feat: DPS310 driver #463
feat: DPS310 driver #463
Conversation
Requested at UncleRus#456 Sponserd by beriberikix Implemented I2C digital interface.
|
Thank you Sakurai-san. Very nice and neat code, as always with you. |
while here, fix a bug in the example by initializing the device descriptor with zeros.
* configure the device, and warn when something looks incorrect * fix bugs in two_complement_of(). * rewrite the example for more realistic use cases` while here, replace all the `DSP` and `dsp`.
pointed out at: UncleRus#463 (review)
components/dps310/src/dps310.c
Outdated
| esp_err_t dps310_is_ready_for(dps310_t *dev, uint8_t reg, uint8_t mask, bool *ready) | ||
| { | ||
| esp_err_t err = ESP_FAIL; | ||
| uint8_t reg_value; |
There was a problem hiding this comment.
initialize the variable (and do audit in other places).
components/dps310/src/helper_i2c.c
Outdated
| static const char *TAG = "dps310/helper_i2c"; | ||
|
|
||
| /* count the number of trailing zero in a value, usually bitmasks. */ | ||
| static uint8_t count_trailing_zero_bits(uint8_t v) |
There was a problem hiding this comment.
prefix it with _ (and document what it means).
|
I was able to test the sample on a ESP32-S3 & Adafruit sensor board and had no issues! |
also add more comments
compensation can be skipped for performance. i believe the use case is not common. no plan to implement faster-but-less-accurate versions. there are bits and pieces in the public header for that.
and document the function
trombik
force-pushed
the
feat-driver-dps310
branch
from
ae852c4 to
6ff7b98
Compare
run measurements in the loop, which is what most users expect from an example.
with `_oversampling_[tp]` and `_rate_[tp]`. and fix a cirtical bug, in which sampling rate was incorrectly used (instead of oversampling).
trombik
force-pushed
the
feat-driver-dps310
branch
from
30e3d33 to
d7ad7d9
Compare
|
@beriberikix the public APIs are considered as stable now. please have a look at the public header, and the example. I would like to request for your comment, specifically; do the naming conventions in functions, enum, and typedef look clear? I kept original words where appropriate, but modified some (such as I will continue to work on background mode and interrupt, but the driver is now usable for simple use cases. SPI mode is a bit complicated because incompatible SPI APIs in ESP8266 RTOS SDK and esp-idf. the SPI interface is probably designed for use cases in which performance is critical (such as drones). but command mode is enough for your use case, i guess (and if you really need it, it is possible to implement your own by using public APIs). |
|
@UncleRus I believe that everything you mentioned is fixed. if you are okay with the PR. please merge. I have already noticed that some functions do not have necessary |
|
Thank you! |
Requested at #456
Sponserd by beriberikix.
Implemented I2C digital interface.