Hi, I have been looking at the schematic for the Mega post on the main site. It has two regulators MC33xx. On the Mega board that I received and on the photos on the site, I can't seem to find these two regulators, only the NDT2955 and FDN340 or equiv. Have these been eliminated from the design? Is there a difference schematic? Or am I just blind?
I will be modifying the board to be 3.3v based for LCDs etc.
There are several differences between the board and the schematic (at least looking at my board, which I got from Sparkfun so I'm assuming it's relatively authentic).
The NDT2955 and FDN340 aren't regulators, they are transistors used in steering power to the +5V bus from either the external DC power input or the USB power source.
You should study the schematic carefully as there will be several implications for changing board power to 3.3V. For example, the USB will still supply 5V and if that finds its way to the FT232, its serial output pin will drive 5V into your 3.3V microcontroller. Not a good thing.
--
The Quick Shield: breakout all 28 pins to quick-connect terminals
Yes, I did see about the "power steering / pumping". Just making sure that I was not missing something on the regulators being not there.
The schematic has "doubled" components for a couple of things (regulator and mosfet) so as to create a PCB that can support several different packaging options. One of the regulators is in a DPAK and the other is in a SOT223 package. The PCB will support either one, but will only have one installed at a time.
The actual board should have one regulator and one MOSFET.
If the Arduino HW was designed by, say, NASA, this would be "clearly documented" somewhere within the thousand page hardware specification, and the Duemilanove would cost $13347... 
Yes, it would would cost $13,000 to put text near the regs saying "use res or Mosfets but not both". I assumed as much and have some regulators on the way.
Um, I hope not, since that would be wrong!
Here's a picture: The regulator occupies the bottom position, near the power plug, and the MOSFET (which is part of automatic power switching circuitry) is on top near the USB connector. In each position, there are two possible packages: a smaller package (highlighted) and a bigger package. You can only put on one at a time, but the regulator has to be in the regulator position and the mosfet has to be in the mosfet position.
If you have a Mega that has a mosfet in both positions, it is incorrectly assembled!! Name names and embarrass shady dealers (or help the world by calling attention to a broken run. Depending on whether you have a shady dealer, or there's a more pervasive problem.)
Ok I will study the schematic in more detail and this sounds reasonable (maybe), but then why does the one that I have have two mosfets AND the main picture on the web site has two, the schematic has two of each. I have not taken the time to really work out the schematics, will have to do so...
I will take some more time and study the schematic and see what they were really trying to go for......
In any case, I will be take BOTH out and putting in a 3.3v reg and that should do the trick. According to the schematic with I will have to make sure that the 5V from the USB is taken care of regulated to 3.3V
You can tell that the two devices are both MOSFETs from the website picture? Note that a mosfet in a sot223 package is going to look exactly like a regulator in an sot223 package, except for the lettering on the package...
The whole automatic power selection circuit is probably pretty questionable at 3.3V. (I don't like it even at 5V.) You're probably better off figuring out how to bypass it entirely...
Ok lets start from the beginning on this:
Power can come from the following sources.
The power jack. This is expected to be unregulated DC and then regulated to 5V.
The USB port. This is 5V and does work by itself with now power jackl power
External 5V or 3.3V. With nothing else plugged in I have tested the 3.3V, and it works. I have not plugged the USB in at the same time, from what I can tell it will supply V5 and I did not want any "back feed" current to the external 3.3V.
I see that by the trace level, there maybe should be one Mosfet and / or one regulator, but the names on the schematic are T1 and T2 and IC4 and IC5 and these are not label on the board, etc.
Yes, I know that this is all open source, and that is why I am asking the questions.
The board that I have seems to work to spec using the two Mosfet installation and I would guess that others are like this.
Will get back when I ask some others about the design, etc.
@ort11
What you write is not very consistant and seems a little bit dangerous. Please understand the complete power circuitry on your board which comprises a pair of comparators as well.
You might read through this thread to feel some of the issues that exist.
http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1284034373/12
When I now read FDN340 I get very very uneasy...
Ok, it's time to get a little bit picky here. There are things that need to be discussed.
What is the power section supposed to do?
What is it designed to do>
What it does given the components populated>
Goals:
It seems that the goals of the board are such that different power sources can be used to power the board.
There is also the question of priority if > 1 power source is used, especially if the USB source is used.
On the schematic there are the following power points
Now the question comes in for some of the chips that are being populated on the board. The one that I have has the following
Larger Location (Reg) 269-5G RU942 ON
This does seem to be a 5V regulator, previous searches for this part did bring up Mosfet which brought up some of the confusion.
Smaller Location (Mosfet) R94 2955E ON
This is a MOSFET.
So for sure, the larger location is for the regulator and the smaller one is for the Mosfet.
The schematic does show two options for either the regulator or the Mosfet, but you need one of each.
The whole reason for looking into this is to convert the board to a 3.3v power bus even when the USB connector is plugged in. Will probably have to route the 5V from the USB to the regulator section to do this. Should not be too bad, or if using an external barrel connector, simply disconnect the USB power all together.
Will be replacing the regulator with a 3.3V version.
Any thoughts on this?
Any thoughts on this?
Well there are several details you have to take into consideration. One would be the voltage support required by the FTDI chip:
From the FTDI 232R datasheet notes: "1. The minimum operating voltage VCC must be +4.0V (could use VBUS=+5V) when using the internal clock generator. Operation at +3.3V is possible using an external crystal oscillator."
I personally think the standard Arduino voltage switching function is over complicated and somewhat restricted in it's design. I kind of like how the current Seeeduino boards are designed from a power management perspective:
Manual switch is used to select between USB power +5vdc or external power via a on-board +5vdc regulator. Also makes a convenient on/off power switch if only one power source is being used.
On-board 3.3vdc regulator providing much more current then the internal FTDI 3.3vdc regulator.
The FTDI chip is only powered by USB +5vdc. Makes sense because if you are not connected to a PC USB port there is no reason for the FTDI chip to be powered on.
The latest Seeeduino 328 board also has a switch the board to operate as a 3.3v or 5.0 board. However keep in mind that operating a 328 chip at 3.3vdc and 16mhz is not within the 328 datasheets operating specifications, but few have reported problems running in that mode.
Unrelated to power design, the Seeeduino also provides a switch to enable or disable the Arduino auto-reset function which can be quite useful at times.
Just throwing out some ideas and thoughts. 
Lefty
OT: And wrt to 4 further aspects the Seeeduino is also the better design
@ort11
Have you understood how the comparators work?
- The FTDI chip is only powered by USB +5vdc. Makes sense because if you are not connected to a PC USB port there is no reason for the FTDI chip to be powered on.
One reason could be that you have other uses for the Rx/Tx pins, or is this a non issue with the Seeduino design?
One reason could be that you have other uses for the Rx/Tx pins, or is this a non issue with the Seeduino design?
Not sure I understand the question?
Lefty
Not sure I understand the question?
You may want to use the AtMega USART for communication with serial devices other than FTDI and then run the risks of parasittic power through to the FTDI chip. This may be harmful to the FTDI chip and anything else you may have in the circuit.
When the FTDI chip is powered, we can trust its datasheet to reflect how it will behave. Without power we don't really know - or do we?
You may want to use the AtMega USART for communication with serial devices other than FTDI and then run the risks of parasittic power through to the FTDI chip. This may be harmful to the FTDI chip and anything else you may have in the circuit.
Possible I guess, but I would think that the two 1k ohm series resistors between the FTDI and pins 0 & 1 would prevent any high current situations that could cause any damage.
Lefty