G400-S custom PCB project advice

NETMF & Gadgeteer
1

For those of you who have developed custom PCB projects using the G400-S, what pitfalls have you encountered that you wish someone would have told you about prior to layout?

For instance:

  1. “All needed is a 3.3V power source and some connections to take advantage of the G400’s long list of available features.” [Section 2, G400 manual]

Yet the Raptor schematic has power supplies for VCORE and VDDIO with non-3v3 voltage outputs. Is the manual incorrect?

  1. Were there any other signals required for basic functionality that were easily missed when reading the manual?

etc…

FYI: I am using Altium Designer.

2

This is a copy paste error from the D to the S variant. We will correct this asap.

3

Thanks for the reply on a Sunday!

May I assume the required voltages for VDDIOM and VDDCORE are 1.8V and 1V respectively? And what kind of current do they draw?

4

The FEZ Raptor uses the G400S, and the schematic is readily available in the Resources section. You can probably use that as a reference for your board design.
https://www.ghielectronics.com/catalog/product/499

5

The full datasheet for the CPU is here but I could not find any reference to the current drawn on each supply only a example of the total power when running at 400Mhz with all peripheral clocks active. From this I would assume the supplies are in the mA range. I’d do what iggmoe suggests and use the same IC’s or something with similar specs.

Reverse engineering the FB inputs of the voltage regulators gave me 1 volt for the VDDCORE and 1.8 for the VDDIOM supplies.

6

The Raptor schematic has been very helpful. It would still help to have more specific data on the G400-S itself, like a datasheet. Custom projects seldom require the level of robustness of a demo board like the Raptor. For instance, the switch mode power supplies, along with all the requisite inductors and capacitors, could easily be replaced with simpler LDO regulators for cost and space savings if one could ensure the current demand would be met. Tolerances and leakage currents would allow for better worst case analysis on the final product. Etc…

Edit:

Dave, I didn’t see your post until after I posted. I got the same values, but it’s always good to get verification! :slight_smile:

As to the Atmel reference, can I assume that pretty much all the interfaces are straight pad-to-micro with no additional circuitry changing things? It would be good to be able to use the Atmel datasheet.

I’m really excited to use this device! I do a good bit of embedded prototyping work on test fixtures and load simulators for much more involved ECUs, and I’m hoping the G400 can drastically improve design and software development time, while increasing capability.

Dave, I saw you were doing some projects like this. Were there any pitfalls you ran into that you could share?