Esphome component related updates by p1ngb4ck · Pull Request #274 · kanflo/opendps

p1ngb4ck · 2026-02-03T21:04:56Z

As the title says - my esphome component allows for ethernet connectivity as well, therefore I propose the following rework.
Mostly just renaming from wifi to network and add ethernet logic & symbol.
Also, display backlight value was added to response-content and I changed default baud to 19200 after intensive changes (please just deny that change of course if you dont agree)
Removed the status code change and rather changed my esphome component to be fully backwards compatible, no need to reflash if someone doesnt need display backlight in query response etc.

Esphome component finished now as well, last remaining point is to create a nice HA custom card
ESPHome component features:

handle sensors for VIN/VOUT/CUR_OUT
handle inputs for voltage, current, output switch (enable/disable), lock, display
provides functionalits for data logging to sd
provides calibration value request & save to sd storage + auto-restore after flashing fw
fw upgrade function capable of handling different bootloader speeds (if bootloaders uart baud rate setting is different from "main" opendps one) and capable of using esphome-devices "local" storage (sd/usb/internal flash data/extra spi flash attached)
to read fw bin file from
provides tcp bridge on configurable port, auto-pausing and resuming when/after external tcp client connects - use in dpsctly.py with param "-d tcp:"

As esphome can deal with ethernet/wifi + mqtt + api connections (ha & custom) -> these features are automatically available as well when replacing the esp8266 wifi-serial- proxy with an esphome based device using my opendps component

Hope this helps =)

kanflo · 2026-02-14T15:22:44Z

Thanks! I like this PR and only have a few minor comments.

p1ngb4ck · 2026-02-14T21:23:38Z

Should be all done now, thanks so much again for this awesome project and still maintaining it !!!!!!!!
jftr - for readers that may see this:
After calibration with a cheap 8ohms 200W load using dpsctl.py, my OpenDPS (DPS5005 model) is extremely exact again (0-3mV/mA both compared to expensive Fluke multimeter measurements).
Together with the high-speed network update rate and integration into HA (automations etc) this is actually the lab psu I always wanted to have !

Both dpsboot and opendps now always start at 9600. Main firmware reads past_uart_baud on boot and switches to the saved operational baud. - protocol: add cmd_set_baud (22) for runtime baud negotiation - pastunits: add past_uart_baud for persisting operational baud - opendps: add hw_set_baudrate(), opendps_set_uart_baud(), opendps_is_valid_baud() - protocol_handler: add handle_set_baud() - ACKs at old baud, then switches - dpsboot: add hw_set_baudrate_boot(), handle cmd_set_baud session-only - Rename cmd_success_with_response (was cmd_success_but_i_actually_sent_my_own_status_thank_you_very_much) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

p1ngb4ck · 2026-04-18T16:20:35Z

Hey - would you mind having a look at this again?
Changed baudrate stuff as desired :

Bootloader & Firmware always start off at 9600
Bootloader can be set to different baudrate via CMD before upgrade if desired
Baud for main fw can be "set" and will be saved and re-applied. If master reset is done, it will revert back to 9600.

This way no baud config during compile is required, all bauds can be changed, it remains "simple" and is very safe as well, no bricking possible - not even "hard" to reach bl should be possible this way

- protocol.py: add CMD_SET_BAUD (23), VALID_BAUD_RATES, create_set_baud() - dpsctl.py: add --set-baud to set/save device UART baud at runtime - dpsctl.py: add --upgrade-baud to switch bootloader to faster rate after upgrade_start ACK for faster data transfer - tty_interface: add set_baudrate() to switch serial port without reconnecting - handle_response: handle CMD_SET_BAUD response Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Eliminates per-sample ADC noise from display and query values. OCP/OVP still operate on raw per-sample values for fast response. i_out averaging uses offset-corrected value matching OCP logic. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

…display - hw.c: i_corrected was applying adc_i_offset twice when pwrctl_i_limit_raw was set (i already had offset added for OCP check). Fix: only apply offset in averaging path when the OCP block was skipped (limit not yet loaded). - opendps.c: increase Vin display from 1 to 2 decimal places. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

ADC offset calibration leaves a ~8mA residual baseline on this unit due to ADC_CHA_IOUT_GOLDEN_VALUE being tuned for the developer's unit. Clamp averaged i_out to 0 when <= 10 counts; 11mA+ is unaffected. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

…urrent Derived from calibration data: A_ADC_K=1.64672816, A_ADC_C=-106.07878113 Zero count = 106.08/1.647 = 64.41 → 64 (was 69/0x45, causing +8mA baseline offset) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Set correct value (64) directly in dps-model.h for DPS5005 instead of overriding via Makefile -D flag, which caused redefinition warnings (dps-model.h defines unconditionally, hw.c has #ifndef guard, Makefile -D was a third conflicting definition). Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

kanflo · 2026-04-19T07:45:14Z

+                }
+                uint8_t valid = 0;
+                switch (baud) {
+                    case 9600: case 19200: case 38400: case 57600: case 115200:


Or use opendps_is_valid_baud(...) ;)

kanflo · 2026-04-19T07:45:25Z

+void hw_set_baudrate(uint32_t baud)
+{
+    switch (baud) {
+        case 9600: case 19200: case 38400: case 57600: case 115200:


Or use opendps_is_valid_baud(...) ;)

kanflo · 2026-04-19T07:47:19Z

      * UART meaning this message got lost. The ESP will however send the
-      * wifi_connected status when it connects but if that does not happen, the
-      * ui module will turn off the wifi icon after 10s to save the user's
+      * twork_connected status when it connects but if that does not happen, the


kanflo · 2026-04-19T07:48:43Z

Just a few minor wrinkles and then we should be able to merge this before the weekend is over. Thanks for your contribution!

p1ngb4ck · 2026-04-19T12:41:19Z

<3 thanks so much fort reviewing so quickly - a little ashamed of these "stupid" mistakes I left in there rn - gonna fix them quickly now

Edit says - should hopefully be fine now. Also added the function to dpsboot bootloader, just to make it all more consistent across project - and doesnt require "more" space for bootloader .. so I hope that`s fine
Edit 2 - compile error due to missing include fixed

gengge2010 · 2026-04-19T12:50:43Z

退订 At 2026-04-19 15:45:39, "Johan Kanflo" ***@***.***> wrote: @kanflo commented on this pull request. In dpsboot/dpsboot.c:

@@ -261,6 +261,33 @@ static void handle_frame(uint8_t *payload, uint32_t length)

} } break; + case cmd_set_baud: + { + uint32_t baud = 0; + if (payload_len >= 5) { + baud = (uint32_t)payload[1] << 24 | (uint32_t)payload[2] << 16 | + (uint32_t)payload[3] << 8 | (uint32_t)payload[4]; + } + uint8_t valid = 0; + switch (baud) { + case 9600: case 19200: case 38400: case 57600: case 115200: Or use opendps_is_valid_baud(...) ;) — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

…_baud to dpsboot, use it in both firmwares, fix spelling Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

…or ui_parameter_t Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

p1ngb4ck · 2026-04-19T17:52:33Z

@kanflo I think u will also be proud as soon as u see what is possible with your opendps and my esphome component.
LVGL 9.5 display is almost fully complete, doing the last tiny additions/fixes at the moment.
Functions (in addition to obvious "default" functionality of opendps):

datalogger to sd/usb/nfs
fw upgrade via sd/usb/nfs-share/http-url
calibration via display menu & assistant
custom theme support (pre-defined themes + custom colors)
main/graphs/settings-pages
ethernet/wifi support in addition to "on screen controls), including ha-api & mqtt & voice-assistant support
very high modability - due to many unused gpios, spi/i2c available etc etc

kleineckeoliver-boop added 4 commits January 26, 2026 21:16

add latest esphome related changes

67c246b

latest changes & fixes

d919c2f

latest changes to fw

c17e2ff

fix status response code

e53fd42

kanflo reviewed Feb 14, 2026

View reviewed changes

Comment thread dpsboot/Makefile

kanflo reviewed Feb 14, 2026

View reviewed changes

Comment thread opendps/Makefile

kanflo reviewed Feb 14, 2026

View reviewed changes

Comment thread opendps/Makefile

kanflo reviewed Feb 14, 2026

View reviewed changes

Comment thread opendps/Makefile

Change default baudrate and disable splash screen

b92f6de

kleineckeoliver-boop and others added 2 commits April 18, 2026 16:12

Merge branch 'kanflo:master' into esphome_related_updates

a611686

kleineckeoliver-boop and others added 7 commits April 18, 2026 16:21

[dpsboot] Enable LTO and trim hw_set_baudrate_boot to fit in 5KB flash

4a9a172

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

kanflo reviewed Apr 19, 2026

View reviewed changes

kleineckeoliver-boop and others added 2 commits April 19, 2026 12:54

[dpsboot/opendps] Address PR kanflo#274 review: port opendps_is_valid…

d32fc82

…_baud to dpsboot, use it in both firmwares, fix spelling Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

[opendps] Fix compile error: include uui.h before opendps.h in hw.c f…

70ee93d

…or ui_parameter_t Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Esphome component related updates#274

Esphome component related updates#274
p1ngb4ck wants to merge 16 commits intokanflo:masterfrom
p1ngb4ck:esphome_related_updates

p1ngb4ck commented Feb 3, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

kanflo commented Feb 14, 2026

Uh oh!

p1ngb4ck commented Feb 14, 2026

Uh oh!

p1ngb4ck commented Apr 18, 2026

Uh oh!

kanflo Apr 19, 2026

Uh oh!

kanflo Apr 19, 2026

Uh oh!

kanflo Apr 19, 2026

Uh oh!

kanflo commented Apr 19, 2026

Uh oh!

p1ngb4ck commented Apr 19, 2026 •

edited

Loading

Uh oh!

gengge2010 commented Apr 19, 2026 via email

Uh oh!

p1ngb4ck commented Apr 19, 2026 •

edited

Loading

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Conversation

p1ngb4ck commented Feb 3, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

kanflo commented Feb 14, 2026

Uh oh!

p1ngb4ck commented Feb 14, 2026

Uh oh!

p1ngb4ck commented Apr 18, 2026

Uh oh!

kanflo Apr 19, 2026

Choose a reason for hiding this comment

Uh oh!

kanflo Apr 19, 2026

Choose a reason for hiding this comment

Uh oh!

kanflo Apr 19, 2026

Choose a reason for hiding this comment

Uh oh!

kanflo commented Apr 19, 2026

Uh oh!

p1ngb4ck commented Apr 19, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

gengge2010 commented Apr 19, 2026 via email

Uh oh!

p1ngb4ck commented Apr 19, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

p1ngb4ck commented Apr 19, 2026 •

edited

Loading

p1ngb4ck commented Apr 19, 2026 •

edited

Loading