Error in Teensy 4.0 circuit diagram (?)

uTasker

Well-known member
Hi All

Would it be correct to say that the labels on these two pins (Teensy 4.0 circuit diagram) are incorrect and need to be swapped?

teensy_4_SD.png

Regards

Mark
 
Hmm. If I'm reading the manuals correctly, SD_B0_05 (J2) is supposed to be DAT3, not DAT2 and SD_B0_04 (H2) is supposed to be DAT2, not DAT3. So something does appear to be swapped. If the labelling on the chip side of things is correct then yes, the DAT2 and DAT3 lines at the SDIO pads would appear to be swapped. But how on earth would the a uSD card hooked up to the SDIO lines even work then?
 
Hmm. If I'm reading the manuals correctly, SD_B0_05 (J2) is supposed to be DAT3, not DAT2 and SD_B0_04 (H2) is supposed to be DAT2, not DAT3. So something does appear to be swapped. If the labelling on the chip side of things is correct then yes, the DAT2 and DAT3 lines at the SDIO pads would appear to be swapped. But how on earth would the a uSD card hooked up to the SDIO lines even work then?

Maybe if data written if swapped then read if swapped is the same. But commands? Curious...
 
As comparison, this is the diagram from an NXP i.MX RT 1062 board:

teensy_4_SD_2.jpg

which is different from the Teensy circuit diagram and matches the chips' documentation.

I have worked with the same SD card connecte to a Teensy 4.0 and to this NXP board and both worked/behaved identically.

Therefore i expect that the HW is correct but doesn't match with the circuit diagram.

I'll do a quick test by programming one of the two suspect line as GPIO output with toggle to see whether it comes out on the pad as expected or not; I'll report back in a few minutes.

Regards

Mark
 
Maybe if data written if swapped then read if swapped is the same. But commands? Curious...

Commands were what I was thinking of too; swapped data lines you likely wouldn't notice until you took the uSD card out and plugged it into something else. But commands shouldn't work at all if they make use of all 4 lines and the 2 lines were swapped. Someone would have noticed months ago. I wonder if it's just the labels on the IC side of things that are swapped on those 2 lines?
 
When I configure the i.MX RT 1062 pin SD_B0_05 as CLKO2 and connect it to an internal clock I can measure this signal on the SD card's DC/DAT3 line.

I believe that this proves that the HW is correct but the circuit diagram doesn't correspond to the HW.

The PCB pads and labels are correct but the pins that they are connected to are swapped:
- SD_B0_04 (H2) is connected to Pad 39 / DAT2
- SD_B0_05 (J2) is connected to Pad 38 / DAT3

I have modified my reference and put it below.

Regards

Mark


Teensy4_0_1.png
 
Thank you for checking that out. I'd hoped that the labelling on the IC side was what would turn out to be wrong.
 
Commands were what I was thinking of too; swapped data lines you likely wouldn't notice until you took the uSD card out and plugged it into something else. But commands shouldn't work at all if they make use of all 4 lines and the 2 lines were swapped. Someone would have noticed months ago. I wonder if it's just the labels on the IC side of things that are swapped on those 2 lines?

I personally think it is. But if expansion and breakout boards for the T4 are designed based on that schematic without the designer checking the validity of the pinouts, there could be problems with the boards in regards to the SDIO port. There is one such breakout board that I have that is problematic with the SDIO port. Paul's T4 BETA breakout board works without any problems using uSDFS, SDfat and SD which again leads me to believe it's just a schematic error:)

EDIT: This is nothing more than observations. Not proven yet.
 
Last edited:
Back
Top