2N3094 as a relay driver? Better alternatives?

[Frost]

Hey there,

I'm using RG's 74C373 channel switching article as a basis for my amp's channel switching. I'm pretty sure I cannot drive the relay's directly from the 74C373. I noticed that he uses BS170 mosfets to drive relays. Problem is, I've got a lot of 2N3904s on hand but for BS170s, I would have to send an order to Mouser. Can I use a circuit which uses 2N3094s to drive my relays (Omron 12 Volt G5V-2-H1) like this one?

http://www.uoguelph.ca/~antoon/circ/relay5.html

Would it be a more robust alternative to uses mosfets instead? I'm concerned with durability above all else. Thanks,

-Heiko

[R.G.]

Ohm's law to the rescue again!

By typing in "Omron relay" into Google then looking for signal relays, G5 series, I find that the Omron G5V-2 12V relay has a coil resistance of 288 ohms.

12V/288ohms - 47ma. (thank you, Georg)

So any driver needs to pull 47ma through the coil without dying.

I then typed "2N3904" into google and found the Fairchild 2N3904 datasheet at http://www.fairchildsemi.com/ds/2N/2N3904.pdf

According to the datasheet, the 2N3904 in a TO-92 package can withstand up to 200ma of current - good! But can it do this without dying and how much base drive does it need?

Checking page 3 of the datasheet, I find that the typical pulsed current gain of the 2N3904 is over 100 at 40-50ma - also good, because this tells us that the base drive is less than 47ma/100 = 470uA. This is well within the capability of the CMOS 373.

So - will it overheat?

The data sheet doesn't help. It shows the collector-emitter saturation voltage as under 0.1V at 50ma of collector current at a beta of 10. That means that they expect to measure the 0.1V at 50ma if you force 1/10 of the collector current into the base. That's 4.7ma. There is no data on beta = 100.

We have to look further. The dissipation chart on page 4 shows power dissipation of 0.5W for the TO-92 in 50C ambient. So we take the current (47ma) and divide that into 0.5W, and get 10.6V across the device before it dissipates 1/2 W. We can easily get it down below 10.6V (!!!) so yes, it will be below 1/2W and will not burn up. Lots of room available.

Good news! It works fine! You will need to change the base resistor to provide at least half a milliamp of base current. Ohm's law can help you there, too.

[cd]

If you want to be super safe, you can use two 2N3904s in a darlington configuration, right?

[beans_amps]

Not really,  If you use two 2N3904's as a darlington pair, you still have a 3904 used as the pass element.  The darlington pair just helps to improve the base drive of the second transistor and provides much higher gains for the pair than a single transisitor would have.

If necessary you could pair up a 3904 with a higher power transistor, but it will not be necessary in this case.

Sean Weatherford
Bean's Amp Repair
Central, SC

[Frost]

So RG,

for the base resistor value: 12 volts/0.0005 amps = 24K, would this be right? By the way, thanks not only for help on this forum, but your website has been an incredible resource as well.

-Heiko