• State
  • Replies 4 replies
  • Subscribers 47 subscribers
  • Views 79 views
  • Users 0 members are here
Support feedback
Options
Options
Related

TLV3603: TLV3603 Comparator Output Floats High

Naddy MY
Prodigy 30 points

Part Number: TLV3603
Other Parts Discussed in Thread: LMH6559, OPA846

Hi,

I have a pulse (38Mhz, 12ps) going through a TIA (OPA846) and then to a buffer (LMH6559), before going to some RC filters circuit to average out/amplify the voltage of the pulse. The average voltage then will go through a comparator (TLV3603) and expected is to give a high 5V provided that the average voltage of the pulse (IN+ of comparator) is higher than the reference voltage (IN-).

It works fine and as expected after the RC filters circuit. However, as it reaches the comparator, i could see the output is always high at the rail value (5V) despite removing my input pulse at the TIA. The Vavg after the RC circuits goes to 0 if i remove my input, but the same is not observed after the comparator where it stays at 5V, and im not entirely sure why.

For reference, my Vavg is 3V and my Vref is 1V. The output is high 5V. But when I increase my Vref to 4V, I expected it to go low (0V) as Vref>Vavg, but it stays high. The same thing was observed when I removed my incoming pulse connected to the TIA, where the output from comparator remains high at 5V.

Troubleshooting that I did:

  • Probe and made sure the supply pin is 5V and GND
  • Check voltage at pin 5 - it was 1.06V so its in hysteresis mode and not latch mode
  • Made sure the potentiometer wiper is working and connected to the IN- of comparator (saw it from scope the ref voltage moves up and down)

 

       

Please could I ask for some advice on this comparator issue. Thanks.

  • 0
    Prodigy 30 points

    Apologies, I forgot to mention something that's important for the context of my troubleshooting point no.3:  the reference voltage is controlled by potentiometer 

    Plus these circuits were built on a PCB.

  • 0 in reply to Naddy MY
    TI__Guru**** 174576 points

    Hello Naddy,

    Try measuring pin 5 voltage? Is it over 1.5V? Measure top of pin2, is it 0V?

    If pin 5 voltage is lower, try removing R2, so that pin 5 is open. Does output change properly? Is R2 really 620k?    

  • 0 in reply to Ron Michallick
    Prodigy 30 points

    Hi Ron,

    The voltage at pin 5 is 1.06V. I thought this value is fine for it to work in hysteresis mode ? Or I could be wrong?

    Pin 2 does give me 0V reading, and R2 is indeed 620k, I double checked with multimeter.

  • 0 in reply to Naddy MY
    TI__Guru**** 174576 points

    Naddy,

    How many TLV3603 were tested that have this issue?  If just one, then there is a good chance the comparator is now broken.

    Naddy MY said:
    The voltage at pin 5 is 1.06V.

    Per data sheet >=1.5V is definitely not in latch mode and <=0.35V is definitely in latch mode.

    Between these voltages, it could be latched or it might not be. For trouble shooting purposes, having no R2 should be considered not latched (on a good sample).    

      

    Comparators are pretty simple output follows input difference unless...

    1) sample is damaged; then almost anything is possible. 

    2) Input difference is smaller than Vos + hysteresis ; not a problem here

    3) Inputs outside common mode range ; not a problem here. 

    4) Power not on or not connected; not a problem here. 

    5) Input is not as simple as it seems. For averaging, I see 10k 15pF = 1MHz and 160 15pF 66Mhz low pass filters. Does Vavg have a large peak to peak AC component?