• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 21 subscribers
  • Views 883 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

CD4504B booting

Gianluca Pagano15
Prodigy 20 points
Other Parts Discussed in Thread: CD4504B

Hi,

I have an issue with this IC. The IC is connected to an MCU operating at 5V. The IC function is to drive a half bridge consisting of a NMOS-PMOS couple with common drain. So the power mosfet gates are separated and connected to two different output pins of this IC.

Sometimes, when I apply the power to the electronic board, one of the NMOS-PMOS couple is shortcut because it may happen that voltage is high enough on NMOS gate and low enough on PMOS gate putting the board on a sort of latch-up because the PSU does not increase the voltage anymore due to excessive current absorption. Adding a capacitor at the input reduces drastically the probability of this event.

I have 3 hypothesis for this:

1. The driver IC is not working while input power transits from 0V up to 5V, so the status of the output is unknown but output impedance looks good enough to keep the same voltage level.

2. The driver IC is correctly working at lower voltages keeping all outputs safe at zero but the MCU is in unknown state. Therefore some output pins of the MCU may be at 1 while others may be at 0.

3. The ripple given by the PSU's voltage ringing may turn on the PMOS and NMOS thanks to the voltage differences created by the picks.

When this happen the lab. PSU goes in current protection and supply voltage level is around 3.5V to 4V: the minimum level to turn on the MCU, this suggests the second hypothesis is more realistic. I believe this issue can be solved adding pull-up/down resistors on the gates, but I need to understand where: at the MCU pins or at the power mosfet gates pins.

Can you exclude the first hypothesis to confirm the second? CD4504B datasheets are not accurate enough to specify if the IC has an undervoltage protection.

Thank you.

Regards,

                          Gianluca

  • 0
    TI__Expert 6335 points
    Gianluca,

    I've assigned this to the appropriate engineer. They will get back to you. Thanks.

    Regards,
    Nirav
  • 0 in reply to Nirav Patel
    TI__Guru 62250 points
    Hi Gianluca ,

    All devices have ESD protection diodes to gnd . They help in suppressing undershoots but isn't quick enough to turn on for fast transients .
    The recommended voltage supply conditions for this device is from 5V to 18V and the behavior is not guaranteed under that . This doesn't mean that it would stop functioning below 5v but that the specs might be differ from the values on the electrical spec table under 5V .
    I would assume that the transistors internal to IC turn on around 3V node but with higher output impedance and prop delays .
    The output state will be unknown or indeterminate until fully powered up /valid inputs .
  • 0 in reply to ShreyasRao
    Prodigy 20 points

    Hi,


    yes, I understand that the datasheets do not specify behavior at low voltages, that's the reason I am writing here, I need a way to boot the power up with ICs outputs on a known safe state.

    In the while I had this new idea: I could connect the IC's low level supply voltage VCC to a PMOS drain with gate driven by the MCU and source connected to 5V rather than my actual design with IC's low level permanently connected to 5V. This way I could keep this part of the IC off as long as all transients have gone. Looking at the schematic of the IC, giving 0V to the first stage buffer inverter will cause to have 0V also to the inverter output, therefore also 0V after the level translator and finally VDD at the second inverter stage. This will keep the NMOS on but the PMOS off making impossible a shortcut on the half-bridge. Since my design already contains this PMOS for different purposes I could just use it without increasing board's cost.

    Can you tell me if this will work? In the datasheets it is nowhere specified this is allowed nor how the IC would perform connecting 0V to the first stage while the second stage is supplied at VDD.

    Kind regards,

    Gianluca

  • 0 in reply to Gianluca Pagano15
    TI__Guru 62250 points
    Hi Gianluca,

    could you please send me your schematic and any other scope shots available .