ADC12DJ5200RF: TMSTP+ signal

FRANK1

Expert 7815 points

Part Number: ADC12DJ5200RF

Hi sir,

Do you have any recommendation for biasing the TMSTP+ signal on ADC12DJ5200RF?
Non-differential setting.

Also the Which Voltage should it be biased with if any specific is required?

Thanks.

Regards,

over 2 years ago

0 Neeraj Gill over 2 years ago

TI__Mastermind 20541 points

Hi Frank1,

I would recommend to either use the 1.8V HSTL or SSTL standard IO on the FPGA. Generally the FPGA has internal series 50ohms resistors and the ADC also has 50 Ohms to ground for each leg when TMSTP_LVPECL_EN = 1. So in theory the circuit shown below should work and there shouldn’t be a need to add any component . If the FPGA does not have the Series 50ohms inside the FPGA you can add them on the board on each line.

Regards,

Neeraj

0 FRANK1 over 2 years ago in reply to Neeraj Gill

TI__Expert 7815 points

Hi Neeraj,

Thank you for your response.

Well our implementation is a little different.

We have an original design that we would like to change.

The input is a single eded analog signal from an RFoF and AC coupled. Which means that we only get a pulse on the input.

And in datasheet is states that the inputs (TMSTP) must be externally biased.

Any recommendations on the biasing?

Thanks.

0 Neeraj Gill over 2 years ago in reply to FRANK1

TI__Mastermind 20541 points

Hi Frank1,

They way you have biased TMSTP looks ok to me. Do you want to the signal on TMSTP pins to differential or just single ended?

Regards,

Neeraj

0 FRANK1 over 2 years ago in reply to Neeraj Gill

TI__Expert 7815 points

Hi Neeraj,

1) We would like to bias using 1.1V or 1.9V instead since 3.3V carry noise into the ADC.
2) The input signal is just a single ended pulse (1PPS) which means that when the pulse dies out (due to AC coupling) the inputs to the ADC will only see the difference in bias (0.096V) which is less that the specified 0.4V diff requirement for the ADC.
Will this work then?

Thank you.

0 Neeraj Gill over 2 years ago in reply to FRANK1

TI__Mastermind 20541 points

Hi Frank1,

Yes, you can bias using 1.1V or 1.9V.

Yes, I think you proposal should work.

Regards,

Neeraj

0 FRANK1 over 2 years ago in reply to Neeraj Gill

TI__Expert 7815 points

Hi Neeraj,

Does this mean that for the ADC to detect a "0" on TMSTP it only require -0.096V between + and -?
Why is it then specified 0.4V in data sheet?

Is there a hysteresis on the inputs?

Thanks.

0 FRANK1 over 2 years ago in reply to Neeraj Gill

TI__Expert 7815 points

Hi Neeraj,

Can you take time to review the addition?
Thanks.

0 Neeraj Gill over 2 years ago in reply to FRANK1

TI__Mastermind 20541 points

Hi Frank1,

If the differential voltage is below 0.4V the input will register it as a "0'. if it more than that it will register it as "1".

Regards,

Neeraj

0 FRANK1 over 2 years ago in reply to Neeraj Gill

TI__Expert 7815 points

Hi Neeraj,

customer wanna to confirm the following to ensure his understand correctly.

Please forgive me for asking more questions about this, but I must be sure that I understand you correctly.
What you say is that if TMSTP+ is 0.4V higher than TMSTP- it will be register a "1".
And if TMSTP+ is 0.3V higher than TMSTP- it will register a "0". Also if the difference ("TMSTP+" - "TMSTP-") is 0V or even -0.4V it will register a "0".
Is this correct?

Thanks.