[Urgent] Unexpected error in LTM4613 parallel operation

Hi

I'm DFAE in Japan. Our customer is reviewing unexpected error in LTM4613 parallel operation like Figure 21 on D/S. Because LTM4613 is burned under their evaluation. Could you please let me know what happen if the pin is opened ? 

Assume that only one of the two LTM4613s has the phenomenon as follows. If this happens, are there any possibility that LTM4613 will burn?

1. One of FB open.

2. One of RUN open.

3. One of COMP open.

4. One of TRACK/SS open.

5. One of MARG0 open.

6. One of MARG1 open.

7. One of PLLIN open.

Regards,

Hiroyuki

 

  • +1
    •  Analog Employees 
    on Nov 1, 2019 4:51 PM

    Hi Hiroyuki. D, 

    1. If one FB pin is open, then with 0.6V reference and bottom FB resistor of 2.61V, the top FB resistor is now 100k so Vout programmed is 23.6V which exceeds Vout abs max (16V). The other channel with its FB pin open will see that its FB voltage is low so will also try to drive the output higher. If Vout rises above the abs max 16V then both parts can potentially be damaged. 

    2. Not expected. 

    3. Not expected. 

    4. Not expected. 

    5. Not expected.

    6. Not expected. 

    7. Not expected. The module with PLLIN pin open will operate at default operating frequency 600kHz, same as the other module which is synchronized to the external clock, which is not concerning, the switching frequency just won't be interleaved between channels so you don't get the extra benefit of ripple current cancellation and minimized possibility of beat frequency issues. 

    Note the LTM4613 is pin compatible with the LTM4601. The LTM4601 has a poly-phase single output demo board (DC1043A-A) for 2 modules paralleled. With the LTM4601 module swapped for the LTM4613 this setup can be used as another testing reference. 

    Thanks. 

  • Dear Young,

    Thank you very much for your quick response. In this case, output overvoltage protection won't work ?

    Regards,

    Hiroyuki 

  • Dear Young,

    Sorry for my pushing. Tomorrow I'll have a meeting with the customer about the issues. Could you please let me know your advice about output overvoltage protection ?

    Regards,

    Hiroyuki

  • +1
    •  Analog Employees 
    on Nov 7, 2019 6:37 PM in reply to Hiroyuki.D

    Hiroyuki, the overvoltage protection can't help in this case because your specified output voltage is 23.6V, this is defined by the feedback resistors. The LTM4613 won't take this as Vout overvoltage condition. But 23.6V is already higher than the maximum Vout the uModule allowed.