• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 62 subscribers
  • Views 163 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.

TPS2373: MPS doesn't appear to engage

Aaron Lane
Prodigy 90 points
Part Number: TPS2373
Other Parts Discussed in Thread: TPS2372,

Hi there,

I have a prototype utilising the TPS2373-4RGWT. The plan is to utilise it with a 802.3bt situation in the future, but as I understand it, the .bt standard is also backward compatible. Can you please confirm this is the case with this chip? My issue is that my design has a very low start-up draw (<10mA) for a few seconds, and the chip doesn't appear to be maintaining the PoE negotiation. If I put a little extra load on the load side of the regulator (after the PoE chip), it appears to work without an issue every time. I was under the impression that the MPS feature was supposed to keep PoE alive in low power situations? Setup is as per below.
DEN - 24k9 - VDD
CLSA - 63R4 - VSS
CLSB - 90R9 - VSS
REF - 49k9 - VSS
AMPS_CTL - 1k3 - VSS
AMPS_DUTY - SHORT - VSS
APD - SHORT - RTN
VC_IN - Floating
VC_OUT - Floating
UVLO_SEL - Floating

The regulator isn't turned on until PG is active, and the #BT, TPH, TPL outputs are monitored to ensure the load remains less than the available power.

I am currently using an 802.3at switch. It powers for about half a second, then drops power, then renegotiates again.

There doesn't appear to be any voltage applied across the AMPS_CTL resistor at any time.

Input voltage - Approx. 53VDC

Idle current - Approx. 10-14mA @ 53VDC

"little extra load" 150R @ 3.3V.

Appears to raise the current draw by 2ma when measured on the 53VDC side.

There is a second regulator also powered by the PoE chip, but it's controlled by the MCU. Apart from the input side capacitors on it, it doesn't seem to be in play.

I've tried putting a decent sized capacitor (2200uF) on the load side of the 3.3V reg, but it didn't seem to make a difference.

I've tried removing both the AMPS_DOTY and APD shorts, but it doesn't seem to make a difference.

Can you please assist and perhaps offer some suggestion as to what might be happening, and how to enable MPS to keep the PoE enabled?

Kind regards,

  • 0
    TI__Mastermind 21320 points

    Hello, 

    First, the MPS signal for the .bt and .at signal is different. You currently have it configured for a .bt PSE, so your .at PSE is not receiving enough current for the higher MPS signal of the .at standard. Please see Table 4 in the data sheet and section 7.4.7.

    You will probably need a blocking diode. The MPS signal is dependent on many factors. 

    Additionally, some .at PSE's only monitor the AC signal, which may remove power in this case. 

    Also, the VC_in and VC_out pins cannot be left floating. You can use the TPS2372 which does not have these pins. 

    Yes this device is backwards compatible -- the fact is can negotiate proves this fact. 

    If this post answers your question, please indicate so by marking this thread as resolved. Thank you.

     

    Regards, 

     

    Michael P.

    Applications Engineer

    Texas Instruments 

  • 0 in reply to Michael Pahl
    Prodigy 90 points

    Hi Michael,

    Thanks for your prompt response.

    If I'm reading section 7.4.7 correctly, I thought that the chip would be putting load on the AMPS_CTL output pin if there's less than ~28mA load. This doesn't seem to be happening.. From what I understand, it automatically determines whether Type 1-2 or 3-4 PSE is being used, and apply the appropriate MPS signal?

    From what I undersand with table 4, is that the duty cycle will automatically be 26% if it detects PSE type 1 or 2. It doesn't appear that AMPS_CTL is ever outputting to increase the load. Based on using a 1k3 resistor with a 53VDC load, I would have expected an extra load of around 40mA, unless it's happening so quick that my meter isn't picking it up? The MPS_DUTY section is blank for the PSE type 1 & 2 line. Does this mean that in theory it will automatically geneterate the pulses at a 26% duty cycle?

    When you mention the blocking diode, can you please elaborate on this? I am using a SMBJ58A-13-F directly prior to the chip across VDD and VSS.

    Thanks for the suggestion with the TPS2372. I'll look into it. I assume I'll have the same issues though if the above isn't sorted?

    Kind regards,

  • 0 in reply to Aaron Lane
    TI__Mastermind 21320 points

    Yes the TPS2373 can identify the PSE Type and should adjust the MPS signal accordingly. 

    The blocking diode is inserted between the IC and the bulk cap. So connect a diode between VDD and the bulk cap. 

    Again, VCin and VCout should be dealt with. VCin is the input power for the IC, so without it the IC is unpowered for normal operation, so that is probably the source of the problem. 

    If this post answers your question, please indicate so by marking this thread as resolved. Thank you.

     

    Regards, 

     

    Michael P.

    Applications Engineer

    Texas Instruments 

  • 0 in reply to Michael Pahl
    Prodigy 90 points

    Hi Michael,

    Thanks for that. I have ordered a TPS2372-4 to swap it out and give it a go. Thank you for your assistance. Can I hold off on the resolution buttons until the replacement chip arrives?

    Kind regards,

  • 0 in reply to Aaron Lane
    TI__Mastermind 21320 points

    Aaron, 

    You can always post on a thread, even if it is closed. Thanks!

    If this post answers your question, please indicate so by marking this thread as resolved. Thank you.

     

    Regards, 

     

    Michael P.

    Applications Engineer

    Texas Instruments