SN65LVDS104 Worst Case Single-Ended VOH and VOL

Bob Kriner

Other Parts Discussed in Thread: SN65LVDS104

The SN65LVDS104 datasheet specifies the LVDS output parameters Differential output voltage magnitude (at 100 Ohms differential load) and Steady-state common-mode output voltage, and provides min and max values.

One (extreme) way to estimate worst case VOL is to take the minimum common-mode output voltage and subtract the maximum Differential output voltage magnitude (1.125 - 0.454 = 0.671V). Similarly, a way to estimate worst case VOH is to take the maximum common-mode output voltage and add the maximum Differential output voltage magnitude (1.375 + 0.454 = 1.829V). However, a typical LVDS signal swings from around 1V to around 1.4V, with a common mode of around 1.2V. It seems unlikely the output of the SN65LVDS104 would ever swing from 0.671V to 1.829V. In other words, the above approach seems it may be stacking too many extremes. It seems more likely the actual swings will be bounded to something less extreme.

Figures 13 and 14 show how VOL and VOH can vary with load current (at room temp and nominal voltage). However, in our case, we expect a differential load of around 100 Ohms is always present, so those figures don't help. They do show what can happen if the load disappears, but our intent is to understand worst case VOH and VOL with a load present. VOH and VOL don’t appear to be specified elsewhere in the datasheet.

Does TI have any characterization data that would show what to expect over the full operating temperature range of -40C to 85C?

Can the IBIS model for this device be trusted to determine this for min and max PVT cases (including over the full operating temperature range of -40C to 85C)? (Not all IBIS models or of excellent quality and/or worthy of use for making design decisions)

Any other options or approaches to bound VOH and VOL less extremely?

Please advise and thanks.

over 9 years ago

0 Miguel Robertson over 9 years ago

TI__Guru 62920 points

Hi Bob,

Sorry for the delay in getting back to you; I have been out of the office.

You're right that those calculations are a bit extreme. One thing to note, though, is the the VOD spec of 454 mV represents the maximum differential amplitude (difference between between P and N), not the peak-to-peak swing. The single-ended voltages would then be VOC + (VOD/2) for the high level and VOC - (VOD/2) for the low level. The highest single-ended output voltage that any LVDS104 unit could be expected to give would be 1.375 V + (.454 V / 2) = ~1.6 V. The lowest voltage that any LVDS104 unit could be expected to give would be 1.125 - (.454 V / 2) = ~0.9 V. You would not expect expect this range of voltages from any single device, however, since a single device (operated at a fixed supply voltage) would not have both minimum and maximum common-mode voltages. The most the common-mode voltage can vary between high and low states of a single device is 50 mV.

The IBIS model for this device was intended to cover both minimum and maximum PVT cases, so it should give you the same answer. In general I would tend to favor calculating this using the datasheet's electrical characteristics, though, since these are based on actual characterization/test data (whereas some simplifications might have been made in the model).

Please let me know if you have any additional questions.

Best regards,
Max

0 Bob Kriner over 9 years ago in reply to Miguel Robertson

Prodigy 180 points

Thanks Max, that answers the question, and thanks for pointing out the /2 issue in my original question.

Interface

Interface forum

SN65LVDS104 Worst Case Single-Ended VOH and VOL