TIOS1013: Digital Input Clamping When Un-powered (TIOS1013/TIOS1015)

Hi Martin,

The TIOS101x family of devices are powered primarily through the VCC pin and it uses an internal LDO to power the digital core or Diagnostics and Control block shown in the datasheet block diagrams. 

For the TIOS1013 and TIOS1015 devices, the internal LDO is also used to set the Control Logic voltage levels to match the MCU with the respective 3.3V and 5V parts.  This LDO can also be used to supply other devices in the circuit. 

The TIOS101 is also powered from the VCC but uses an external voltage reference to be supplied on the VCC_IN pin to set the digital signal levels.  But this external 3.3V or 5V voltage is not used to power the digital core and the device still requires VCC to operate.  There is not a required power up sequence between the VCC and VCC_IN, and VCC_IN can be powered before VCC such as in your case where this is coming from a field-side supply.  This should not be a problem.

The IN and EN pins should always be within the abs max limits, but otherwise, there are no restrictions.  The High/Low logic levels are also such that they should work with either a 3.3V or 5V logic signal, so either device could be used in your application.

Regards,

Jonathan

Hi Jonathan,

Thanks for the speedy reply and I think you have confirmed what I believed to be correct, that is -

The VIL and VIH logic levels (IN & EN) are always 0.8Vmax & 2.0Vmin irrespective of the VCC IN voltage (TIOS101) or internal LDO voltage (TIOS1013/TIOS1015) and will therefore always work with our 3.3V logic signals irrespective of the device used.

With the TIOS101/TIOS101x completely unpowered (No VCC or VCC IN) we can still apply our 3.3V logic signals to IN & EN without damaging the device or loading down the logic drivers which feed IN & EN (as long as they are within the -0.3V to 6V absolute max range). We will probably default our logic to the low state in this scenario anyway, possibly add series resistors too.

A further valid scenario is our system is powered off, however the field-side supply remains connected and powered (24V). This would effectively keep the TIOS1013/TIOS1015 powered up however the IN & EN inputs will be floating since our system is unpowered. I assume from your reply this is also an acceptable safe condition.

For the TIOS101 configuration there will be no VCC IN so a low Vf diode in series with our supply would prevent any back-feeding if this is deemed necessary. For the TIOS1013/TIOS1015 the internal LDO powers everything (from VCC, VCC IN diode indicated above would be removed to disconnect VCC IN/OUT from our supply) and as we aren't using VCC OUT this shouldn't cause an issue. IN & EN will have pull-down resistors to terminate the floating inputs under this condition.

Obviously normal operation will only resume once a valid VCC (and VCC IN) voltage are present.

Regards

Martin

Hi Martin,

Everything in your message is correct.  I want to note that if you use a TIOS101 and do not provide a VCC_IN voltage, you will want to add a Vf diode because the device will back feed a small voltage out the VCC_IN pin. The LDO and non-LDO versions of this device share a core design and the device monitors for a voltage on the VCC_IN pin to determine if it is a TIOS101 non-LDO version, or a TIOS1013/5 LDO version.  If it does not sense a voltage, it will try to enable the internal LDO to source out the VCC_OUT pin.  However, there is a different bond wire configuration on the TIOS101 that prevents the LDO from sourcing a large current through the VCC_IN pin.  However, there is a very small leakage current path in the monitoring circuit that can slowly charge the external decoupling caps.  Once these caps charge to approximately 3V, the monitor circuit detects this voltage and disables the LDO because it is actually a TIOS101 non-LDO device.  Once the caps discharge, the process repeats. 

The TIOS101 was not designed to operate without both supplies present, and there is no damage to the TIOS101 device from a powered VCC and unpowered VCC_IN pin.  The voltage will also not raise above approximately 3V and will be less than the 3.3V supply level.  However, you will want a Vf diode to prevent back-feeding to the other devices.  This "feature" was removed in our new TIOS102 device that is now available as a preview and will be fully released soon.  The TIOS102 is footprint compatible with the TIOS101 with only a slight increase by 0.5mm in the width, but keeps the same pin function and spacing.

Regards,

Jonathan

Hi Jonathan,

Thank you for your detailed explanation and confirming the device operation under the various powered/unpowered scenarios which may exist in our application. Looks like we can certainly go ahead with incorporating these devices into our design.

One other question...is the output current actively current limited to the value set using the RSET resistor, i.e. if RSET is set for a 100mA current limit, during a short circuit on the output will the output current be limited to 100mA for the blanking time before the output is turned off and the auto retry activated?

To be honest the new TIOS102 device would have been my preferred choice because, with the minimum VCC requirement of 4.75V and the lower residual voltage levels, it would allow a 5V push-pull option in addition to supporting industrial 24V switching. We have a 5.5V supply which we could diode-OR onto the VCC pin to support this. Unfortunately, it's not available at this time but I'm certainly going to ensure the design will allow for this for future use assuming there isn't an issue with the small footprint differences.

Do you have any idea when the TIOS102 will be finally released for production?

Regards

Martin

Hi Martin,

You're welcome, I'm glad to help.  There is a difference between the TIOS101x and TIOS102x devices with respect to the current "limit."

The TIOS101x uses the "limit" as a threshold that if exceeded for longer than current fault blanking time of approximately 200uS will disable the output driver, but it does not actively restrict the current to the 100mA "limit" as an example.  This can be seen as a benefit in some applications allowing faster charge times for capacitive loads  that need larger inrush current.  But in the case of a short or fault condition, the device will still disable the driver.

However, the TIOS102x will actively restrict the current to the set "limit" for the current blanking time of approximately 200uS and would only allow 100mA to flow in your example. This may be seen as a benefit because it will limit the maximum current to a set level.

The trade off is the active monitor loop can reduce the maximum capacitive load that can be driven because it will take longer to charge and it may exceed the 200uS window.  Also, there is an increased risk that some complex inductive and capacitive loads may become unstable and oscillate with an over-current enable/disable scenario due to the current changes associated with the phase delays in the L/C elements.  This is rare and more of a general risk with any current monitor loop, but it is one of the reasons the TIOS101x did not actively restrict current.

From the last information I have, the TIOS102 should be released in about 1 month, or roughly the end of October, beginning of November time frame.

Regards.

Jonathan

Hi Jonathan,

That's interesting to know regarding the current "limit". I guess the main thing is the TIO101x can survive a permanently shorted output without damage plus I can see the benefit of not clamping the current to the "limit". It looks like the TIOS102 data sheet needs updating to indicate the slightly different functionality as I don't believe I've seen this explained in it.

I'm going to proceed with a design which will allow any of the TIOS101x or TIOS102x devices to be fitted, that way we can try all devices and increase the available devices which we can use in the final product. I know the footprints are ever so slightly different but I'm hoping this will work out fine...now I won't ask why TI have used two different packages...lol.

The TIOS102 is interesting because of the lower VDSon and more importantly the lower minimum VCC requirement. In our application VCC will be normally 24V (10V - 30V nominal range) however if 5V can be supported that would be an extra bonus for us.

The TIOS102 release dates look ok for us, we will probably be prototyping near the end of this year, possibly sometime in December. If samples are available they would be useful although each board will require 16 devices...I'm also hoping the lead times start to come down as these are playing havoc when it comes to designing our boards.

Once again thank you for the assistance, you have been very informative.

Regards

Martin