Please help me understand this partial schematic from Mega R3

  1. There are 6 lines going into the USB connector here. Why? USB connectors only have 4 lines. What do the two lines on the bottom of the USB connector indicate?

  2. BLM21 seems to be a ferrite inductor. But they are available in many different resistances. How do I tell which one is appropriate for this part?

  3. What is this part? I don’t recognize it. What does it mean?

Thank you. I want to put together a clone of the Mega R3 to test out if I can mount BGA and QFN parts successfully. This seems like it would be a nice challanging project.

  1. The 2 extra wires are going to the USB shield.
  2. Not sure.
  3. Is a solder bridge so (in this case) you can bridge between grounds or keep them seperate.

Riva: 1. The 2 extra wires are going to the USB shield. 2. Not sure. 3. Is a solder bridge so (in this case) you can bridge between grounds or keep them seperate.

By shield do you mean the metal enclosure around the pins? Thanks.

JoeN:

Riva: 1. The 2 extra wires are going to the USB shield. 2. Not sure. 3. Is a solder bridge so (in this case) you can bridge between grounds or keep them seperate.

By shield do you mean the metal enclosure around the pins? Thanks.

Yes.

Is the metal enclosure for USB even an active part? Is the circuitry for this just an attempt at lessening EMI? Should I just leave that stuff off? It saves on two parts that I don't own including the ferrite bead with an unknown value.

JoeN: Is the metal enclosure for USB even an active part? Is the circuitry for this just an attempt at lessening EMI? Should I just leave that stuff off? It saves on two parts that I don't own including the ferrite bead with an unknown value.

Sure, give it a shop. Commercial designs often have to apply for and pass FCC certification for EMI emission and rejection standards that a DIY may not have a problem with. I would assume you are not going to try and get FCC certification for your project?

Lefty

retrolefty:

JoeN: Is the metal enclosure for USB even an active part? Is the circuitry for this just an attempt at lessening EMI? Should I just leave that stuff off? It saves on two parts that I don't own including the ferrite bead with an unknown value.

Sure, give it a shop. Commercial designs often have to apply for and pass FCC certification for EMI emission and rejection standards that a DIY may not have a problem with. I would assume you are not going to try and get FCC certification for your project?

Lefty

Nope, it's just a advanced soldering /custom PCB project because I need more experience with those things. I'd like something useful if it works out is all. An extra few Mega R3s compatibles would be great to have if I can make it work. I'm going to leave off the power parts because I have a pretty decent bench supply and I am looking for other things I can trim off. I see the PRC fuses are cheap so I will keep that, all the caps and resistors, of course, most of the SMD leds, and the two main chips and headers. That USB grounding and regulator circuitry goes. The opamp driven LED on PB7 goes too, I don't think that is worth enough to bother. Actually, I am not even sure what that is the indicator for. It's marked 'L' and is a yellow LED.

http://arduino.cc/en/uploads/Main/arduino-mega2560_R3-schematic.pdf

I would have to say, if you have to ask about these kind of details, why do you want to mess with QFN and BGA packages ? If you want to lay out a board for such a complicated part, please learn basic electronics and schematics first. Just a small hint to get you started...

// Per.

Zapro: I would have to say, if you have to ask about these kind of details, why do you want to mess with QFN and BGA packages ? If you want to lay out a board for such a complicated part, please learn basic electronics and schematics first. Just a small hint to get you started...

// Per.

OK, since you are the schematic genius, what is the value of the BLM21 inductor?

It's a well kept secret the Arduino folks haven't divulged anywhere. I've seen hundred of commercial designs where this part is omitted though...

It's not recommended to connect the USB shield to ground in the design, and should be avoided...

https://forum.sparkfun.com/viewtopic.php?f=14&t=22122

// Per.

It's not recommended to connect the USB shield to ground in the design, and should be avoided.

It is recommended that the USB shield and the ground design has ONE common point and only ONE.

All shields must be connected to the ground in only ONE point to avoid circulation of parasistics current. And, of course, ground must be a "good" ground : large diameter wire "star-connected" or wide area of copper on the printed circuit.

From the FTDI datasheet:

A ferrite bead is connected in series with the USB power supply to reduce EMI noise from the FT232R and associated circuitry being radiated down the USB cable to the USB host. The value of the Ferrite Bead depends on the total current drawn by the application. A suitable range of Ferrite Beads is available from Steward (www.steward.com), for example Steward Part # MI0805K400R-10.

This is associated with USB RFI compliance, so it won't change significantly for different USB chips...

The opamp driven LED on PB7 goes too, I don't think that is worth enough to bother. Actually, I am not even sure what that is the indicator for. It's marked 'L' and is a yellow LED.

That is the standard pin 13 LED made famous and standard to support the infamous Blink example sketch.

Lefty

retrolefty:

The opamp driven LED on PB7 goes too, I don’t think that is worth enough to bother. Actually, I am not even sure what that is the indicator for. It’s marked ‘L’ and is a yellow LED.

That is the standard pin 13 LED made famous and standard to support the infamous Blink example sketch.

Lefty

Oh, OK, it’s the indicator indicator :). I think I will leave that off. I don’t want to have to bother with an opamp just drive an 0805 LED on the board since I don’t need the opamp for the power circuitry that I am going to leave off too.

So drive the LED directly, like the Duemilanove and early Uno's did, before the op amp buffer was used.

CrossRoads: So drive the LED directly, like the Duemilanove and early Uno's did, before the op amp buffer was used.

I gave that a bit of thought. My thinking is that they put that opamp there for a reason - it doesn't upset that pin very much at all because of input impedance. The spec gives 1K resistors for these leds, so a direct-drive LED is going to bleed off 5mA. Maybe that is not so bad, but it can't be the best solution. Obviously not good enough for the Arduino engineers.

Even less: (5V -Vf)/1K = 3mA with Vf = 2V. Not much load on a pin that can support 20mA minimum.

I think they just decided to use the op amp that had been sitting unused on the board for several design iterations. And now we see questions like "why is the LED on all the time?" when D13 is floating and trips the comparator level.

Good point, I always forget to factor in the forward voltage. Thanks

Maybe I should throw in an LEDs on each GPIO then the board can look like an old 1950's computer. Hmmm....

I've never found that to be all that useful unless things were fairly static, otherwise they just flash quickly, or if flashing quickly enough just look to be in various states of dimness.

JoeN:

retrolefty:

The opamp driven LED on PB7 goes too, I don't think that is worth enough to bother. Actually, I am not even sure what that is the indicator for. It's marked 'L' and is a yellow LED.

That is the standard pin 13 LED made famous and standard to support the infamous Blink example sketch.

Lefty

Oh, OK, it's the indicator indicator :). I think I will leave that off. I don't want to have to bother with an opamp just drive an 0805 LED on the board since I don't need the opamp for the power circuitry that I am going to leave off too.

Driving the led via a op-amp section buffer is a newer method that came upon the scene with the R3 versions of the Uno and Mega boards. Prior to that all standard arduino boards wired the series led/1K ohm resistor directly from pin 13 to ground. The on board pin 13 led can be useful as a troubleshooting tool and an indication of the bootloader running, etc. Having it is not a requirement but also it's not just blingy jewelry either. ;)

Edit: By the way this newer way of driving the led via a op-amp section in Rev3 boards is not without a flaw. Many users report the led being randomly on or off even when not being used in their sketch. As the pin will default to input mode the op-amp input is effectively seeing a 'floating input pin' condition and therefore no insurance of what state to drive the led, on or off. So to insure the led is off if you aren't using it in a sketch then one must set pin 13 mode to output and write a LOW to the pin. That's pretty poor design, they should have wired a 50k ohm pull down resistor from the opamp input to ground.