• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 62 subscribers
  • Views 1036 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LM5070-50 trouble on detection

Paul Martin5
Prodigy 70 points
Other Parts Discussed in Thread: LM5070

Hi, 

I've designed a custom board around the LM5070-50 PoE chip, but I'm having trouble from two separate PSE devices in detecting my PD, making me believe I've done something wrong with my circuit.  I've followed the reference design for the diode rectified isolated circuit almost exactly (I forgot the ground coupling cap between power and signal ground, but my understanding is this is just for EMI and noise reduction).  One of the PSEs I'm using gives an error message that says undervoltage (overload) fault, making me believe my circuit is drawing too much current in some location, but to the life of me I don't see any problems and DMM probes indicate there isn't any (obvious) short.  I've double checked cap and diode polarities as well. 

I realize this is a pretty vague question, but I'm attaching my schematic here in the hopes that some kind soul from TI can lend a hand.  The ethernet connector I'm using here (not pictured) is a bel fuse 0813-1x1t-57-f 10/100 MagJack PoE. Thanks in advance, I truly appreciate it. 

Followup Question:

It seems most PoE PSE devices provide power as (+) = 0V, and (-) = -48V. I'm having trouble understanding the voltage requirements of the LM5070---is it compatible with Vin = 0, Vee = -48V?

0677.poe.pdf

  • 0
    Prodigy 70 points

    To add some more details, it appears that when 10V is applied across Vin/Vee, 70mA is consumed and the PSE signal is loaded.  Looks like without the LM5070 in place, one of my PSEs is applying 10V while waiting for Rsig and then aborting.  With the LM5070 in place, the PSE is applying 10V to check for Rsig but this is being loaded and reduced to about 4V somehow and then the PSE aborts.  

    I could really use some help with this. I plan to order about 20 more of these boards, but I'm stuck at stage 1 here. I've even overnighted more LM5070s to solder on, thinking I might have a bad IC, but that didn't help. 

  • 0 in reply to Paul Martin5
    TI__Mastermind 42795 points

    Hi Paul,

    Does this occur on a single unit or multiple? Meaning, could it be a single failure of the LM5070 device or a design related failure (occurs on multiple units)?

  • 0 in reply to Eric Wright
    Prodigy 70 points
    I only have one PCB prototype manufactured right now, but I've tried replacing the LM5070 with a new one and observed the same behavior, so most likely it's a design error on my part. Cross checking my schematic with the suggested schematic in the data sheet, however, I can't seem to find any difference. I've triple checked all part footprints and polarities as well.
  • 0 in reply to Paul Martin5
    TI__Mastermind 42795 points

    Have you tried operating from just a DC 48V power supply? You can connect +48V to TP1 and 48V GND to TP2 and check whether the PD portion (hotswap of LM5070) and DC/DC portion operate. If that operation seems fine, then operation from a PSE can be further investigated.

    Also, for PSE operation can you provide the voltage waveform taken between TP1 and TP2 during start up? This can help identify which portion of the PSE-PD negotion is failing. For example, if detection of the 24.9k RSIG is failing then the waveform should only be in the 0-10V (detection range) range and then abort.

    Another possibility is that the PSE briefly turned on the PD but then turned off because the load current from the PD was not enough (the PSE can disconnect if the load current is below ~10mA).

  • 0 in reply to Eric Wright
    Prodigy 70 points
    Thanks for your replies--I really appreciate it. I unfortunately don't have a supply more than 25V in my lab, so I can't hook up to a 48V supply. I did however hook up to a 10V and 20V supply across TP1 and TP2. With 10V, there was a current of 70 mA, and with 20V the current was not measurable. I couldn't get the voltage high enough to just check the DC-DC conversion portion, unfortunately.

    Rather than take a screen cap from the scope, I can pretty easily describe what the waveform looks like: There is a short pulse of about 2.5V from the PSE across the input to my diode rectifiers. The PSE (www.phihong.com/.../PSA16U-480-POE.pdf) then blinks a fault message indicating "Low voltage sensed during detection (overload)," and indeed if I test the output of the MagJack with no circuitry connected the pulse goes to about 4V to begin with and then (with no overload fault) immediately jumps to about 9V for Rsig classification. In other words, the PSE is not even getting to Rsig stage with my current circuit--something else is loading the voltage too much.
  • 0 in reply to Paul Martin5
    Prodigy 70 points
    Eric,

    I've found the problem. The RSIG value was 24.9 (I had ordered the wrong P/N even though my description said 24.9k)---stupid mistake on my part. The RSIG was SO far off that rather than get an incorrect RSIG error, I got an overload fault. Thanks for helping me think through this!