BUF802: BUF802 being switched ON/OFF

Hello Dr. Alan Victor,

Thank you for connecting on the BUF802 buffer and your proposed use-case.  I am examining possibilities alongside the broader High-Speed Amplifier team.  Once we know if there is a path forward or not for this application, we will provide that update.

Best,

Alec

Thank you Alec. I have used your Spice model.  I noted that if the CLL pin is used and toggled from -6 to 0 volts, the output port will provide a proper replica of the input as well, no signal transfer as desired. The Transient response of this switching of the input signal to the output appears well behaved. Of course, this is the model and it may not be an appropriate method to provide for switching control of the amplifier. 

The CLL pin appears to provide a transient free enable function. I need to verify 

that the model correlates with on bench test and measure. Relay will not work. 

Not nearly fast enough! Thanks! 

Alec, a question to the TI design team, is the NL model for the BUF802 correct as to the operation of the CLL pin. 

Taking the CLL pin from -6 V to 0 V forces the output to follow the input signal OR NOT and in a precise manner with no 

significant transient response. The  output as CLL goes HIGH is adjustable to some degree and within minimum DC offset, ~ zero V. 

I can forward a screen shot of the simulation. 

Here is a screen shot of the SIM using the CLL pin function. 

Hello Dr. Alan Victor,

Thank you for the additional information; I will align with the design team to discuss your findings.

Best,

Alec

One other item, the model has a port for IN_BIAS, but not R_BIAS. Is the model incomplete or is it to be revised? 

Hello Dr. Alan Victor,

I am still awaiting the discussion with members of my team.  I will make sure to address whether the modeling of R_BIAS is integral to the model's operation.

Best,

Alec

Thank you Alec. 

I did some work on this and looked over the model list and the data sheet and I think I 

have a better handle on the model/data sheet items. In the BUF mode the need for the 

10 Meg IN_BIAS is not required. The FET input can be DC coupled or not and it appears

that the R_bias value of 17.8 k may be internal to the model. All that said, it appears that the CLL 

line can be used as a signal strobe control to the output, at least the model permits that. Bench 

measurements may prove otherwise. Thanks again, Alan

Hello Dr. Alan Victor,

I have examined the model as well; I would concur that functions of the BUF802 are internal and not 'pinned-out' in the SPICE model.  I am following-up on this internally to better understand why some HSAMPs models disclose all functions that are modelled or not modelled, and others provide some assumptions alongside the explicitly modelled parameters.

Once I have additional information from team members who own/specified the operation of the BUF802, I will report back on that case.

Best,

Alec

Thank you Alec. 

Hello Dr. Alan Victor,

I will be away from the office this afternoon & Monday.  I will be back and available on Tuesday.  My colleagues will look to make sure your responses are addressed in the interim.

Best,

Alec

Thank you. 

Hello Dr. Alan Victor,

I am back from travel and I am waiting to meet about your questions.

Best,

Alec

Hello Dr Alan , you are right the 10M @ IN_Bias is not needed in BUF mode as mentioned in the Pin Out table .

The BUF802 Main Input can be DC or AC coupled , it supports both .

17.8k ohm is a value suggested to bias the 1st stage JFET in which it offers enough bandwidth to support AC specs mentioned in datasheet.

Hello Dr Alan ,

This is Shreenidhi here , product definer of BUF802 and apologize for the delay form our end in our reply .

Yes you can use CLL and CLH to ensure output is driven to Zero when signal is not needed to appear at output.

As shown in the datasheet CLL and CLH both are capable of being driven to any value between 0V to VEE / VCC respectively.

However do note these same clamp pins also clamp the input via the ESD diodes.

These diodes are rated for 100mA and i would want to ensure the current through the input diodes do not cross 100mA.

So basically you need to have enough input impedance such that the current through the input impedance is limited to 100mA.

Also the AC clamp response is shown in 8.3.1 Input and Output Over-Voltage Clamp

Please feel free to email me ( shree@ti.com)  / raise any question on BUF802 on this thread.

Great and thank you for the feedback. Next steps will involve bench testing. There should be no issue

in preventing limited current through the ESD diodes. Regards, Alan