope - it's because the scope is forming an earth loop with the computer.
I also tried powering mine from a USB injector, and it makes absolutely no difference to the noise levels.
The power supply in the Focusrite is actually very quiet, even with (reasonably) noisy USB ports.
You could try plugging the scope into the same power adaptor as your computer, to keep the earths as close as possible.
Alternatively, try using a laptop is you have one.
Okay, that makes sense, because I don't hear the noise despite it contains low frequency noise as well. I've tried float computer, float oscilloscope, or in the same plug, same result. but loopback test shows pretty beautiful graph.
have you ever open the unit? kind of interesting inside. output connector is filtered by 3 pin filter to earth, even the signal gnd was filtered. when I measure op amp voltage relative to earth, +-5V, relative to signal gnd is +-2.5V. Don't know if I remember it right, but it puzzles me.
Your circuit works super duper, it eliminates those weird spikes. can't thank you enough.