I have decided to play a bit with RightMark Audio Analyzer or simply RMAA measurements. Well, it appeared that none of my own sound cards has a line-in, which is a “must”. First, I scanned eBay for the alternatives which would work with current versions of Windows. As I am not sure either I will continue with those RMAA measurements, I decided not too over-invest and my choice ended with 10USD Chinese CM6206LX-based USB card (via eBay), for which I am waiting right now 🙂
Just before buying the USB card, I spent good time investigating in what type of sound-card I am getting etc. So here, I will share what I managed to find online.
Few numbers about CM6206LX chip
Just keep in mind that there is two pretty similar named chips: CM6206 and CM6206LX from C-Media Electronics, Inc., a Taiwanese company. I speak here only about the CM6206LX chip.
The chip itself has 6 channel DAC output with 16 bit resolution, S/PDIF input&output; supports stereo MIC, most likely internally runs 8 KHz sampling rate (I need to check this) and of course has Line-In 🙂
Provided SNR and dynamic range for DAC’s are 96dB, what is a theoretical max for 16-bits. Of course in the reality, many things (everything?) depends on the noise level. According to the datasheet, given THD+H value is -88dB for both DACs and ADC. That would be an equivalent of 0.0039811%, what would be OK’ish. One parameter is a bit worrying and that is Power Supply Rejection Ratio (PSRR), which is only -40dB. What does it mean? The noise in poorly filtered power supply will quite easily find its way to output signal.
Few of those who tried this card before me
If you a Mac fan, you might be interested to read this post about enabling CM6206 card under Mac OS.
Some of the card measured parameters are given here. Quite informative article with comparison with few other USB-based devices from higher price range.
Laurence used to input a digital signal through S/PDIF and record it on PC.
Here, a DIY forum user shared his tweaking experience of this card.
PCB view and missing parts
Here is top view of PCB, photo source:
If we scan the PCB we see that few parts are missing. One of them is U2 but this is for EEPROM chip and some customization. Let’s move on. L1, L2 and L3 inductors are missing too. That is more interesting. They supposed to create the LC filters similar as it shown here:Ok, as I do not have the card yet, I cannot confirm all inductors purpose. But at least it looks such LC filter would be really nice to implement after some modifications 🙂
Are there any real-life RMAA measurements of this card?
Yes, there are few. The first of them are given here.
First of them is here, where a guy performed the RMAA measurements with couple suggestion for the tweaks. Here is another thread with the RMAA measurements before and after tweaking CM6206LX-based card (via eBay).
The frequency response curve for unmodified version of this card, figure source:
We can see couple things. Normally, the frequency only from something like 250Hz is reproduced, whatever is below most likely won’t be. The curve is not linear in the region 4kHz-19kHz and its variations are within 0.5dB. This is limited by the capacitors size which are installed at Line-In input.
The noise level of the card before any modifications, figure source:
Wow, the noise level is impressive. I was not expecting anything like this but just remember PSRR -40dB. According to a second link with RMAA measurements, such noise level is only related to noisy power supply from the USB port.
Few measurements after tweaking
Here is the frequency response curve after putting in parallel 22uF capacitors to existing 0.1uF at Line-in input:
We can see that the curve was shifted to the left and right now the frequencies starting from 20Hz should be reproduced.
The 220uF inductors were added in missing places on the pcb and the noise spectrum became like this, figure source:
Great improvement compared to previous noise curve 😉 What is interesting, there are visible noise at 1Khz and its harmonics 2kHz, 3kHz etc. Most likely, it is the noise coming from USB port.
My upgrade plan
1) To put 330uF in missing inductors L1, L2 and L3 places and to put low ESR 2200uF (via eBay) in the places of C14 and C29. Together with the inductors, they should create LC filters with cut-off frequency 187Hz. I would expect visible noise of 1kHz and its harmonics to disappear.
2) Instead of C1, C2, C3, C4, C5, C6 and C30, C31 put Nichicon ES series 100uF caps (via eBay).
3) Plan B – to build a linear 5V power supply, as the guys on head-fi.org found also that another sound-card with an external PSU sounds a lot better than powered through USB port.
I hope my USB card (via eBay) will arrive soon too 😉