Weird Things Happening At My Relay @.@

[sunmix]

after starting a thread about 24v - 5vdc convert, i finally managed to get a regulated 5v- cheers to everyone who contributed information on that thread!

the 5v is meant for my 7406 buffer, so it produces enough currents and voltage to kick up my big and bulky solid state relay, which is connected to a mechanical relay.

Then, i am trying to diversify all these circuits into electronics, and the problem starts here.

A parallel port pin supplys 5v signal to the buffer, and after buffer comes the pull up of 10k ohms, and next is to a blue 5v electronic relay, shown on the pic. I am running mach3, so i am able to stimulate the process by turning on and off the active low and high settings.

As i turn it into active high, nothing happens.

I asked a local electronics guy regarding this problem and he says that I should be using a 3v relay, for this application. So i bought the white 3v relay, and tried it out. nothing happened again.

But i've notice that if my pin is at active High, the voltage is at 4.2vdc at free load, and after I connect it to those 2 relays, it drops to 1.5-2vdc.

I took both relays and tried it using 3x1.5v batteries (4.5v) and it manages to trigger the relay.

What exactly happened? Help me..! going nuts soon!

[HillBilly]

The 7406 is a open collector devise. You are relying on the 10k resistor to turn on the relay. The way to do this is to put voltage on the relay and let the 7406 ground it.

Darek

[sunmix]

HillBilly,
Do you mean by taking away the 10k resistor or putting the 10k resistor before the 7406?

[Al_The_Man]

As Darek said, with open collector (7406) the 10k pull up resistor should be replaced by the relay coil. Make sure the 7406 can sink the current required by the coil (ohms law).
You may also want to look at relay drivers like the DS2003 these have reverse emf diodes installed, are darlington output and can be used to switch 24vdc relays, with 5v input. They come 7 to a chip.
You also need to wire a 1n4005 or similar across the coil, reverse biased, or you may lose the 7406 due to back emf on switch off.
Al.

[sunmix]

em, can you kindly draw out a simple schematic plan? I couldn't really understand! Sorry, my bad, my english is not that good..

[Al_The_Man]

This should explain it. http://www.kpsec.freeuk.com/components/relay.htm
Al.

[NC Cams]

A way to look at it is this: sourcing current versus sinking current.

A parallel port can fairly easily "sink" current (pull down a signal that is pulled high via an external pull up resistor).

When you ask a parallel port to "source current" as in provide enough power to actually drive the device high that you're trying to drive, that's when things can become difficult.

Might want to google "parallel port interfacing".

Yes, parallel ports can do neat things BUT sometimes the actual way they "communicate" with external devices is NOT intuitively obvious.

[Al_The_Man]

Originally Posted by NC Cams A parallel port can fairly easily "sink" current (pull down a signal that is pulled high via an external pull up resistor). When you ask a parallel port to "source current" as in provide enough power to actually drive the device high that you're trying to drive, that's when things can become difficult. .

But It looks like he is already looked at buffering, hence the 7406.
Al.

[pastera]

The Songle relay in your picture needs 78mA for the coil and the 7406 is only rated for 48mA.

Take a look at the link Al posted. Under the heading "Protection diodes for relays" you will se a schematic. This is what you need. Use a 2n3904 for the transistor and a 1.6K base resistor.

Aaron

[Al_The_Man]

This is what makes the DS2003 nice is that it will drive up to 350ma.
and has the rev. diode in the chip.
For single driver, I prefer the 2N7000 as no other resistors are required usually.
Al.

[NC Cams]

The 7406 is both current limited in its ability to source current AND it depends what it is using for its +5vdc source. I don't think we know what that source is, do we????

A common trick is to use pin 17 (I think) off the LPT port as a 5v source for pull up reference. In a reference book I have (that I can't find or else I'd pos the ISBN for it), it clearly states that this pin is NOT a reliable voltage source - yet, folks persist on using it.... Hence, it may also NOT provide the current needed to properly drive a buffer IC.

If there is a stiff power supply powering the 7406, it is probably a decent buffer IF its current sourcing capabilities are not exceeded.

For a decent amount of current, you might be better off driving a 2N4401/2N4403 totem pole or better yet, perhaps a fet driver. In either case, a wall bug would be what I'd use for the +5vdc voltage source.

[Al_The_Man]

Originally Posted by NC Cams For a decent amount of current, you might be better off driving a 2N4401/2N4403 totem pole or better yet, perhaps a fet driver. In either case, a wall bug would be what I'd use for the +5vdc voltage source.

I disagree with the 2n4401 over the 2n7000, the 2n7000 is a 'Fetlington' Darlington equivalent in a fet and is TTL compatible input, this is what it is designed for as a TTL driver to buffer high current loads.
Personally if you are going to alternatively source a relay power supply, why go with 5vdc? I would recommend at least 24vdc.
But if you want to stick with the PC power supply, there is already a high current 5v source using one of the power supply HD connectors.
Al.