Added basic implementation for precharge. One function for initializi… #34

DimitryP6 · 2025-11-14T00:13:28Z

…ng a given struct to given values, another one for running in an infinite loop.

Changes

Function that takes an empty struct and initializes it with given values + function that runs an infinite loop for voltages.

To Do

Any remaining things that need to get done

delays during infinite loop?
is the check for dipping below a lower bound ratio necessary for capacitor?
not sure about arrangement of function parameters.

Checklist

It can be helpful to check the Checks and Files changed tabs.
Please reach out to your Project Lead if anything is unclear.
Please request reviewers and ping on slack only after you've gone through this whole checklist.

[ *] No merge conflicts
[ *] All checks passing
[ *] Remove any non-applicable sections of this template
[ *] Assign the PR to yourself
[ *] Request reviewers & ping on Slack
[* ] PR is linked to the ticket (fill in the closes line below)

Closes #5

…ng a given struct to given values, another one for running in an infinite loop.

caiodasilva2005

Awesome work!

caiodasilva2005 · 2025-11-15T20:30:19Z

Core/Inc/precharge_routine.h

+	float lower_ratio;
+	nertimer_t debounce_timer;
+	uint32_t debounce_time;
+} prechargeconfig_t;


Love the struct for this, could you maybe add some docs for what each field is for?

caiodasilva2005 · 2025-11-15T20:30:54Z

Core/Inc/precharge_routine.h

+ * @param precharge_config is the empty configuration to configure.
+ * @return a precharge configuration for running the precharge thread.
+ */
+prechargeconfig_t *precharge_init(cell_asic_2950 ic, SPI_HandleTypeDef *hspi,


Can we add an assert that the transition ration is <= 1 and >= 0 and that hspi is not null

caiodasilva2005 · 2025-11-15T20:32:46Z

Core/Inc/precharge_routine.h

+ * @brief Runs the pre-charge until the threshold-ratio is reached, after which the AIR switch is set to open.
+ * @param precharge_config is the configuration to run.
+ */
+void precharge_run(prechargeconfig_t *precharge_config);


nit: rename to vPrecharge like the tasks defined in shep_tasks.c and define it there. You can also set the thread parameters like delays (let's probably do about 500 ms to start) and other settings there. Use other tasks as a reference

caiodasilva2005 · 2025-11-15T20:33:05Z

Core/Inc/precharge_routine.h

+/**
+ * @brief Given the cell, SPI, AIR switch, initializes the pre-charge structure for the given BATT/TS voltage capacitor threshold ratio, 
+ * and with the given debounce time to wait when the AIR opens before closing again.ADI1_delay_ms
+ * @param ic idk what this is T-T.


this is the struct for the 2950 chipon hv plate for docs

caiodasilva2005 · 2025-11-15T20:33:31Z

Core/Inc/precharge_routine.h

+ * @brief Given the cell, SPI, AIR switch, initializes the pre-charge structure for the given BATT/TS voltage capacitor threshold ratio, 
+ * and with the given debounce time to wait when the AIR opens before closing again.ADI1_delay_ms
+ * @param ic idk what this is T-T.
+ * @param hspi idk what this is either T-T.


this is the struct for the spi handler that goes into the isospi channel wae are using to comm to the chip for docs

caiodasilva2005 · 2025-11-15T20:34:10Z

Core/Src/precharge_routine.c

+
+void precharge_run(prechargeconfig_t *precharge_config)
+{
+	if (precharge_config == NULL) {


Make this an assert here, the code should crash if something critical like this is missing

caiodasilva2005 · 2025-11-15T20:34:39Z

Core/Src/precharge_routine.c

+				    &precharge_config->debounce_timer)) {
+				cancel_timer(&precharge_config->debounce_timer);
+			} else {
+				// SOME SORT OF DELAY?


See above comment for delay timings, but yes a delay here is necessary

caiodasilva2005 · 2025-11-15T20:34:49Z

Core/Src/precharge_routine.c

+					    precharge_config->debounce_time);
+			}
+		}
+		// DELAY?


See above comment about delay time.

caiodasilva2005 · 2025-11-15T20:35:03Z

Core/Src/precharge_routine.c

+					  precharge_config->hspi,
+					  precharge_config->gpo);
+				air_switch_closed = false;
+				// I would assume there could be a voltage spike here as well.


this is good logic, comment can be removed

caiodasilva2005 · 2025-11-15T20:36:04Z

Core/Src/precharge_routine.c

+			}
+		} else if (!air_switch_closed) {
+			// Check if charged up until threshold.
+			// Should there be another check for depleted battery voltage?


ah do you mean like welded AIR check? Let's not have that here now (not really in scope). This comment can be removed

caiodasilva2005

Funny thing @jr1221 I am almost sure that our debouncing control code in Cerberus 1.0 did not work sometimes because we were using the same timer for the device on debounce and device off debounce lol

jr1221

nice stuff. idk what the chipdata diff means so didnt review that.

jr1221 · 2026-01-09T04:38:01Z

Core/Src/precharge_routine.c

+	precharge_config->close_debounce_timer =
+		(nertimer_t){ 0, 0, false, false };
+	precharge_config->debounce_time = debounce_time;
+	precharge_config->air_switch_closed = false;


where is the return here?

jr1221 · 2026-01-09T04:39:21Z

Core/Src/precharge_routine.c

+	hv_plate_t *hv_plate = precharge_config->hv_plate;
+	bool should_precharge = // TODO: mutex hv plate data
+		hv_plate->ts_volts * precharge_config->transition_ratio >=
+		hv_plate->batt_volts;


shouldnt it be batt_volts * transition ratio? Because its when ts gets above 90% of battery?

Ah whoops yeah thanks for catching that

jr1221 · 2026-01-09T04:40:18Z

Core/Src/shep_tasks.c

+
+	prechargeconfig_t precharge_config;
+	precharge_init(&precharge_config, hv_plate, 0.9f,
+		       50 /* ms debounce time */);


lets start more like with 200ms. time is not of the essence but safety sure is imo.

jr1221 · 2026-01-09T04:42:59Z

Core/Src/precharge_routine.c

+		 precharge_config->debounce_time, close_relay,
+		 precharge_config);
+
+	debounce(!should_precharge, &precharge_config->close_debounce_timer,


im going to do some review and start a thread with markow and pablo tmrw night, however I think we need to de-assert the GPIO whenever shutdown is open, because when shutdown closes we dont want to have our GPIO already high in effect only precharging the first time.

Added basic implementation for precharge. One function for initializi…

911b794

…ng a given struct to given values, another one for running in an infinite loop.

DimitryP6 requested a review from caiodasilva2005 November 14, 2025 00:13

DimitryP6 self-assigned this Nov 14, 2025

caiodasilva2005 requested changes Nov 15, 2025

View reviewed changes

caiodasilva2005 and others added 5 commits December 23, 2025 15:22

Merge branch 'develop' into feature/pre-charge

a2c1b82

5 - make new task for precharge

fb45ec4

use hv plate calculated data

2be28cd

5 - fix init

f01d316

5 - fmt

29126e5

caiodasilva2005 marked this pull request as ready for review January 9, 2026 03:30

caiodasilva2005 added 3 commits January 8, 2026 22:35

up to date

7f93e45

fix getting chip data

f6fae01

fmt

02facbd

caiodasilva2005 requested review from KSMehta11 and jr1221 January 9, 2026 03:47

caiodasilva2005 reviewed Jan 9, 2026

View reviewed changes

jr1221 requested changes Jan 9, 2026

View reviewed changes

caiodasilva2005 added 2 commits January 9, 2026 15:40

fixed precharge eq

633e0d8

fixups

db42d40

Added basic implementation for precharge. One function for initializi… #34

Are you sure you want to change the base?

Added basic implementation for precharge. One function for initializi… #34

Uh oh!

Conversation

DimitryP6 commented Nov 14, 2025

Changes

To Do

Checklist

Uh oh!

caiodasilva2005 left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

caiodasilva2005 Nov 15, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

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!

caiodasilva2005 left a comment

Choose a reason for hiding this comment

Uh oh!

jr1221 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!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

caiodasilva2005 Nov 15, 2025 •

edited

Loading