LMH7322: Comparator output is fixed

Hello Purnenduk,

To get a LVDS compatible output, you need a negative supply of greater than -2V to VEE. Note the LVDS output example (figure 8) has a -5V supply on VEE.

Also note that the termination resistors (100 ohms) need to go to ground at the receiving end.

Hi, thanks for the reply. 

Just to test I have masked the pad and connected VEE -4.5V (available on board). So now when,

IN+ > IN-;  Q+ = 1.285V, Q- = 0.859V

IN- > IN+; Q+ = 0.916V, Q- = 0.934V

Do I need to AC couple MAX9111 somehow?

NOTE: I have tried with different output pull down resistor values (62 to gnd, 50 to gnd, 88 to gnd, 410 to VEE) but result is more or less similar.

Kindly help, I am running out of ideas here

Also note that the termination resistors (100 ohms) need to go to ground at the receiving end.
--Can you elabolate. do I need 100ohm from both Q+ and Q- to GND in parallel to 62ohm

Hello Purnenduk,

The termination resistors should only be 100 ohms to ground *at* the load. Remove any differential terminations.

How far is it between the LMH7322 and the MAX9111? If less than a centimeter, then just remove R52 and change R44 & R48 to 100 ohms.This should give you the 1V & 1.4V LVDS swing.

If the MAX9111 is a long distance away, then the 100 ohm resistors should be placed *at* the MAX9111 from each input to GND to avoid reflections.

Also, do not put an inductor in the ground line (VEE) without a bypass cap..this can cause ringing. Short L17.

With MAX 9111

Without MAX9111

1. Pull down to VEE is close to LMH7322

2. trace length between two IC is less than an inch and 100ohm differential characteristics impedance is used.

3. 100 ohm termination is near MAX9111 as per datasheet

4. Positive and negative terminals seems to be unbalanced. (Loaded differently) As circuit works okay when MAX9111 is de-soldered

for testing we have used 50MHz square wave input with threshold set in between. Waveforms are attached with and without MAX9111. (We have used normal 1M probe hence signal looks more noisy. Also input is provided with some external wire soldering hence some noise is due to that.)

Output of MAX9111 is staying at 3.2V with small swing of 0.1V.

All inductors are shorted.

Inputs are captured at input of MAX9111

Hi Purnenduk,

Could you please clarify the two waveforms that you are showing in the screenshots. Also, have you removed the differential termination(R52) as mentioned by Paul and kept only the termination resistors at the load (R44 and R48 to ground)?

Regards,
Jaskaran