THS4303: mistake in datasheet, S-Parameter table ?

Hello Mete,

I am looking into the S21 (Ang) on page 7; I will give you an update on if we have another document to reference or can attest to a more accurate value.  

Best,

Alec

Thanks Alec. One more request. It is also said that S-parameters are "(Measured using standard THS4303EVM, edge number 6454762, with VS = 5 V in a 50 Ω test system)". But I guess only the opamp is measured, so there has to be some kind of correction, delay/port-extension etc. Is it possible to clarify this and if it is the case, is it possible to give the approx. delay the board is causing so maybe I can get this out in my measurement too.

Best regards,

Mete

Hi Mete,

Thank you for the follow-up!  I will also look into any board delay and also clarify some of the measurement decisions.  I'll get back to you soon.

Best,

Alec

Hi Alec,

Thanks for the update.

What I mean by delay is actually if you can share more info about how s-parameters are measured. For example, it says they are measured using standard THS4303EVM but in that case I guess s21 (dB) cannot be 20dB due to series resistor matching 50-ohm line, so I guess some correction is applied. Same for other values as well, since there is no calibration points on the EVM board, I guess additional correction is performed eg port extension/electrical delay. It is not that important, I just wondered if I can measure the same values.

Best regards,

Mete

Hi Mete,

I better understand your question now.  I can ask my team about our typical approach when characterizing devices in case there is a good explanation.  Thank you for helping me see the point you were asking.

If I get a meaningful answer I will append it here.

Best,
Alec

Hi Mete,

Just a final update here: when we make measurements in the lab using EVMs and other test boards, we do our best to properly calibrate the devices so we are only measuring the device under test (DUT).  As best as we can, this minimizes effects that are not introduced by the IC and allows us to confidently display measurements in datasheets.

Some of the steps we may take include using a blank EVM to attempt to characterize how it performs without the IC, understanding how its effects can add/change the IC performance.

For your calculations and design, you are working with data which is representative of how the IC performs; the various calibration and measurement steps taken provide that clarity and knowledge.

I hope this helps you understand a bit of what we do to account for board behavior and various delays during measurement.  

Best,

Alec

Hi Alec,

I was also thinking you are using something like a blank EVM as you say. Thanks a lot for clarifying this as well.

Best regards,

Mete

Hi Mete,

Absolutely, I am glad we were able to have a good discussion and clear things up.

Best,

Alec