Respect pn532 i2c timing #367
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The pn532 documentation differs slightly from the included ascii art documentation on how a packet looks like. The preamble was omitted however the postamble is mentioned. This patch adds the Preamble to the ascii frame documentation. The code later refers incorrectly to the start byte as the preamble. This variable was renamed to more descriptively state that it is the preambe and the start bytes. Signed-off-by: Olliver Schinagl <[email protected]>
The pn532 user manual states that after a i2c stop condition and before a i2c start condition there should be a delay of minimally 1.3 milliseconds. This is probably a limitation of the i2c peripheral or the firmware. In any case, each i2c_read and i2c_write creates the packets which are complemented with start/stop markers. It is thus required to take care of timing in these two functions. We solve this by wrapping the lower i2c_read and i2c_write functions for the pn532, as this requirement is not for all chips. Currently, we keep time using local variable, and thus the code is not thread-safe. With libnfc being single threaded and only one instances of libnfc can open a bus anyway, this is not yet a problem. Signed-off-by: Olliver Schinagl <[email protected]>
Currently, we very occasionally can EXNIO errors from pn532_i2c_write() -> i2c_write() -> write(). This may happen about once every 30 seconds. Based from the kernel sources, EXNIO happens if the chip no longer responds to its own address. To make sure we do not loose any sent packets, we retry to send PN532_SEND_RETRIES number of times. Since we miss 1 every 30 or so seconds, doing 1 retry should be fine. This might be considered a hack as the failure may be some other timing related issue. Signed-off-by: Olliver Schinagl <[email protected]>
|
Nice work ! |
|
Thanks, and I forgot this thread, but the timing is wrong. The datasheet says 1.3 ms and even though due to delays we are at 1.4-ish, this is still to fast for the I2C block inside the chip. We are now trying to find the actual timings, but NXP said they are not sure what is going on here. We are now at 4ms and my guess is we will find the sweet spot at 2.5ms. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The pn532 states in its datasheet that the i2c stop and start conditions require atleast 1.3 ms between them. Currently, there is a nanosleep of 90 ms which seems arbitrarily chosen (nothing in the comments/git log from it). A delay however is always needed, as otherwise we write to fast for sure.
Additionally, this delay is not always added everywhere, so we still sometimes send data way to fast for the pn532. This patch creates wrappers around i2c_read and i2c_write to ensure proper timing of read and write commands.
Additionally we clarify the preamble/start condition a bit, as there was a bug in the text. No code changes where done in that regard.
Note, I have tested it on one board on one machine and not yet endlessly. So I'd like to first properly review this first before merging :)
edit: I cleaned some more on the preamble bit, and added a retry mechamism to the pn532_send() routine.