Transistor Selection

[gtrplaya101]

Ok I have moved to a new thread because I am taking a new route in fixing my issue. I am now going to use a transistor to control a relay. My question is what base resistance do I need to ensure the transistor goes into the saturation zone and what transistor should I use?

Heres my specs:
Supply voltage: 5v
Relay resistance: 50 ohms
Voltage to base : 2v

Thanks in advance.

[R.G.]

Ok I have moved to a new thread because I am taking a new route in fixing my issue. I am now going to use a transistor to control a relay. My question is what base resistance do I need to ensure the transistor goes into the saturation zone and what transistor should I use?

Heres my specs:
Supply voltage: 5v
Relay resistance: 50 ohms
Voltage to base : 2v

If the supply voltage is 5V and the coil resistance is 50 ohms, the current is going to be close to 5V/50 = 100ma. That is big enough to start worrying about whether the transistors you're using are up to it. And that worry is settled by looking at the datasheet.

You don't mention the transistor, but I would use a 2N3904, which I happen to know is specified for up to 100ma. The higher gain/lower noise transistors like the 2N5088/5089/MPSA18 may not work at that current. A quick trip to google gives this: http://www.fairchildsemi.com/ds/2N/2N3904.pdf

First we check the absolute maximums to see if it's qualified: Vceo max is 40V, and Ic max is 200ma. Both are not only OK, they have some excess capability, or safety factor, depending on how you look at things.

Since (as you know from my harangues on hfe in this forum) hfe varies with Ic, we need to see what gain we get at 100ma. Fortunately, the 3904 lists it in the normal operating specs: Hfe at 100ma with Vce=1V is a minimum of 30. Translating that into non-EE-ese means: you can get any one of these transistors to pull 100ma down to a Vce of 1V by inserting 100/30 = 3.33ma of base current into the base. What it doesn't tell you is that most of them will pull 100ma to lower than 1V with 3.33ma of base current,  you're just expected to have been listening in class that day.

Ordinarily we could just hand-wave the saturation voltages Vcesat and Vbesat; however, you're using 5V, and that's going to make the variations big enough compared to 5V to have to check them. So we go look at the typical curves section of the datasheet. On page 3, upper right corner we find a good graph: Collector-emitter saturation voltage vs collector current. This tells us what the actual saturation voltage will be. Looking over at 100ma, we see that even at 125C, the saturation is under 0.15V. This is good! But that's that "beta=10" thing mean? What it means is that the curves are all graphed with a base current equal to 1/10 of the collector current supplied to the base. But the gain was 30, right?

No. The active area gain was at least 30 with Vce=1V. 1V isn't very saturated. To get the collector-emitter voltage down from 1V to 0.15V, you have to shovel in more base current to saturate it harder. The subliminal message is to count on a saturation gain of 10. So we now know to get 100ma down to 0.15V on the collector, we need to pump 10ma into the base.

Hang on, we're getting there.

What's the base emitter voltage under these conditions? Glad you asked - that's the graph one row down and on the left, Base-Emitter saturation voltage vs collector current. We find that Vbe could be 0.8 to 0.9V (I don't think you're gonna run it a -40C!). So now we can compute the base resistor.

You say you have 2V to provide base drive, right? The Vbe will be as much as 0.9V, and we have to provide 10ma. The answer is  - Ohm's law. 2v-0.9V = 1.1V across the resistor, and 10ma through it, the resistor has to be 1.1V/0.01A = 110 ohms.

Some interesting question have no doubt popped up in your head.
- but I'm using a 43DS21342 transistor; how do I figure this out for those transistors? [find a datasheet or use a different transistor; or just try it and if it blows up, then get another transistor]
- But my 2V base driver can't supply 10ma! [bummer; you need a better base drive or a higher resistance relay coil]
- 110  ohms seems very low [yep, it does; I don't like drives like this, but that's what you told me you had to work with]

[WLS]

What about the 2N3565 or BC237B. This might suit your needs!

[Gus]

http://www.diystompboxes.com/smfforum/index.php?topic=74956.0

gtrplaya101 is using a mm74c373.   I am guessing the 2V is because of the limited drive current I did not ask if it was a loaded or unloaded output voltage reading.

I did give some terms for a google search.

  I would think a 2n3904 like R.G. posted or other transistors with  hfe Ic that are useful with the relay coil in the collector leg with a damping diode across the coil would work.

[R.G.]

gtrplaya101 is using a mm74c373.   I am guessing the 2V is because of the limited drive current I did not ask if it was a loaded or unloaded output voltage reading.

Ah. That's a different kettle of fish. In that case, what's needed is higher input impedance, not lower drive resistors. The 'C373 has very limited output current at any voltage. We want something that can pull 100ma with a 5 volt drive. And the winner is - a MOSFET. Take a look here: http://www.fairchildsemi.com/ds/BS/BS170.pdf, page 3, "On Region Characteristics". With 5V on the gate and the source grounded, we look at the line for Vgs=5V, and see where that line crosses 100ma. It's way over on the left, where all the lines start running together. And it's under 0.2V, best I can eyeball it. For this, the gate resistor is zero - it runs from the 5V the 'C373 puts out.

If you want to slow it down a little, you can do this directly at the gate without a feedback capacitor. You put a series resistor back in (!?) and another capacitor from gate to ground. Now the square, sharp signal from the 'C373 output goes through the R-C and presents a much-slowed-down signal to the gate, which slowly pulls the coil on and off. There will be much less transferred ticking in the relay.

Remember a gate zener of about 10V, cathode to gate, anode to source, and a coil catch diode across the coil to snub any coil flyback pulses.

I would think a 2n3904 like R.G. posted or other transistors with  hfe Ic that are useful with the relay coil in the collector leg with a damping diode across the coil would work.

It would. But the BS170 is literally designed for things like this, and they're no longer expensive.

[gtrplaya101]

Wow you guys are amazing! I guess my question now is why am I only getting 2v out of my 74c373 chip? I am running it at 9v so shouldnt I be getting 5v? Not sure if this matters but I am using a huge resistor to pull the inputs to ground when there is no voltage. Somewhere around 4.5M ohms. Also (im still learning as you can tell) shouldnt I be able to supply 12v and get a 10v output from the chip? Maybe Im miss reading the data sheet altogether. Thanks