# AD9670 : Arranging cutoff frequency of DC Offset Correnction

Hello,

Our customer is considering to use AD9670 for his application.
But not for Medical equipment.

In that application, the input signal frequency range is 1kHz to 2MHz.
Thus, he must change CLG higher more than 10nF.

To move the cutoff frequency lower than 1kHz the CLG value will be estimated as 480nF according to eq.(6)

Could you answer the following two questions?

Q1. What is maximum value for CLG ?

Q2. I confirmed the cutoff frequency of table10 using eq.(6).
The conditions are gm=0.5mS , LNAgain=15.6dB
fhp = (1/2*PI)*LNAgain(=6.03)*gm(=0.0005)/10nF = 48kHz
fhp of table10 is 41kHz , it is lower than the result of eq.(6)
Other conditions are similar result.
Why is there this difference ?

Thanks,
y_suzuki

## Top Replies

Parents
• Hi y_suzuki,

This part is not really meant to operate at such a low frequency down to 1KHz. The main application is for medical ultrasound imaging that with probes starting at 1 MHz. The performance cannot be guaranteed at 1 KHz and there are several blocks that act as a HPF in the signal chain including the dc offset removal, the AAF, and the digital HPF. The calculation discrepancy is probably due to rounding errors but the equation is meant to scale the cutoff frequency with a new Clg value.

Regards

Ashraf

• Hello Ashraf,

The difference between table10 and calculation result are 17% . (42kHz vs 48kHz)
It is too large as a round error.
Our customer wants to confirm that the equation(6) is correct and he can  get the cutoff performance he expect using equation (6).
Does he able to get the expected cutoff frequency using equation(6) ?

Best regards,

y_suzuki

• Hi Suzuki-san,

The table values should be within +-20% accuracy. The customer can use the table to scale the cutoff frequency but they need to understand it will be +-20 accurate to reflect the gm variation over the process.

Thanks

Ashraf