zl50212 Zarlink Semiconductor, zl50212 Datasheet - Page 11

• Manufacturer Part Number: zl50212
• Description: 288 Channel Voice Echo Canceller
• Manufacturer: Zarlink Semiconductor

Data Sheet
2.2
After echo cancellation, there is always a small amount of residual echo which may still be audible. The EVP uses
Zarlink’s patented Advanced NLP to remove residual echo signals which have a level lower than the Adaptive
Suppression Threshold (TSUP in G.168). This threshold depends upon the level of the Rin (Lrin) reference signal as
well as the programmed value of the Non-Linear Processor Threshold register (NLPTHR). TSUP can be calculated
by the following equation:
where NLPTHR is the Non-Linear Processor Threshold register value and Lrin is the relative power level expressed
in dBm0. The NLPTHR register is 16 bits wide. The register value in hexadecimal can be calculated with the following
equation:
where 0 < NLPTHR
When the level of residual error signal falls below TSUP, the NLP is activated further attenuating the residual signal
by an additional 30 dB. To prevent a perceived decrease in background noise due to the activation of the NLP, a
spectrally-shaped comfort noise, equivalent in power level to the background noise, is injected. This keeps the
perceived noise level constant. Consequently, the user does not hear the activation and de-activation of the NLP.
The NLP processor can be disabled by setting the NLPDis bit to “1” in Control Register 2.
The comfort noise injector can be disabled by setting the INJDis bit to “1” in Control Register 1. It should be noted
that the NLPTHR is valid and the comfort noise injection is active only when the NLP is enabled.
The patented Advanced NLP provides a number of new and improved features over the original NLP found in
previous generation devices. The differences between the Advanced NLP and the original NLP are summarized in
Table 1.
The NLPSel bit in Control Register 3 selects which NLP is used. A “1” will select the Advanced NLP, “0” selects the
original NLP.
The Advanced NLP uses a new noise ramping scheme to quickly and more accurately estimate the background
noise level. The noise ramping method of the original NLP can also be used. The InjCtrl bit in Control Register 3
selects the ramping scheme.
NLP Selection
Reject uncanceled echo as noise
Reject double-talk as noise
Noise level estimator ramping scheme
Noise level ramping rate
Noise level scaling
Non-Linear Processor (NLP)
Feature
(dec)
< 1
Table 1 - Comparison of NLP Types
NLPTHR
TSUP = Lrin + 20log
NLPSel (Control Register 3)
NLRun1 (Control Register 3)
NLRun2 (Control Register 3)
InjCtrl (Control Register 3)
NLInc (Noise Control)
Noise Scaling
(hex)
Zarlink Semiconductor Inc.
Register or Bit(s)
= hex(NLPTHR
10
(NLPTHR)
(dec)
* 32768)
1
1
1
1
5(hex)
16(hex)
NLP Default
Advanced
Value
0 (feature
not supported)
0 (feature
not supported)
0 (feature
not supported)
0 (feature
not supported)
C(hex)
74(hex)
Default Value
Original NLP
ZL50212
11