uSD card frame Slave Select SS

An advanced Arduino-compatible board with ATmega1284P and ATmega32u2 MCUs for experienced users. More RAM than a Mega, but in a convenient Uno-style form factor. [Product page]
Post Reply
Joined:Tue Nov 05, 2013 1:24 am
uSD card frame Slave Select SS

Post by ozra » Tue Nov 05, 2013 1:37 am

I was a backer for the Pozible page for 2 Goldilocks boards.

Just wondering if I ask the question below here or do I approach Freetronics now?

Great boards thank you. Looking forward to playing with them during my annual leave.

I noticed an Errata was published regarding the SD card wiring.

The uSD card frame Slave Select SS, is not connected to PB0 (Digital 14) as per design, but rather to the normal Arduino Digital 4 (PD4).This impacts the use of 16bit PWM (found on PD4 and PD5) when the SD card is in use. This also impacts the use of multiple SD cards, as the Goldilocks SD card SS line is now shared with the standard Arduino layout.

Do I now send my unused and unopened boards back to get the correct boards as per the original design?

As you can imagine, impacting a pwm pin is quite a severe drawback compared to an ordinary I/O pin. For my intended project I need BOTH the SD card and all PWM functionality.

Thank You

Joined:Fri Jul 19, 2013 6:17 am

Re: uSD card frame Slave Select SS

Post by feilipu » Tue Nov 05, 2013 10:44 am

Hi Ozra,

I understand your concern about the uSD SS pin not being as designed. It clearly wasn't the plan. The impact is quite high for PWM applications, which is why I have now called this issue out clearly in the User Manual.

The Pozible project is closed now, as all the boards have been sent out. For better or worse, I'm not a manufacturer but rather an individual acting with the best of intentions. The offer to be a supporter through crowd funding always comes with some risk.

If you buy a Goldilocks from Freetronics, as you can do now, it is subject to their return policy which you could take up with the team directly.

For your interest, anyone could have spotted the problem in my design document, publicised in Pozible updates, and called it out. It is the light blue trace running horizontally across the top of the 1284p Crystal intersecting with PD4, in the May 3rd Design rules check. I know I didn't spot it until someone else on this forum noted that the uSD card seemed to be working when it shouldn't have been. I've been kicking myself since then. ... Gtqfw/edit#

How to fix what you need? The only solution that I've thought of is to remove the SMD resistor linking the uSD SS trace to PD4, and then use a jumper wire to PB0. It doesn't seem easy to cut the light blue trace without damaging the board, as it is an internal trace.

Other than that, I'm working on a new Goldilocks design with some additional features, and fixes for the USART sync issue, and uSD card issue. I'm getting some samples soon, but probably nothing will be ready for public consumption until next year.

Sorry I can't be of direct immediate help.


Post Reply