Powering Demo-board during development

henkvbeek

Hello,

I've just received the propeller demo board and ran the withcoming demo program.
I also tried the first example and succeeded, Hurah.

Now I am wondering: Is there a possibility to power the board from USB during development.

Looking at the schematics, there is not.
In my opinion it can be done easely or do I oversee something ?

Henk

Mike Green

Theoretically this is possible, but there are rules about drawing current from the USB. USB peripherals are required to disconnect themselves from the bus except for a very small amount of current at certain times and there is a limit of 100ma in most circumstances normally. A device can declare itself as a high current device (up to 500ma), but it has to start out as a low current device first. To avoid damaging the USB master (PC or hub), there needs to be some protective circuitry and the demo board just doesn't have it.

henkvbeek

When I start the demo-board it draws about 85 mA. This is, I think the most it will draw in the development phase. The demo uses more cogs. When I run a more simple program the current drops below 25 mA.

Do I make a serious fault when I connect a diode between the USB-connecter and the input of the 3.3 V regulator or will it damage the 5 V regulator ?

It is just so much simpeler when I can connect the board simply to the PC alone.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
<img src=C:\Documents and Settings\Henk\Mijn documenten\Mijn afbeeldingen\Henk.jpg>

What God wants, God gets . . .

cgracey

That would probably work just fine. I imagine you could solder a piece of wire wrap wire from the USB Mini-B connector to the input of the 3.3V regulator, no problem. Please tell us if this works. Maybe we should put a jumper on the next rev of the Demo Board for this purpose.

henkvbeek said...
When I start the demo-board it draws about 85 mA. This is, I think the most it will draw in the development phase. The demo uses more cogs. When I run a more simple program the current drops below 25 mA.

Do I make a serious fault when I connect a diode between the USB-connecter and the input of the 3.3 V regulator or will it damage the 5 V regulator ?

It is just so much simpeler when I can connect the board simply to the PC alone.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


Chip Gracey
Parallax, Inc.

Phil Pilgrim (PhiPi)

You also need to limit the inrush current that comes from charging the filter capacitors in your power circuit at powerup. Otherwise, the FTDI chip will reset itself when it's Vdd dips below a certain threshold. This entails a "soft-start" circuit of some sort, switched by one of the FT232's outputs. FTDI shows one in their tech support docs, but it's not a very good one. To avoid making the necessary connection to the FTDI chip's power enable pin, you could probably get by with just a delay circuit that switches on the soft-start transistor a fixed time interval after being powered up.

-Phil

henkvbeek

Chip,

It all works great !

The only disadvantage I see is:

What happens if the board is connected to the PC and external power is applied. In some cases the board will supply current to the USB-port.
It can be circumvented by 3-pole strap, External vs USB supply.

===============================================

Phil,

I can't see that problem. It seems to be too theoretical (and I hope it is).

Thanks

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
[img]C:\Documents and Settings\Henk\Mijn documenten\Mijn afbeeldingen\Henk.jpg[/img]

What God wants, God gets . . .

Mike Green

This is not theoretical. Most USB masters (PCs and hubs) use one of a series of power switches (see Maxim's website for examples) that consider any current draw more than 100ma as a fault condition and disconnect power from the USB port at least briefly, then may try again after a delay. Others I think do limit the current to 100ma (until the USB device is recognized as a high power device, then they switch to a 500ma current limit).

Phil Pilgrim (PhiPi)

If it works, it works, and I'm happy for you! We do things differently sometimes when it's for our own use and not a commercial product. When designing a commercial product for others to use, the "theroretical" suddenly becomes important!

-Phil