Triggering a 5v relay board

Jee Labs Talk » Batteries & power

Triggering a 5v relay board

(7 posts) (3 voices)

Tags:

No tags yet.

craig

After learning how to fire a relay I found a nice little PCB at Sparkfun that I decided to deploy.
It's designed for a 5v circuit and works great with Paul Badger's RBBB.
I tried it "right out of the box" with a Jeenode, but although I could light the indicator LED on the relay board, I could not fire the relay as expected.
So I hooked the Jeenode up to a 9v battery and pulled it off one of the Jeenode's ports (using the 'P' pin) and ran it through a LM7805 regulator, smoothed it with a couple caps and a diode and have a real nice 5v supply to the relay board.
However, I'm getting the same behavior - the LED lights, but the relay doesn't close.
To clarify, I'm using 3 pins on a single Jeenode port. 'P' and 'G' serve as input to the 7805, 'D' is connected to the 'control' pin of the relay board. The relay board gets the 5v and ground out of the 7805 (after being smoothed by the caps and diode).
Voltages across all boards are what I expect them to be.
On another project, I have mixed 3.3v and 5v devices on an I2C bus and had to do some level shifting with a couple mosfets.
Am I running into the same problem here? It doesn't seem so because the voltages look good.
Thinking through, it seems that I have confused the transistor on the relay board and it is dumping electrons to the wrong ground.
If anyone would care to help lift the veil of ignorance it would be greatly appreciated.
There's obviously a fundamental requirement that I'm not grasping.

Posted 15/04/2010 15:51:07
joop

electric seems ok , maybe your software doen't control the right D ? , see port pinmaping .

Posted 15/04/2010 17:33:01
jcw

Puzzling. I agree with Joop - electric side seems ok. All I can think of is that the transistor gain is marginal - with 3.3V driving it instead of 5V, you are providing less current through the 1 k? resistor. Measure the collector voltage when turned on, it should be 0.4..0.7 V when driven in saturation. Compare with driving it from a 5V unit. If it's not driven strongly enough, consider adding a 1 k? resistor in parallel to the existing one - it'll still easily be within the current sourcing specs of an ATmega output pin.
All the different grounds are connected together, right?

Posted 15/04/2010 18:58:53
joop

no message

Posted 15/04/2010 19:01:30
craig

Exactly what do you mean by "All the different grounds are connected together, right?"
Intuition tells me that the ground coming out of the LM7805 mini-circuit is causing the problem. I thought I had connected it everywhere on the board, but perhaps not.
To which ground on the JeeNode should I connect it?
Thanks for the help.
Everything else checks out and makes sense.

Posted 18/04/2010 13:32:47
jcw

Oh, sorry, didn't mean to confuse. All GND pins on the JN are connected internally. Just wanted to make sure you connected one of those JN ground pins with relay ground.
Can you make the relay turn on by shorting its input pin to +3V (or +5V)?

Posted 18/04/2010 17:42:59
craig

I used to be able to. Pretty reliably for a while. It stopped.
You wrote...."Just wanted to make sure you connected one of those JN ground pins with relay ground."
I'm an electronics novice, but I've done a ton of reading. A lot of it I've read twice and a full understanding is just beyond reach ... but I keep getting closer.
I've also designed and breadboarded a "transistor as switch" circuit and am getting the same results. That's really what drove me to the instinct-based "transistor dumping to wrong ground" conclusion.
Makes sense. I have no problem doing it with these voltages. You know... the kind I can put a tongue on.
Let's say, though, that I'm switching 24volts (using the appropriately sized transistor and protective components, of course).
I'm having a hard time bringing myself to put that ground to my JeeNode. That's where it goes, I suppose.

Posted 20/04/2010 02:16:07