Thanks HpW, added 5572:
Lastly, how does the AK5574 compare to *5572* [5397] and (5394A)?
S/(N+D) at -1 dB: 112 *112* [108 but 110 at -2 dB] (110)
S/(N+D) at -20 dB: 97 *97* [102] (100)
S/(N+D) at -60 dB: 57 *57* [64] (60)
DR (A-weigthed): 121 dB single channel but 127 in 4-to-1 *121 but 124 2-in-1* [127 but 130 mono mode] (123)
So the 5572 seems to be every bit the same as the 5574 except it does not get the benefit of the averaging, and the card is a lot cheaper. Will need to read up on the implementation you linked to.
Lastly, how does the AK5574 compare to *5572* [5397] and (5394A)?
S/(N+D) at -1 dB: 112 *112* [108 but 110 at -2 dB] (110)
S/(N+D) at -20 dB: 97 *97* [102] (100)
S/(N+D) at -60 dB: 57 *57* [64] (60)
DR (A-weigthed): 121 dB single channel but 127 in 4-to-1 *121 but 124 2-in-1* [127 but 130 mono mode] (123)
So the 5572 seems to be every bit the same as the 5574 except it does not get the benefit of the averaging, and the card is a lot cheaper. Will need to read up on the implementation you linked to.
Yes, that is the explanation, see this.
I did consider (but not test) the PCM4220. The performance seems to be quite good, but I didn't like the increase in noise level at high frequencies (above 60 kHz).
The AK557x family also has a noise increase at the upper end, even starting at lower frequencies than 60 kHz, more like just around 20 kHz.
The AK5397 does not have a noise increase of this magnitude. So, apart from the higher distortion compared to the AK5394A, it is still an option worth considering.
Thanks for the enlightening link, Jens. Ouch, that hurts!
Now what I don't get:
The second picture (first graph) shows the 97 consistently 3 dB better up to about 60 kHz and still about 2 dB better near 96 kHz, while the 94A and the 97 run in parallel, i.e. both form a shallow bathtub.
The fourth picture has the noise of the 97 a lot worse throughout the 80 to 60 kHz range, but with a flat distribution. What was going on there? Noise modulation?
Did this effect, whatever it was, raise the harmonics also? If so, was the picture the same at 44.1 kHz?
Now what I don't get:
The second picture (first graph) shows the 97 consistently 3 dB better up to about 60 kHz and still about 2 dB better near 96 kHz, while the 94A and the 97 run in parallel, i.e. both form a shallow bathtub.
The fourth picture has the noise of the 97 a lot worse throughout the 80 to 60 kHz range, but with a flat distribution. What was going on there? Noise modulation?
Did this effect, whatever it was, raise the harmonics also? If so, was the picture the same at 44.1 kHz?
As an aside, the Lynx E22 / E44 has better claimed AD performance than the L22 (-110 dB at -1 dB fs vs -108), and it changed from AK4394(A) to CS5381. Looking at the CS data sheet, it would seem worse then the AK4398
-110 dB at -1 db
- 97 dB at -20
-57 at -60
-110 dB at -1 db
- 97 dB at -20
-57 at -60
I assume that the increased noise in the last picture is due to some kind of quantization noise, since it is level dependent. I didn't investigate it further.
The increase in my case was rather large. A similar increase, although not quite as large, can be seen in the AK5397 evaluation board manual, e.g. in figures 14, 20 and 21.
I don't think I tested it at 44.1 kHz.
The increase in my case was rather large. A similar increase, although not quite as large, can be seen in the AK5397 evaluation board manual, e.g. in figures 14, 20 and 21.
I don't think I tested it at 44.1 kHz.
Yes, it is even evident in your measurements of the 94A: - 157 dB at 1 kHz without signal to - 146 dB with signal. It's just that the 97 went from -160 to roughly -137. Clearly a trade-off in modulator design that probably looked better on the data sheet 🙂
I have the demo board for the PCM4220 and while good the distortion performance is not near that of the AK5394A. I did some work on the demo board to get the best performance and those insights are in the RTX.
I was baffled by the claims from Lynx about the Cirrus chip, however the AK5394A has been discontinued so Lynx has to keep pushing what is available.
The best performance re: distortion is at about -12 dBFS and all the chips are compromised at close to FS. Being inside a PC does not seem to limit the performance using the balanced/differential interface. The EMU 1212M is a good alternate here if you can find it. There are patches to get the emu running on win 10 that work well. Easier to mod as well.
I was baffled by the claims from Lynx about the Cirrus chip, however the AK5394A has been discontinued so Lynx has to keep pushing what is available.
The best performance re: distortion is at about -12 dBFS and all the chips are compromised at close to FS. Being inside a PC does not seem to limit the performance using the balanced/differential interface. The EMU 1212M is a good alternate here if you can find it. There are patches to get the emu running on win 10 that work well. Easier to mod as well.
Thanks guys, you have been very helpful. Looks like I need not look any further then.
I have bought the odd EMU card over the years but have shied away from using them because of the clunky software and purported win10 issues (apparently, the patches stop working with most major windows updates).
I also just bought an L22 pci, so I will try to compare the two.
Demian, have you posted your insights on the 4220 buffer anywhere?
Thanks
I have bought the odd EMU card over the years but have shied away from using them because of the clunky software and purported win10 issues (apparently, the patches stop working with most major windows updates).
I also just bought an L22 pci, so I will try to compare the two.
Demian, have you posted your insights on the 4220 buffer anywhere?
Thanks
The remarkable thing about the AK5394A is that it outperforms the datasheet and the demo board. If it was just datasheet specs we were comparing then other ADCs would be comparable, but they aren't.
RTX6001 will be fully supported in the upcoming kernel 5.8 linux/pcm.c at v5.8-rc5 * torvalds/linux * GitHub , a few weeks away.
It's all DIY in the end 🙂
I have to thank to Demian Martin and Matthias (mbrennwa) for their help and patience with testing.
I have to thank to Demian Martin and Matthias (mbrennwa) for their help and patience with testing.
It took me a moment to realise you were asking about the TI ADC. I did not post anything that I remember. It was all years ago.
I have a 4220 demo board that I would pass along. However shipping to the EU would make it more expensive than its worth.
Overload protection issue
Hi everyone,
I have been caught out by what I suspect is the overload protection mod on my RTX6001. It appears the overload protection drops the level by 30db when it activates. You can hear a relay click if you send the RTX a high signal level, both channels operate the same. What caught me is there is no indication that the overload protection has activated i.e. the orange OVF LED does not light. An increase of indicated THD when protection is active is what caught me out.
Have others experienced this? there a simple mod to illuminate the OVF LED in this circumstance?
Apologies if this has already been discussed. With almost 300 pages in this thread it would be easy to miss.
Thanks
Bruce
Hi everyone,
I have been caught out by what I suspect is the overload protection mod on my RTX6001. It appears the overload protection drops the level by 30db when it activates. You can hear a relay click if you send the RTX a high signal level, both channels operate the same. What caught me is there is no indication that the overload protection has activated i.e. the orange OVF LED does not light. An increase of indicated THD when protection is active is what caught me out.
Have others experienced this? there a simple mod to illuminate the OVF LED in this circumstance?
Apologies if this has already been discussed. With almost 300 pages in this thread it would be easy to miss.
Thanks
Bruce
The red OVF LED's should light up, when overflow is detected.
What input setting did you use?
What was the input signal level?
Have the mods been installed in your unit?
What input setting did you use?
What was the input signal level?
Have the mods been installed in your unit?
Overload protection issue
Thanks Jensh,
Yes I have done the overload mods. My RTX was one of the initial group buy.
If I have a sine signal of +33dbu (36V) and feed it to the RTX with input attenuator set to 10dbV (3.16V), the following occurs - I hear relay click as signal is applied, the orange OVF LED sometimes flashes briefly as relay clicks (but then stays off), in software a level of -12dbFS is registered. Same situation on either LEFT or RIGHT input.
With above signal, if I change the attenuator switch I get these results - 40dbV -12dbFS no LED, 30dvV -2.3dbFS with OFL LED, 20dbV -0.6dbFS no LED, 0dbV -2.3dnFS with OVF LED, -10dbV -0.6dbFS no LED, -20dbV -0.2dbFS no LED.
By trying a few different signal levels, it appears a moderate overload e.g. up to 20db, the OVF LED will illuminate but with large overloads the LED stays off. Not so helpful when testing power amplifiers!
Thanks Jensh,
Yes I have done the overload mods. My RTX was one of the initial group buy.
If I have a sine signal of +33dbu (36V) and feed it to the RTX with input attenuator set to 10dbV (3.16V), the following occurs - I hear relay click as signal is applied, the orange OVF LED sometimes flashes briefly as relay clicks (but then stays off), in software a level of -12dbFS is registered. Same situation on either LEFT or RIGHT input.
With above signal, if I change the attenuator switch I get these results - 40dbV -12dbFS no LED, 30dvV -2.3dbFS with OFL LED, 20dbV -0.6dbFS no LED, 0dbV -2.3dnFS with OVF LED, -10dbV -0.6dbFS no LED, -20dbV -0.2dbFS no LED.
By trying a few different signal levels, it appears a moderate overload e.g. up to 20db, the OVF LED will illuminate but with large overloads the LED stays off. Not so helpful when testing power amplifiers!
It looks like your unit is still using an old firmware version. You should update to the latest firmware (V1.19), which you can find here:
RTX6001 firmware
at the bottom of the page.
RTX6001 firmware
at the bottom of the page.
Thank you JensH
That fixed the issue. Much easier and less hassle than I had feared! Really appreciated.
I have had my RTX6001 about 2.5 years now. Gotta say, it is an excellent piece of test equipment. And when not being used for its primary purpose I use it as my music DAC in my workshop. Sound quality is fantastic🙂
Warm regards from Australia
Bruce
That fixed the issue. Much easier and less hassle than I had feared! Really appreciated.
I have had my RTX6001 about 2.5 years now. Gotta say, it is an excellent piece of test equipment. And when not being used for its primary purpose I use it as my music DAC in my workshop. Sound quality is fantastic🙂
Warm regards from Australia
Bruce