TPS2597: Power management forum

Having issues testing the overcurrent protection feature on the eFuse. Testing done by connecting the electronic load to the output of the eFuse (across VBUS and GND) and running it in Constant Resistance mode at various values of resistance to simulate different current levels. 

RILM = 1.74 kOhm, CITIMER = 1.5 nF, CDVDT = 1.5 nF, RDVDT = 0 Ohm

This is the plot of load current when load was turned on and off at different resistance values. 

The first four peaks are basically flat because the load current was below the overcurrent limit (3.3 A +/- 11%) and so the load was turned off after waiting around 30s.

The last peak is where the load resistance was set such that load current was well within the overcurrent range (around 3.2 A) and so the eFuse immediately triggers and turns it off within the ITIMER delay (1.14ms). This behaviour aligns with the expected outcome.

However, there are three peaks in between where the load resistance was low enough that load current was just within the range (minimum of 3.3 +/- 11% is 2.937 A). The eFuse regulates the load to around 2.25A and then 2A before shutting down, and for varying durations. This was tested across a few boards, and all showed the same behaviour. 

On connecting a wire to RILM resistor (where it connects to ILM pin) and measuring the VLM voltage through a scope, the following scope shots were seen -

Load current 2 A

Load current 2.3 A

Load current > 3A

Questions:

1. Why does the ILM current switch at such a high frequency at higher load currents? Is this a noise issue, or is it expected?
2. Could this switching be affecting the internal current measurement?
3. The eFuse operation meets our requirement of an overcurrent trip when within the 3.3 +/- 11% range but is still curious enough to spur an investigation.
4. Why does the eFuse regulate the load to around 2.25A and then 2A before shutting down?

Thanks in advance for any help you can offer!

Regards,

Meghana

Hi Meghana,

Thank you for this detailed description. 

Internally we have a senseFET with different width compared to mainFET for a scaled down current. This goes to RLIM pin for current monitoring and a comparator with fixed threshold at other pin is used for current limiting. 

Here for the 2A case, the waveform looks as expected. Is this current limiting variant or circuit breaker variant? 

Meghana Asha Krishnan said:

Could this switching be affecting the internal current measurement?

I am not fully sure if this is due to noise or due to internal operation. I will have to check with the team on exact operation during the blanking timer and current limiting. 

Meghana Asha Krishnan said:

Why does the eFuse regulate the load to around 2.25A and then 2A before shutting down?

This shutting down looks like thermal shutdown.

I will ask them about this after you provide me the confirmation on the variant. 

Best Regards,
Arush

Arush,

Thank you for your response. This is the TPS25974LRPWR, so the circuit breaker variant. Additionally, in our design, the eFuse does not have its own separate ground plane island on the board. Could this possibly contribute to noise at higher current levels?

Regards,

Meghana

Hi Meghana,

I will get back on this by end of this week. I am stuck in lab with some testings. 

Best Regards,
Arush