5
Redefine SPI signal names
- Open
- Subscribe
|
J |
Jason |
I suggest that the Logic 2 SPI analyzer use SDO/SDI for data line names instead of MISO/MOSI. For the rationale and more information see this page:
www.oshwa.org/a-resolution-to-redefine-spi-signal-names/
Activity Newest / Oldest
Jonathan
Would support adding a "naming convention" selector to the config menu since the device might be using either of 3 conventions (MISO/MOSI, SDO/SDI, CIPO/COPI)
Joey
The terms "master" and "slave" have been controversial for years, and, in my opinion, can be easily replaced, in the sense that they are labels and I have no emotional attachment to them. However, I agree with opponents of changing, in that choosing any other name should be a careful process. The replacements should be unambiguous, just as MISO and MOSI are.
By the way, there are more than the mentioned variants. For example, the article mentions NXP using "DIN/DOUT" but not that NXP also uses "SIN/SOUT" ('S' standing for "Serial") on other recent microcontrollers. I imagine there are more exceptions. Fun fact: NXP allows SIN/SOUT pins to be swapped. A useful feature when the hardware-engineer connects SDI to SDI and SDO to SDO. MISO and MOSI were better in that regard. :^)
Using any of the single-device naming conventions, in my opinion, makes it ambiguous as to which line is being looked at from the capture point-of-view. I would propose using controller/peripheral (CIPO, COPI) or primary/secondary (PISO, POSI), in order to keep the naming unambiguous. Otherwise you would make the distinction by writing lengthy qualifiers, such as "Controller SDO, Peripheral SDI", but that would waste screen space and make quick distinctions harder.
Peter Jaquiery
I never thought of Master/Slave in MOSI in terms of human relationships or slavery until you posted this idea. Is it your intent to seed discontent and force slavery to the forefront of people's minds?
For many years I worked with RS-232 devices and there was always confusion about what Tx and Rx meant and which end was listening and which end talking. MOSI/MISO completely avoids that. SDI/SDO recreates that confusion.
Jason
Rather than debate it from a political perspective, I would just put the suggestion as follows:
- Many chip manufacturers* have replaced MISO/MOSI with SDO/SDI in their datasheets. The reason why is not important, and the directionality does not matter for a logic analyzer (since you can choose which signal you want to call "in" or "out").
- As a consequence, it would be great if the built in SPI analyzer followed what's now becoming the more common convention (or at least allow users to decide which they prefer)
*There is a list in the article I linked to, but actually even that is out of date, as both TI and Microchip use SDO/SDI now.
Peter Jaquiery
I think this renaming is unfortunate for the reason I gave in my first reply. MISO makes the data direction completely clear. SDO doesn't. With MISO the wire can have that label and be clear. With SDO there is no label that makes the direction of the signal on the wire clear without going back to something like the MISO nomenclature.
Politics and knee jerk reactions are driving bad decisions. This is one of them.