TPS7H5005-SEP: DCR current sensing for CS_ILIM

Hey David,

How did you connect the R/C in your schematic?

The only time I saw the CS_ILIM value spike up that fast I put the R/C backwards.

Thanks,
Daniel

Hi Daniel. Thanks for your reply. Firstly a correction: I'm using a TPS7H5006 evaluation board, not TPS7H5005 as I previously stated. Apologies for that but I don't think it should make any difference for this.

The resistor and capacitor are connected as shown below. This is from my simulation and I've confirmed that the components are connected in the same way in hardware, with the capacitor closest to the output capacitors (which are currently 2x 680uF). I believe this matches the schematic in the application note. The component values used were 100 ohm and 10uF and the CS_ILIM signal is connected through some jumper wire to the TPS7H5006 evaluation board. I've subsequently tried 1k and 10uF and then 1k and 100nF and in all cases the CS_ILIM rises rapidly, so I think something else is probably wrong.



In my simulation CS_ILIM would reach a maximum of just less than 1V, rising gradually over a large number of switching cycles. In hardware, I'm seeing the CS_ILIM signal shoot rapidly above the limit every switching cycle, as shown in the attached oscilloscope screenshot (CH1 (yellow): 7H5006 OUTA, CH4 (blue): 7H6013 HO, CH3 (orange): CS_ILIM). I can't see any obvious issues with my setup, and my understanding is that the CS_ILIM signal should simply reflect the current through the inductor, so I don't understand where this behaviour is coming from. Out of the box, the TPS7H5006EVM has a circuit connecting CS_ILIM to OUTA but I've removed all those components prior to connecting the DCR sensing circuit. Are there any internal connections to CS_ILIM or other issues with the controller configuration that might cause this?

Hey David,

The CS_ILIM pin is effectively a high impedance input during the start of the cycle until the output flips off.
There is a small discharge circuit after the OUTA/OUTB switch off.

What this means is the voltage doesn't shoot to 1 V without something external pulling it up that fast.
This meaning either there is still something connected to the pin, or there is damage to the device.
If changing the R/C values doesn't change the signal, than its probably not the R/C

Thanks,
Daniel

Hi Daniel. Thanks - I'll continue to investigate with that in mind.

I have some additional questions about the design in SLVAFJ8, please:

1. How is the current limit set for this configuration? The TPS7H500x-SEP datasheet gives an equation (eqn 10) for this, for a current sense resistor from CS_ILIM to AVSS but that equation doesn't seem to apply here.
2. Three parallel 0.001 ohm resistors have been placed in series with each of the output inductors. What is the reason for including these rather than just using the inductor's DCR?
3. I think there's a difference between the RC and L/R time constants with the chosen component values. Were these deliberately chosen to be different? If so, what was the reasoning for that?

Many thanks for your assistance with this.

Hey David,

1. There isn't a current limit. Its a downside of the current sense implementation

2. Increasing the inductor's DCR increases the size of the current sense signal into the CS_ILIM pin.

3. They are meant to be the same.

Thanks,
Daniel

Hi Daniel. Okay, that's clear. Thanks very much for your help on this.

I think the problems I'm seeing with the current sense signal may be due to poor ground connection between the two evaluation boards, so I'm looking at how to improve this. Does the TPS7H5006EVM have a ground plane layer?

Hey David,

The board has a ground, but its very cut up.
If you look at the users guide you can see the EVM layout.
This is because the EVM was originally the same daughter card we used for radiation testing.

Thanks,
Daniel