TPS3620: TPS3620

Hi Edward,

Can you provide some scope photos on VOUT, VDD, and VBAT?  What happens if the 220uF capacitor is removed from the circuit?  Does the TPS3620 flip to VBAT faster?  Any additional information provided may be very useful.  Thanks!

Ben

Hi

Trace 1 (Yellow) is VOUT, Trace 2 and 3 (Cyan and Pink) is VBAT, Trace 4 (Green) is VDD. All traces are 1V/DIV. In this plot VDD is removed instantly and drops to about 0.5V

In this trace VDD is switched off at the bench supply and drops more slowly

Here is the circuit. L38 is just a ferrite high impedance Z

Here is the circuit. L38 is just a ferrite high impedance Z

Hi Edward,

Thank you for the additional information.  Have you tried removing the RF choke L38?  I want to see if the TPS3620 is working properly before adding additional components to the design.

Ben

Here is the waveform with L38 removed. It looks to be working fine

Also /MR is connected to VDD

Hi Edward,

So removing the RF choke solved your issue?

Ben

Ben, it has fixed the fault only in the fact that the TPS3620 output is now disconnected from the load. Replacing the L38 with a shorting link brings back the problem! I am going to leave nMR floating and see what happens. Currently it is connected to VDD. Edward

HI Edward,

I dont think leaving the /MR pin floating will fix your issue.  The /MR pin is internally tied to VDD via internal resistor.  See below:

The /MR pull-up resistor value is calculated to be around:

Where the midpoint IIL can be used to calculate the internal pull-up resistor.

The issue is that the RF choke is inductance and current through an inductor cannot change instantaneously.  The equation for current through an inductor is:

V = L * (di/dt)

where (dt) is the time it takes for the current to reverse direction.  The time delay that you are seeing is the time it takes for the current to stop flowing from VDD to VOUT and change direction where VBAT becomes the source of VOUT.

Ben  

Hi Edward,

Please let me know if you have any updates with your application?  Thanks,

Ben

Ben problem was not with the TPS3620 but loading of supply, Br Edward

Hi Edward,

Glad that the problem has been resolved.  Can you close the thread by clicking on "resolved"?  Thanks and good luck to you!

Ben

Ben I am not sure if the TPS3620 is completely innocent. I disconnected L38 in our circuit which meant there was no load on the TPS3620. Then I connected a 330uF capacitor to VOUT of the TPS3620. I applied power to VDD pin and it charges the capacitor up. When I switch the VDD off you can see the voltage on the 330uF drops in-line with VDD! untill the threshold voltage, at which the switch over to VBAT occurs. Then the capacitor charges up to VBAT. This would prove that for a short while before switchover occurs current is able to flow from the 330uF cap back through the VDD pin? Pls see plot

Here VDD is green, VOUT is yellow

Hi Edward,

Is there a load on VOUT?  The dip on VOUT does not make sense.  In your previous pic, it showed that VOUT did not dip.  What is the  difference between the two scope shots?  Looks like VOUT dips around 300mV.

What is the voltage on VBAT?

Ben

Ben there is no load on VOUT. The previous pic shows VOUT with no-load. The difference between the twos scope shots is I have connected a 330uF cap to VOUT. VBAT is connected to the battery via a 100R resistor.

The only way 330uF cap can discharge would be back through the TPS3620. Pls note there is a 100R resistor between battery and VBAT

Is there a way for you to change the slew rate of VDD when VDD goes low?  My guess is that there is some delay where VOUT will back feed to VDD.  On the setup, can a diode be placed in series with the VDD power supply?

Ben

Ben unfortunately not the only thing we can do is replace L38 with diode Edward

Hi Edward,

Looking forward to your results with the diode.  

Ben

Ben the max VDD to VOUT is 400mA, so depending on how fast VDD falls then this could be exceeded. I am thinking that the disadvantage of the diode is the constant drop, what about putting a 2R2 in L38 position which would reduce the backflow into VDD e.g. for a 300mV difference between VDD and VOUT backflow would be reduced to 150mA

Hi Edward,

That sounds a like a good idea.  Keep me posted.

Ben

Ben what would happen if the battery were removed then VDD removed. Would the TPS3620 still switch over to the battery even though it is not there?

If I read Table 1 correctly, if VDD is above or below VIT and also VDD is above VBAT, then VDD is always connected to VOUT. This would mean if I removed the battery whilst VDD were present, then removed VDD the TPS3620 will not switch to VBAT but stay connected to VDD. This would cause VOUT to discharge completely into VDD? Do you agree with this?

Hi Edward,

Based on what you have said, the TPS3620 will not switch over to VBAT and will stay connected to VDD.

Once VDD goes below VIT-, the switch connecting VDD to VOUT will be turned off and my guess is that VOUT should stay up because the voltage on the capacitor.  

The timing diagram in the datasheet displays the same VOUT behavior as in one of the scope photos you showed two days ago.

Below shows one case where VDD is not greater than VBAT and VDD is not above the threshold voltage (VIT), then VOUT is VBAT.

So, these tables tell me that if you do not have a battery hooked up to the TPS3620, and VDD goes below the VIT threshold voltage, then VOUT is connected to VBAT where VBAT is floating so VOUT will be whatever the voltage seen at the output capacitor.  

I hope my explanation has been clear.

Ben

But if VBAT is floating and below VDD then the second row in the table says VDD>VBAT=1, VDD>VIT=0 VOUT=VDD. So VOUT will remain connected to VDD as long as VDD>VBAT do you agree?

Hi Edward,

Yes, I agree, according to the table.  Are you able to verify the function table?  I think the timing diagram is explaining what you are seeing on your setup.

Ben

Hi Ben, ok we are going to do some tests to confirm. I noticed on the datasheet pg3, that VIT is only specified for TPS3619-33 and TPS3619-50. Should the datasheet not say TPS36xx-33 and TPS36xx-50 to make it clear that this applies to both TPS3619 and TPS3620? Edward

Hi Edward, 

Yes, the VIT- specs should apply to the TPS3620 as well.

Ben

Hi Edward,

Are there any updates?  If not and you are satisfied with the answers, can you please close the thread by clicking on "resolved"?  Thanks!

Ben

Ben, Yes I have tested the circuit and can confirm if the VBAT is not present you get a heavy discharge back into the VDD supply: 

By putting a diode in L38 position the backflow current is blocked:

Edward

Hi Edward,

Thanks for providing the information.  This makes sense because the switch is a PMOS device.  The PMOS device has a body diode that connects the source to the drain.  This is the path of the current that flows from VOUT to VDD.  Without the VBAT voltage, the VDD to VOUT switch stays on and the body diode connection is causing your VOUT to drop.

Ben

Hi Edward,

Let me know if you have any other questions.  If not, can you close the thread by clicking on "resolved"?  Thanks and good luck to you!

Ben

Hi Edward,

Just following up with you.  Please let me know if you have any other question.  Thanks!

Ben