Post Go back to editing

LT7809 eval board and our board has startup problems when device is cold

Thread Summary

The user reports unexplained rising output voltage on the LT7809-AZ eval board and a custom board when the chip is below room temperature, with the voltage stabilizing after 20-50 seconds. The issue is resolved when the chip is warmed above 25°C. The problem occurs with both external and internal voltage references, and the output voltage is approximately 26V with an input of 45-50.4V.
AI Generated Content
Category: Hardware
Product Number: LT7809

With both the eval-LT7809-AZ as well as our own board much like the eval board I see unexplained rising output when the chip is below room temp

The voltage will get to target voltage but it can take 20 to 50 seconds.  The voltage ramps up to some value such as 15V then gradual does jumps up (and back down) several volts over many seconds

If I warm the LT7809 mildly to just above 25C room temp everything works well including expected slow start ramp of 20msec.  

I have used the eval board and our own board on-board voltage source for EXTVCC or use chip internal reference, both seem to have same issue.  This is AFTER the choice of voltage to use for bias and may be some internal chip reference.

If I cool down the chip or even have it start up first time in the morning where room temp is 20C or lower I can see unexpected output rise.

This is on the eval board as well as our own board..  In both the eval and our own board this happens.  

The chip has code  B61779 in both cases.

Is this a known issue with next rev chip planning to look at this or is thing new and unknown issue to your team?

  • We use an input of 50.4V max or down to 45V with same result.   Our output voltage is about 26 volts.

  • Can you provide scope captures of the issue?

  • Hi Mark,

     

    Please tell me the simplest unmodified Eval Board test conditions/setup to duplicate the issue and which waveforms are wrong. Please send scope captures if possible.

     

    We have engineers trying to replicate the issue now. Is the SS pin resetting on startup?

     

    Best,

    Vanessa Roque

  • Original problem on our board then I duplicated on unmodified eval board.  I will only discuss eval board here.

    Problem first seen when I was cooling part with freeze mist to start another test after parts were hot.

    Problem only happens if Lt7809 has been hit with freeze mist and some parts around it would have been hit as well.  Slow start and clean output when all parts room temp or hotter.

    Eval Board Jumpers:  

    Please clarify how to determine this that you ask:   'Is the SS pin resetting on startup'  I will then do things and note the SS pin.

    I will next try to cool only the Lt7809 in a more controlled way with ice-in-a-bag held only to the Lt7809 to avoid this being issue with other parts and/or freeze mist residue.

    First saw this with 10milliOhm shunt however it happens now with 5 milliOhm shunt as well.

  • Hi Mark,

     

    We believe the issue you are experiencing is caused the moisture forming around the IC, rather than temperature alone. 

     

    The product team conducted the following tests:

     

    1. Continuously monitored the output while applying freeze spray to the IC. The output continued to regulate normally.

       

      2. Stopped applying freeze spray while continuing to monitor the output, The IC would thaw and moisture formed around the IC. The output then began to drift, similar to the behavior you observed.

       

      3. Performed a controlled test in an environmental chamber, lowering the temperature without introducing moisture. The device continued to operate normally.

     

    CONTROLLED TEST CONDITIONS:
    VIN=48V, VOUT=12V, RLOAD=1kΩ, TEMP=-40C
    (JP2) MODE=FCM, (SW3) RUN=ON, (JP1) PLLIN=FIXED, (JP7) PULSE POLARITY=COMP, (SW2) INT POWER=ON, (JP4) VOUTADJ SOURCE=INT

     

    Thanks,

    Vanessa Roque

  • Ok, I will try with the method I mentioned where no moisture would form.  I have trusted freeze mist for decades but perhaps there is a high impedance point that is being effected.    

  • You mention a temperature chamber. This is beyond my means but is indeed the proper way to check the part from cold start.    Can you cool the entire board in a powered off state do the lowest temperature the LT7809 should still operate at and then take a trace of application of a very fast rising Vin and monitor the output voltage.  I do not have a temperature chamber like I used to in the old days of my HP hardware design.    The ideal case would be directly while full board is in temp chamber to cold temp apply battery voltage of 50 volts using powerful 12 cell lipo batteries.     If you cannot use a 12 cell Lipo battery the next best this is the fastest rise time your voltage source can supply, not the extremely tame slow rising Vin you show in above trace.     Thank you.

Before You Switch


Switching languages will make ADI Explorer unavailable. Resume your session by switching back to English and reopening ADI Explorer.