When ac coupling the input of the TLK6201EA to the output of another part that is 3.3V CML which is back terminated 50 ohms to 3.3V (i.e. such as another TLK6201EA) and has the typical 400mV single ended CML swing, the input of the TLK6201EA being driven will experience the following:
1.) The TLK6201EA input that is being driven low will go to -200mV (i.e. the 400mV input swing will be swinging around the input termination voltage which is ground, that results in a single ended input swing of +/-0.2V on each leg), this is in violation of the 0.5V absolute minimum rating of Din+, & Din- shown on pg 4 of data sheet.
2.) At initial power up before the ac coupling caps have charged (i.e. there's 0V across the ac coupling caps) the input which is being driven high will see ~(3.3V/100ohms)=33mA which is in violation of the continuous current spec of Idin+, Idin- of 25mA, which is shown on pg 4 of the data sheet. This is a transient condition i.e. not continuous but can't find anywhere in the data sheet where absolute maximum transient conditions are specified.
Figure 4 on pg7 of the TLK6201EA data sheet clearly shows that the part is intended to operate with an ac coupled interconnect, can someone please help me reconcile these apparent contradictions?
John,
Good catch about the absolute minimum level - it should actually be -0.5 V, not +0.5 V. There is an ESD diode that is active around -0.7 V to -1.0 V. You are also right that the initial transition from a 0 V transmit output (before initial power-up) to a high-level output can transfer through the AC-coupling capacitor and potentially result in a relatively large input current at power-up. The duration of this current should be short enough, though, that the "continuous" condition does not really apply. Also, I would expect the slew rate of the initial ramp to be slow enough that the impedance of the capacitor would help limit the current entering the DIN inputs.
Regards,Max RobertsonAnalog Applications EngineerTexas Instrumentsm-robertson@ti.com