hwmon: add driver for ltc4284 by actorreno · Pull Request #3450 · analogdevicesinc/linux

actorreno · 2026-07-21T08:33:05Z

PR Description

The LTC4284 is a high-power hot swap controller designed for -48V distributed power systems supporting up to 2500W applications. It features dual-gate MOSFET drivers with multiple operation modes, comprehensive monitoring capabilities, and advanced protection.

Datasheet: LTC4284

This PR is for upstreaming
This is for dev branch in case some tests/features are requested: #3351
This supersedes the old draft PR #3126

Key features:

Dual current sensing with independent sense resistors
Voltage, current, power, and energy monitoring
Four operation modes: single, parallel, staged start (high/low stress)
Programmable current limiting (15mV-30mV sense voltage range)
Min/max history tracking with reset capability
Comprehensive fault detection and logging
I2C interface with regmap support

PR Type

Bug fix (a change that fixes an issue)
New feature (a change that adds new functionality)
Breaking change (a change that affects other repos or cause CIs to fail)

PR Checklist

I have conducted a self-review of my own code changes
I have compiled my changes, including the documentation
I have tested the changes on the relevant hardware
I have updated the documentation outside this repo accordingly
I have provided links for the relevant upstream lore

nunojsa · 2026-07-28T12:20:42Z

The ltc4283 driver is already upstream! Any reason this device cannot be supported in that driver?

actorreno · 2026-07-29T02:08:53Z

I was pulled late into the project (around May), so I simply continued work until I ended up being the submitter.
(LTC4283 got upstreamed around June it seems)

I wasn't aware of ltc4283 being in the works, and I think we can't easily add on ltc4282.

Digging some history there's this #3126 that seemed to have decided why we have a separate driver code.

but now that I see LTC4283 I am inclined to just add support on it, the code overlap seems considerable compared to probable extra code lines.

actorreno · 2026-07-29T22:59:47Z

v2:

now adding support on top of ltc4283.c rather than its separate driver

For the gpio support, I was thinking since it's exactly the same feature, just mentioning it in kconfig was enough

The LTC4283 is a negative voltage hot swap controller that drives an external N-channel MOSFET to allow a board to be safely inserted and removed from a live backplane. Note that this device has a MODE pin that is determined by differing voltage levels (VEE, VIN, INTVCC, or left open) and cannot be controlled via software or GPIO. Signed-off-by: Alexis Czezar Torreno <alexisczezar.torreno@analog.com>

Support the LTC4284 Hot Swap Controller. It is very Similar to LTC4283 but for higher power. The device features programmable current limit with foldback and independently adjustable inrush current to optimize the MOSFET safe operating area (SOA). The SOA timer limits MOSFET temperature rise for reliable protection against overstresses. An I2C interface and onboard ADC allow monitoring of board current, voltage, power, energy, and fault status. Signed-off-by: Alexis Czezar Torreno <alexisczezar.torreno@analog.com>

Similar to LTC4283, LTC4284 has up to 8 pins that can be configured as GPIOs and behaves the same. Signed-off-by: Alexis Czezar Torreno <alexisczezar.torreno@analog.com>

actorreno · 2026-09-18T01:29:29Z

PR got stale, just following up on review @nunojsa
Also cleaned the commits.
Thanks!

nunojsa

I think the major thing to rethink is sense1 and sense2. I think it should just be a new sense2 variable! The docs are actually wrong IMO as it looks like we can ever have curr1, curr2 and curr3 which is not the case!

nunojsa · 2026-09-25T12:47:40Z

+    description:
+      Value of the SENSE2 current sense resistor. Only applicable to LTC4284.
+      This resistor is connected between SENSE2+ and SENSE2- pins and is used
+      for channel 2 current measurements.


We should not need those "only applicable to ltc4284". The below conditional stuff should make it clear!

I see, will lessen comments like this.

nunojsa · 2026-09-25T12:48:12Z

+    then:
+      required:
+        - adi,rsense1-nano-ohms
+        - adi,rsense2-nano-ohms


Would be nice to also add an example for the new chip

nunojsa · 2026-09-25T12:48:40Z

+      For LTC4284, this is optional and represents the sense resistor value for
+      current measurements via the ADC+/ADC- pins (channel 0), which depends on
+      the hardware connection (can be wired to SENSE1, SENSE2, or an average via
+      resistive divider).


I would do the above change

sorry, can I get further explanation on this?

nunojsa · 2026-09-25T12:49:04Z

+      the hardware connection (can be wired to SENSE1, SENSE2, or an average via
+      resistive divider).
+
+  adi,rsense1-nano-ohms:


maybe we can reuse adi,rsense-nano-ohms to be the first one in here?!

yes, possible

nunojsa · 2026-09-25T12:49:42Z

+curr3_min_alarm		Undercurrent alarm
+curr3_max_alarm		Overcurrent alarm
+curr3_crit_alarm        Critical Overcurrent alarm
+curr3_label		Channel label (ISENSE2)


Why 2 and 3? I would say we just have 1 and 2 no?

Was running on the assumption SENSE, SENSE1, and SENSE2 are all different, as commented below I shall double check this

nunojsa · 2026-09-25T12:49:57Z


 config SENSORS_LTC4283
-	tristate "Analog Devices LTC4283"
+	tristate "Analog Devices LTC4283 and compatibles"


Maybe "Similar devices"?

will change

nunojsa · 2026-09-25T12:51:03Z

 #define LTC4283_POWER_MIN		0x48
 #define LTC4283_POWER_MAX		0x49
-#define LTC4283_RESERVED_68		0x68
-#define LTC4283_RESERVED_6D		0x6D


Why have you removed the above? Aren't they still reserved for ltc4283?

I think I was thinking this was a placeholder especially since ltc4284 uses these, but well, yeah... it doesn't make sense on the view of ltc4283, will revert.

nunojsa · 2026-09-25T12:52:15Z

+	/* in tenths of microohm - For LTC4284 channel 1 (SENSE1) */
+	u32 rsense1;
+	/* in tenths of microohm - For LTC4284 channel 2 (SENSE2) */
+	u32 rsense2;


Again! For the registers defintions, sure they are different. But in here I don't see why we need two new rsense1 and 2. rsense can be repurposed to be the "rsense1" no?

will try to repurpose 'rsense', but yes it should be possible

nunojsa · 2026-09-25T12:54:39Z

+};
+
+struct ltc4283_chip_info {
+	enum ltc4283_chip_id id;


I think that instead of an id, what we might need is has_sense2 boolean?!

will check out this out!

actorreno · 2026-09-28T06:32:48Z

I think the major thing to rethink is sense1 and sense2. I think it should just be a new sense2 variable! The docs are actually wrong IMO as it looks like we can ever have curr1, curr2 and curr3 which is not the case!

yes, I may need to clarify this with apps, double check sense, sense1, sense2. would answer a lot of review points above.

actorreno · 2026-09-30T00:19:57Z

apps replied, SENSE is just the average of SENSE1 and SENSE2 for ltc4284.
my edits would make SENSE from ltc4283 as the ltc4284 SENSE1, then I'll add SENSE2.

So ltc4283 = SENSE.
ltc4284 = SENSE and SENSE2.
similar treatment to the resistors

Will not support the average as it's not an actual real channel

gastmaier force-pushed the mirror_ci/groeck/linux-staging/hwmon-next branch from 8411d1b to a17a53a Compare July 22, 2026 00:00

actorreno force-pushed the staging/ltc4284 branch from 01e1cf6 to c482653 Compare July 22, 2026 02:35

actorreno marked this pull request as ready for review July 22, 2026 03:41

actorreno requested a review from nunojsa July 22, 2026 03:42

actorreno force-pushed the staging/ltc4284 branch 9 times, most recently from b275459 to 5e45c36 Compare July 29, 2026 22:49

actorreno force-pushed the staging/ltc4284 branch from 5e45c36 to dc0ff7e Compare August 6, 2026 01:11

gastmaier force-pushed the mirror_ci/groeck/linux-staging/hwmon-next branch 8 times, most recently from ec684c3 to 16e3a40 Compare August 29, 2026 00:00

gastmaier force-pushed the mirror_ci/groeck/linux-staging/hwmon-next branch 5 times, most recently from c70491f to 666afcf Compare September 3, 2026 00:00

actorreno requested review from CiprianRegus, amiclaus, buha, chegbeli, danmois, machschmitt, rbolboac and stefpopa as code owners September 18, 2026 00:51

actorreno changed the base branch from main to mirror_ci/groeck/linux-staging/hwmon-next September 18, 2026 00:51

actorreno force-pushed the staging/ltc4284 branch from 8e6ac06 to f44a9bf Compare September 18, 2026 01:05

actorreno added 3 commits September 18, 2026 09:05

gpio: gpio-ltc4283: Add support for the LTC4284 Swap Controller

6deff73

Similar to LTC4283, LTC4284 has up to 8 pins that can be configured as GPIOs and behaves the same. Signed-off-by: Alexis Czezar Torreno <alexisczezar.torreno@analog.com>

actorreno force-pushed the staging/ltc4284 branch from f44a9bf to 6deff73 Compare September 18, 2026 01:06

gastmaier force-pushed the mirror_ci/groeck/linux-staging/hwmon-next branch 7 times, most recently from a7c82e2 to a249931 Compare September 25, 2026 09:33

nunojsa mentioned this pull request Sep 25, 2026

hwmon: add driver for ltc4284 #3351

Closed

4 of 8 tasks

nunojsa reviewed Sep 25, 2026

View reviewed changes

gastmaier force-pushed the mirror_ci/groeck/linux-staging/hwmon-next branch 3 times, most recently from b328ade to e943ef6 Compare September 28, 2026 00:01

gastmaier force-pushed the mirror_ci/groeck/linux-staging/hwmon-next branch from e943ef6 to 935f1d8 Compare October 1, 2026 00:02

Conversation

actorreno commented Jul 21, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR Description

PR Type

PR Checklist

Uh oh!

nunojsa commented Jul 28, 2026

Uh oh!

actorreno commented Jul 29, 2026

Uh oh!

actorreno commented Jul 29, 2026

Uh oh!

actorreno commented Sep 18, 2026

Uh oh!

nunojsa left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

actorreno commented Sep 28, 2026

Uh oh!

actorreno commented Sep 30, 2026

Uh oh!

This branch has not been deployed

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

20 participants

actorreno commented Jul 21, 2026 •

edited

Loading