Here it is. One for the generator at -1dBFS and the other at -3dBFS.
Attachments
Good catch, I enabled "Lock frequency to FFT" option in the Generator and obtained same level of noise performance as nyt
Both output and input are configured at 0dB, and REW calibrated under this setting before loopback measurements.
Can you show the calibration setting - steps ( screen shots) - what OS are you using?
kannan
I'm using Windows 7, REW V5.19 Beta 7.
For calibration, you just open "Preferences", click "Calibrate" and it'll have on screen instruction telling you what to do. Other settings for my measurement can be seen at the picture at post #1331.
Make sure your select maximum FFT length, it'll have direct impact on the noise floor you'll get. I also suggest you to choose 44.1kHz, and then enable both the high pass and low pass filter (only the latest Beta version has this feature, the stable version does not).
For what its worth the AK4490 has slightly lower distortion on 44.1 family clocks. Its something to do with internal optimizations. Unfortunately most measurement software is focused on 48K sample rate multiples. The QA401, which also uses the AK4490 DAC, has this limitation. It uses a different ADC, AK5397, that trades ultimate distortion performance for extended response (to 768KHz) and supposedly greater SNR. There are tradeouts in all of these choices. Keep in mind these are the same limitations that all recording systems are up against. You won't find real audio sources that are better than the RTX6001.
There is no "balanced" Victor source. However with much effort you can take a battery powered Victor and connect it balanced. You need a common mode reference however, using a pair of resistors to ground can do it. I would suggest 1K. You should also put the whole affair in a separate isolated metal box.
I'm speaking from experience. However I'm not sure which computer has the plots. If I find them soon I'll post.
Hi everyone,
I'm having a problem calibrating the RTX unit with the ARTA software. I'm using it in the full paid-for mode.
When I measure the output (step 1), the reading is 706.33 mV differentially (expected). When I enter this value, ARTA comes back with an estimated peak value of 1,412.66 mV, which is wrong. I'm using the default output level of -3 dB. The meter I'm using is an HP 34401A. I get 0.7063 V using a Agilent U1273A handheld meter.
Any ideas what I might be doing wrong here?
Thanks ARTA users!, Chris
I'm having a problem calibrating the RTX unit with the ARTA software. I'm using it in the full paid-for mode.
When I measure the output (step 1), the reading is 706.33 mV differentially (expected). When I enter this value, ARTA comes back with an estimated peak value of 1,412.66 mV, which is wrong. I'm using the default output level of -3 dB. The meter I'm using is an HP 34401A. I get 0.7063 V using a Agilent U1273A handheld meter.
Any ideas what I might be doing wrong here?
Thanks ARTA users!, Chris
I tried posting on the Diana Thread but I'm not getting through.
Is this the same problem? Are you both getting the same 1.4V reading and becoming confused even though a different software package is being used?
Last edited:
Hi cwtim01,
The peak value for 0.7071 V rms is 1.00 V. The peak value for 1.00 V pk is 1.00 V pk. The peak to peak voltage is 2 V. It looks like the program is multiplying the rms value twice (by 1.414).
This doesn't make any sense to me. If you're right (and you may be), ARTA is dragging another scale in without announcing it clearly. That is, reading from a -3dB scale referenced to 0 dBu, then switching to that scale and calculating the peak value for that. To what end would they be doing this?
Hi Andrew,
Do you have a link to your posts?
Thank you both for your replies. Hopefully someone can shed some more light on this.
The peak value for 0.7071 V rms is 1.00 V. The peak value for 1.00 V pk is 1.00 V pk. The peak to peak voltage is 2 V. It looks like the program is multiplying the rms value twice (by 1.414).
This doesn't make any sense to me. If you're right (and you may be), ARTA is dragging another scale in without announcing it clearly. That is, reading from a -3dB scale referenced to 0 dBu, then switching to that scale and calculating the peak value for that. To what end would they be doing this?
Hi Andrew,
Do you have a link to your posts?
Thank you both for your replies. Hopefully someone can shed some more light on this.
When I calibrate RTX with ARTA, also at the default -3dB(FS) level, I measure 707.8mVrms between pin 2 & 3 (pretty close to your 706.mVrms). Since that is at -3dBFS level, 0dBFS will be multiplied by 1.414 and hence around 1Vrms.
When I click the "Estimate Peak Output mV" button, I got 1415.6mV, as expected. This is ARTA's estimation of 0dBFS (peak output) based on my input above (and not the peak value of the rms voltage I entered, which is at -3dBFS), but here it's using peak and not rms, and 1415.6mVpk corresponds to 1Vrms also.
Attached is my calibration result.
When I click the "Estimate Peak Output mV" button, I got 1415.6mV, as expected. This is ARTA's estimation of 0dBFS (peak output) based on my input above (and not the peak value of the rms voltage I entered, which is at -3dBFS), but here it's using peak and not rms, and 1415.6mVpk corresponds to 1Vrms also.
Attached is my calibration result.
Attachments
Last edited:
Hi cwtim01,
I hope that's it, because they are changing levels without telling you what they are doing. If they wanted to get to 1 Vrms, why not just set it up that way and be done with the gymnastics?
The one bit of great news is that the difference between L and R channels is so tiny that the error band is much greater. I am very impressed with that.
Thank you for the explanation. Now I can complete the cal procedure.
-Chris
I hope that's it, because they are changing levels without telling you what they are doing. If they wanted to get to 1 Vrms, why not just set it up that way and be done with the gymnastics?
The one bit of great news is that the difference between L and R channels is so tiny that the error band is much greater. I am very impressed with that.
Thank you for the explanation. Now I can complete the cal procedure.
-Chris
Hi lemon,
Actually, that same PDF of instructions did not clearly state that they changed the level for the estimated levels. Once it became clear that they did that, it was easy. The level change came out to x 1.414, which is the same ratio of rms voltage to peak voltage. So when it came out as that ratio too high, it seemed like an error had taken place.
-Chris
Actually, that same PDF of instructions did not clearly state that they changed the level for the estimated levels. Once it became clear that they did that, it was easy. The level change came out to x 1.414, which is the same ratio of rms voltage to peak voltage. So when it came out as that ratio too high, it seemed like an error had taken place.
-Chris