Parasite power

Paul M

I've been using/evaluating/playing with the propeller for the last few weeks and I'm very impressed. I've been using a USB/Serial converter cable (3.3V) from FTDI. I am using only the RX line (+GND) on the serial cable and "Simple_Serial.Spin" to provide a debug port so I can get·info back from the propeller.

One day, I disconnected power from the circuit but had left the serial cable plugged in and noticed that the propeller was still happily running and·sending data back to the PC (I was monitoring a couple of 1-wire DS18B20's which were also powered from the same supply lines).

This behaviour is not unique to the propeller, but what is the mechanism? , is it possible to damage the chip?, and how does one avoid this problem?
Even if the chip may not be damaged, I suspect that depending on the load, data may be corrupted; not very clever in a commercial application.

Paul

···

Mike Green

There are protective diodes on all I/O pins that are reverse connected to Vdd and Vss. These prevent the I/O pin circuitry from going more than a diode drop above or below Vdd and Vss respectively. If Vdd is zero (like when there's no power), any input voltage on an I/O pin above about 0.6V will begin to power the chip and, if above 2.6 + 0.6 = 3.2V, will actually be within spec for minimum supply voltage. The problem is that the diodes and their conductors on the chip are not made for continuous conduction of any significant supply current (see the specs in the datasheet).

You have two options (other than disconnecting the USB to serial converter when the power is off):

1) Put a pull down resistor on the RX line to load it down.

2) Put a Shottky diode between the converter cable's 3.3V output and the Propeller's Vdd line so the cable can properly power the Propeller (and other circuitry) when the main power supply is off. You'll also need a Shottky diode between the main power supply and the Propeller if the regulator is not designed for this situation.

deSilva

It is a common situation that the USB Plug remains connected to PC/Notebook when the Propeller's main power is down. Parallax should have taken means against ESD overload when designing their boards...

But most likely the current from the USB Plug is limited (<1 mA) in contrast to the current from a standard PC RS-232 interface..

Paul Baker

deSilva said...
Parallax should have taken means against ESD overload when designing their boards...

We did, thats why the parasitic power situation occurs, it is the ESD protection diodes that are the conduit. The I/O pins are protected beyond what the test fixture was capable measuring (8kV HBM). If you are thinking that some form of resistance should have been added to prevent the parasitic power situation from burning out the diodes, that would have a deleterious effect on the ESD performance (the diodes are there to provide a safe path to ground, impeding that path will increase the likelyhood of grounding through the oxide and damaging the input structure of the pin).

With the boards themselves, there is no "killer value" of resistor to incorporate, some applications need added protection, but some need the rise and fall times to be as short as possible on output signals.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer

Parallax, Inc.

Post Edited (Paul Baker (Parallax)) : 11/19/2007 9:47:10 PM GMT