I'm using the ADL5205 for a new design.
I have a doubt regarding the Maximum Input Swing of the ADL5205. While in page 3 of the datasheet specifies that the Maximum Input Swing is 8Vpp, in the absolute maximum ratings of page 6, the datasheet states that the Differential Input Voltage ((VINx+) − (VINx−)) is +-1V. Both informations seem contradictory. ¿What is the true maximum input swing?
My concern is about the maximum input level that the device can tolerate without damage (it doesn't matter if it is in linear operation or saturated) for the whole range of gain codes.
Thank you. Joan Ramon
The “Maximum Input Swing” of 8 Vpp on page 3 is for the lowest gain setting of -9 dB. Then as the gain is increased, the Maximum Input Swing has to be reduced correspondingly to maintain the same maximum output swing.
As for the “Differential Input Voltage” in the Absolute Maximum Ratings table on page 6, that takes effect starting around 14 dB of gain and up. At that point, there’s no more resistive attenuation on the input, so whatever signal is presented on the pins that will directly go on an input ESD protection circuit, which is the reason the input signal swing has to be limited.
If the maximum differential input voltage is consequence of the input ESD protection circuit, in my oppinion, a possible solution to avoid to damage the ADL5205 could be to limit the current flowing through the ESD protection by means of a small series resistor. ¿do you have information about the maximum current the ESD protection can tolerate or the maximum power it can disipate?
The series resistance approach should work, because essentially it forms a resistor voltage divider with the input thereby achieving the goal of reducing the input voltage swing at the pins. In fact, it may even make more sense to add a proper resistive attenuator pad (Pi or a T structure), because that way the impedances can be controlled as well, and mismatch can be reduced - if that is of concern.
I'm not sure going to the limits of the ESD protection in terms of current or power is the right direction. A proper design should use voltages nowhere near high enough to engage the ESD circuits during regular operation. Also, the usual human-body model (HBM) and charged-device model (CDM) don't cover continuous ESD operation. It's probably better to stay well within the abs max ratings for voltage swing at the pins, which will result in no current and no power dissipation on the ESD.