Driving pins high or low for power supply.

Oldbitcollector (Jeff)

In the last couple posting I've submitted to the Gadget Gangster blog I've used a technique of driving two pins (one high/one low) for the purpose of supplying power to the project.

Here's the examples:

http://www.gadgetgangster.com/news/56-jeffs-shop/509-easy-game-controls-for-your-quick-module.html
http://www.gadgetgangster.com/news/54-wednesday-project/507-propeller-powered-pumpkin-lights.html

This is not a technique I've seen commonly on the forums, so I thought I'd post it here to let the EE wizards here to chew on.

According to Parallax Technical support, the concept is fine as long as I don't draw more than 40 mA from the pins.

Opinions?

OBC

Mike Green

There's not much to chew over. There's one MOSFET that connects one output pin to Vdd and another MOSFET that connects the other output pin to Vss. There's a small voltage drop across these MOSFETs (about 0.3V each) that causes the resulting supply voltage to be a little lower than the Propeller's ... about 0.6V. If you start with 3.3V, you end up with 2.7V for your supply voltage. Whether that's enough or not depends on what you're connecting to the two output pins. The other issue with this technique is that you've got a shift in the ground voltage. The Vss of your project board is about 0.3V above the Vss of your Propeller. That's not enough to cause problems with the protection diodes on the I/O pins of your project circuitry, but it does reduce the noise margins of the I/O pins on the project board (when you connect logic signals back to the Propeller).

If you don't have to reverse the voltage to the project board, you could do better by just switching Vdd and using the same ground as the Propeller is using. That reduces the voltage loss to only 0.3V and it avoids the ground shift.

Phil Pilgrim (PhiPi)

This is the method I used to power the new altimeter module from the Propeller Demo and Quickstart boards:

http://www.phipi.com/29124_altimeter.html#lbl3

But the altimeter requires so little current that there's hardly any voltage drop.

-Phil

pjv

Sure Jeff, I do it all the time.... for years in fact, starting with the SX.

For some sensor applications it actually has great benefits as you can reduce/eliminate the bonding wire voltage drop in the Prop's internal ground and power leads. I use this technique to power differential bridge pressure strain gauges with good success. At low currents it brings the excitation voltages closer to the rails inside the Prop, and hence improves the Sigma-Delta A/D converter concept.

Cheers,

Peter (pjv)

David B

I've never used two pins to provide both power and ground, but I've powered a Seetron serial backpack LCD, which only draws about a milliamp, from a data pin.

That's handy because if strings of data sent to the LCD that are supposed to be ASCII accidently contain binary data, it can put the LCD into some unknown mode where it won't respond, but with a data pin powering it, it's easy to power down/up the LCD to force a hard reset without having to power down the entire propeller unit.

WBA Consulting

Excellent! I have also done this for the Sensirion SHT11 sensor because of its tiny power requirements (8mA when doing measurements)