This guy got incredible results from the PCM4222EVM board:
http://www.diyaudio.com/forums/equi...ject-audio-measurements-tool.html#post2072409
http://www.diyaudio.com/forums/equi...ject-audio-measurements-tool.html#post2072409
If you look through all his measurements they aren't a lot different from mine. I have not tried the -25 dB measurement. http://www.diyaudio.com/forums/equipment-tools/160432-diy-analog-digital-converter-project-audio-measurements-tool-3.html#post2127696
I still have not figured out where to buy the AKM boards. Maybe we need to create one.
So, the great results were from setting the input signal to only -25dB? I guess that makes some sense.
Larry Burk has an M-Audio 1212M, and will try it out when he has time. Depending on his data on the noise floor slope, that might be a good choice.
It's hard to find good info on these cards and modules -- many sites don't provide any technical data at all -- just reviews like "Great!" or "terrible sound." It seems strange to me that many of the "new" devices out there are still three and four year old products. And some of the "pro" price level devices have very mediocre specs when you eventually find the specs page. The PCI/PCIe cards seem dominated by the wants of gamers -- do they really use 7.1 channel system? But specs are hard to come by, and heaven forbid that they list the chipset used.....
It's hard to find good info on these cards and modules -- many sites don't provide any technical data at all -- just reviews like "Great!" or "terrible sound." It seems strange to me that many of the "new" devices out there are still three and four year old products. And some of the "pro" price level devices have very mediocre specs when you eventually find the specs page. The PCI/PCIe cards seem dominated by the wants of gamers -- do they really use 7.1 channel system? But specs are hard to come by, and heaven forbid that they list the chipset used.....
Do you mean the Creative's E-Mu 1212M?
E-MU Systems - 1212M PCIe - PCIe Digital Audio System
It looks like a great system if you have a desktop computer. I don't, so I would need an expensive PCI expansion box.
E-MU Systems - 1212M PCIe - PCIe Digital Audio System
It looks like a great system if you have a desktop computer. I don't, so I would need an expensive PCI expansion box.
Isn't the LM317 a little noisy? Isn't there a lower noise CCS that could be created? I'm having trouble finding a schematic at the moment.
Frex (member here) says the ESI Waveterminal 192X (not the 192L) uses the AKM5394 -- I've tried to find one and not only is ESI's marketing non-existent and unhelpful, so are the products, except for the Juli@, and those come from China with limited distribution. But at least they are reasonably priced. Frex says the AKM5394s are almost impossible to find -- I wonder why?
The Juli@ is not acceptable to me, vis-a-vis its rapid rise in the noise floor above 30kHz, similar to the EMU 0204 and others. Otherwise a very nice card. As to other cards (not USB modules), there's something called the Lynx which are very expensive, but may be OK. I'm really stumped by the lack of USB sound modules with great performance, seeing the on-going development of faster and better chipsets.
The Juli@ is not acceptable to me, vis-a-vis its rapid rise in the noise floor above 30kHz, similar to the EMU 0204 and others. Otherwise a very nice card. As to other cards (not USB modules), there's something called the Lynx which are very expensive, but may be OK. I'm really stumped by the lack of USB sound modules with great performance, seeing the on-going development of faster and better chipsets.
It's no problem. I'm confused too, about which thing to get for my situation. I still like the PCM4222EVM because it has balanced inputs that I can use with Pete Millett's sound card interface. I think it's splitting hairs when picking an ADC at this level of performance. I'm planning on linking the PCM4222EVM to my laptop's USB via the miniDSP miniStreamer.
I just ordered the parts for figure 13C from his part II article in Audio Express on current sources. I'm going to either remove or bypass the 100 ohm resistors on Victor's board.
Well, reading the spec's on these things isn't always helpful, since I don't recall seeing one rated for better than -120dB at all yet some seem to measure better than that. So, I'm wading in a swamp trying to figure out which to get.
I understand, Dirk, I'm wading in there too. For my purposes -- measuring oscillator distortion in the 0.0001% area and less (sometimes much less), the SNR of these cards is not the ultimate spec, although more is always better. I can live with SNRs of 110dB just fine -- spectrum analyzer averaging helps clean that up to give noise floors of -130dBu and more.
The Gearslutz site has a nice comparo of the "transparency" of sound cards and modules that you'll find interesting. I've lost the link, but if you go there I think you can find it. Nice methodology, and very interesting results.
Larry Burk sent my some plots of an old EMU 1212M (PCI card) and the noise floor is relatively flat out to the 90kHz Nyquist limit. So I may look into one of those. Larry says it's a bear to use, tho'.
The Gearslutz site has a nice comparo of the "transparency" of sound cards and modules that you'll find interesting. I've lost the link, but if you go there I think you can find it. Nice methodology, and very interesting results.
Larry Burk sent my some plots of an old EMU 1212M (PCI card) and the noise floor is relatively flat out to the 90kHz Nyquist limit. So I may look into one of those. Larry says it's a bear to use, tho'.
Last edited:
If I see another sound card that screams "7.1 channels! 24 bits! only 17.99$!!", I will scream.
I'm still convinced that the PCM4222EVM is a pretty good choice for me, since I can put it in a separate RFI and EMI proof box (that I am going to make myself), and it has balanced inputs.
I'm still convinced that the PCM4222EVM is a pretty good choice for me, since I can put it in a separate RFI and EMI proof box (that I am going to make myself), and it has balanced inputs.
Last edited:
1) The current noise of an LM317 won't be an issue with the shunt regulators and parallel caps. It would be if you were dealing with high impedances.
2) I suspect the interface circuit may be limiting the ADC's, The TI should be pretty adequate for most applications. The demo board is finicky and mine sometimes seems to stick in 16 bit mode.
3) A Lynx card is a good alternative but big bucks. I like the ESI card a lot. The HF noise is only an issue in a few cases. I also like the driver and mixer interface. Unlike most of the cards in this class the Juli supports 176.4.
4) The Juli can connect directly to the I2S interface of an ADC and DAC. Its a bit of a hack but it would be possible to hatch a replacement for the on board chips since the analog board is removable. But you are still hooked to a PCI interface.
5) The ESI Waveterminals were discontinued years ago.
2) I suspect the interface circuit may be limiting the ADC's, The TI should be pretty adequate for most applications. The demo board is finicky and mine sometimes seems to stick in 16 bit mode.
3) A Lynx card is a good alternative but big bucks. I like the ESI card a lot. The HF noise is only an issue in a few cases. I also like the driver and mixer interface. Unlike most of the cards in this class the Juli supports 176.4.
4) The Juli can connect directly to the I2S interface of an ADC and DAC. Its a bit of a hack but it would be possible to hatch a replacement for the on board chips since the analog board is removable. But you are still hooked to a PCI interface.
5) The ESI Waveterminals were discontinued years ago.
Last edited:
I've found the 4222EVM rather easy to configure 
The external clock thing is kinda less straightforward. It takes the external clock, divides it by 2, and works with it as internal oscillator. To get the 1x/2x/4x sampling frequencies, you'll have to stick to 22/24MHz oscillators. The "multipliers" actually are dividers built inside the PCM4222 ADC - for 1x (44.1kHz sampling) it divides the MCLK by 4 (4x - 1/4), 2x - 1/2, 4x - 1/1. You'll have to stick to 128Fs/256Fs/512Fs oscillators as the board won't support 192Fs/384Fs modes (limitation of DIT).
My spreadsheet with configuration settings - for PCM mode + it's sampling rates:
http://s3t.it/data/uploads/pcm4222evm_v0.1b.xls
The external clock thing is kinda less straightforward. It takes the external clock, divides it by 2, and works with it as internal oscillator. To get the 1x/2x/4x sampling frequencies, you'll have to stick to 22/24MHz oscillators. The "multipliers" actually are dividers built inside the PCM4222 ADC - for 1x (44.1kHz sampling) it divides the MCLK by 4 (4x - 1/4), 2x - 1/2, 4x - 1/1. You'll have to stick to 128Fs/256Fs/512Fs oscillators as the board won't support 192Fs/384Fs modes (limitation of DIT).
My spreadsheet with configuration settings - for PCM mode + it's sampling rates:
http://s3t.it/data/uploads/pcm4222evm_v0.1b.xls