MX93032 データシートの表示（PDF） - Macronix International

部品番号

コンポーネント説明

メーカー

MX93032

APPLICATION NOTE [M1 VERSION]

Macronix International 

MX93032

Weighted average of all 63 combinations

-7

-10

-13

-16

-19

-22

-25

-28

-31

Sum

-13

3.07E-5

0.0001842

0.0007076

0.0015908

0.0021134

0.0015908

0.0007076

0.0001842

3.07E-5

0.00714

-16

9.013E-5

0.0005408

0.0020771

0.0046699

0.0062042

0.0046699

0.0020771

0.0005408

9.013E-5

0.02096

-19

0.0002528

0.001517

0.0058271

0.0131006

0.0174048

0.0131006

0.0058271

0.001517

0.0002528

0.0588

-22

0.0005126

0.0030754

0.0118127

0.0265578

0.0352832

0.0265578

0.0118127

0.0030754

0.0005126

0.1192

-25

0.0008063

0.0048375

0.0185813

0.041775

0.0555

0.041775

0.0185813

0.0048375

0.0008063

0.1875

-28

0.000915

0.0054902

0.0210885

0.0474118

0.0629888

0.0474118

0.0210885

0.0054902

0.000915

0.2128

-31

0.0016925

0.0101549

0.0390058

0.0876941

0.1165056

0.0876941

0.0390058

0.0101549

0.0016925

0.3935

Sum

0.0043

0.0258

0.0991

0.2228

0.296

0.2228

0.0991

0.0258

0.0043

1

Talkoff Condition 1 : the combination of average near-end level (-19 dBm) and the average far end level (-28 dBm).

This talkoff rate should be less than 1 in 45 hours

Talkoff Condition 2 : the weighted average of the worst one percent of the 63 combinations. The talkoff rate should be

less than 1 in 10 hours.

Talkoff Condition 3 : the weighted average of all of the 63 combinations. The talk off rate should be less than 1 in 35

hours.

Tested Result :

Condition 1 (nominal ) : 1 hit per 38.4 hours

Condition 2 (worst 1 %) : 1 hit per 16.79 hours

Condition 3 (global) : 1 hit per 35.73 hours

( Spec. >= 45 hours)

( Spec. >= 10 hours)

( Spec. >= 35 hours)

3. Recommendation:

All of the test results corresponding to talkdown given above are obtained by configuring our demo board to "Adjunct

Model". That is, our demo board was connected directly to the Tip and Ring on which the far end speech and far end

CAS has been mixed. Therefore, there is no way to tell speech signal from the CAS on that testing configuration

adopted by us. For practical application, Near end speech instead of far end speech should occur on talkdown test. If

the mixed signal captured from tip and ring can be subtracted from the source signal captured from the near end

speech with proper gain. Then the SNR of resultant signal will be highly improved on adjusting the impedance match

between CPE and SPCS adequately. According to our experience, if the quantity of the echo canceller can excess 6

dB, the performance of CAS detector embedded on 93032-M1 will meet the Bellcore specification for talkdown. The

feature of echo canceller will be inherently possessed by phone mode, there will be additional circuit to implement

echo canceller for adjunct model.

P/N:PM0689

REV. 1.0.3, DEC. 13, 1999

62

Share Link: