E-fuse parts with over-current warning period feature?

[HwAoRrDk]

Does anyone know of any e-fuse / hot-swap controller / load switch parts that are similar to the Richtek RT1720?

The key feature I'm most interested in is the fault warning timer functionality. This is where, upon an over-current condition (also over-voltage, but I'm not interested in that) it will start to limit current, assert a fault flag output, and also start a timer (whose duration is dictated by a capacitor). If the over-current condition persists until the timer expires, then the load is shut off, and another (longer) timer period ensues before it resumes. Basically, a warning to the load "Hey, you're going over the current limit, quit it!".

I'm finding it very difficult searching for a feature like this, because it's not as simple as just ticking a box on a parametric search. I have found a couple of parts from TI (e.g. TPS25970ARPW) that have similar, but not equivalent, functionality. They feature a transient over-current timer, but don't limit current or assert a fault output until the timer expires, not when it begins, like I want. Also they only come in annoyingly minuscule 2x2mm DFN packages.

Other requirements:

- Vin >13V
- Iout >6.5A
- Enable pin
- Soft-start
- Internal or external FET, doesn't matter
- Over-voltage protection not necessary, but welcome
- Preferably non-microscopic package (e.g. SO-8, 3x3 DFN, etc.)

Bonus points if the part also has a configurable power-good output. The RT1720 has one, but it's not useful for what I need; it's whether VOUT is 'good' relative to VIN, rather than an absolute voltage level.

[HwAoRrDk]

Oh, and tangentially related, a quick rant about MOSFET specs:

I'm looking at MOSFETs to go with the RT1720's need for an external one, and a requirement is that the FET's SOA needs to handle the configured fault warning timer period under short-circuit load condition - e.g. for a 6.5A current limit, I_D = 6.5A, V_DS = 12V. So, say I want a 3ms fault timer period, I need a MOSFET whose SOA allows greater than that at the given current/voltage.

But MOSFET SOA specs are all over the place for parts which are ostensibly the 'same' model with otherwise 99% identical specs. For example, the common 4410, which over a dozen manufacturers make clones of, half the parts have a SOA chart that shows 6.5A/12V as being safe for up to 10ms, but others for only 1ms, or even less! No idea who to believe, and makes it very difficult to multi-source or substitute parts.

[TimNJ]

https://www.analog.com/en/products/lt4356-1.html, and many other LTC/ADI variants. I think LT4356 was the original...or one of the earlier ones. TI also has a bunch I think.

I had inquired about RT1720 with RichTek before, and IIRC, they said it was planned EOL.

Nexperia has a bunch of "enhanced SOA" parts which might be of interest: https://www.nexperia.com/products/mosfets/application-specific-mosfets/asfets-for-hotswap-and-soft-start#/p=1,s=0,f=,c=,rpp=,fs=0,sc=,so=,es=

[HwAoRrDk]

https://www.analog.com/en/products/lt4356-1.html, and many other LTC/ADI variants. I think LT4356 was the original...or one of the earlier ones.

Seems like it would do the job. Amusingly, the MSOP-10 package is actually pin-compatible with the RT1720, so I guess maybe Richtek copied LT? Haven't read the datasheet in detail to know if they both have exactly identical behaviour. Of course, it's LT though, so expensive as hell...

I had inquired about RT1720 with RichTek before, and IIRC, they said it was planned EOL.

That's disappointing. Still says "active" on their website, though.

I am starting to think perhaps I could instead replicate the necessary over-current protection, limiting and warning functionality with a discrete current sense amplifier and a small microcontroller (with ADC watchdog and/or op-amp comparator), which would be handy as I have some other logic that the MCU could also incorporate too. Would need to use a P-channel MOSFET instead of N for ease of driving.

[TimNJ]

I had inquired about RT1720 with RichTek before, and IIRC, they said it was planned EOL.

That's disappointing. Still says "active" on their website, though.

I am starting to think perhaps I could instead replicate the necessary over-current protection, limiting and warning functionality with a discrete current sense amplifier and a small microcontroller (with ADC watchdog and/or op-amp comparator), which would be handy as I have some other logic that the MCU could also incorporate too. Would need to use a P-channel MOSFET instead of N for ease of driving.

Yeah, and I had asked maybe 3-4 years ago now, so..  I think I had asked that particular question because I was designing for a customer who requested a relatively long lifecycle, like 10+ years.

It's definitely not rocket-science if you're willing to go through the effort.

[temperance]

Did you bother to search for those yourself? Probably not. TI, ST, On Semi, Rohm, AD,... and many others make those in all flavors.

TI TPS1Hxyz, TPS2HBxyz series would fit your requirement including the request for a non QFN.

[HwAoRrDk]

Did you bother to search for those yourself? Probably not. TI, ST, On Semi, Rohm, AD,... and many others make those in all flavors.

Yes, I did. And it is an impossible task without spending hours and hours reading every single datasheet for dozens and dozens of parts - because as I said earlier, the specific behaviour I am looking for is not described in headline feature text or parametric filters. So I thought I would ask on the off-chance someone else is familiar with such parts.

I hadn't looked at AD/LT/MAX because I typically always disregard them due to very high prices, but thanks to TimNJ for telling me about LT4356. As mentioned above, I found a couple of TI parts that nearly, but not quite, fit the bill. I also found the Renesas ISL6115A, which again is almost what I want, but doesn't have any warning output signalling the current-limiting period is in effect. The Rohm BV1HD045EFJ-C is also, again, nearly what I want, but doesn't toggle status output until 'hard' current limit of 21A is hit, and the variable current limit can only be set up to 5.4A anyway. Unless I missed something, ON Semi, ST, and Diodes don't seem to have anything suitable.

TI TPS1Hxyz, TPS2HBxyz series would fit your requirement including the request for a non QFN.

The TPS1H doesn't have the over-current fault warning period functionality that I'm looking for; it just shuts off the load immediately upon over-current. The TPS2H is completely unsuitable (2-ch, SPI interface, min 10A current limit, etc.).