Please tell me. I am using 2n2222 collector load resistor of 125 ohm and +5V, Vcc supply....what should be the value of base resistor to drive to transistor in saturation, if +5V base supply is used?

 

hi c2,
Assuming that all the 5v is dropped across the 125R when the 2N2222 is hard on, the Collector current will be 5v/125R= 40mA.
The 2N222s has a good hFE gain of approx 200 at 40mA, refer to this image, that would make I Base = 0.04/200 = 0.2mA [ 200uA]

To ensure saturation I would set Ibase = 0.4mA.
So from a 5v signal and allowing for a Vbe drop of say 0.6V, the Rbase value = [5v - 0.6v]/0.4mA = 11k say 10k.

E

 

If you want to be sure the transistor fully saturates then assume a gain of 10. That makes the base resistor 1.2k.

 

The curves in post #2 are indicate that you should use a beta of 10 for saturation mode. You want a low Vce to minimize power dissipation in the transistor, and a low Vce has a significant impact on beta.

 

The rule of thumb is that for hard saturation, the base current should be 10% of the collector current. But this rule comes from the 60's, when the average transistor had much less gain, and the gain dropped off rapidly at higher collector currents. Today's parts are much better, so for a small signal transistor with a typical gain of 100, I start with 5% base current.

ak

 

If the goal is to minimize power dissipation in the transistor, then pushing it too far into saturation will work against you. Yes, you reduce the power dissipation associated with the collector current, but you increase the power dissipation associated with the base current. Given that Vbe > Vcesat, the base-emitter dissipation can easily dominate.

Consider the above case with Ic = 10 mA. At Ib = 1 mA the Vce is about 20 mV giving a Pce of 0.2 mW while the Pbe, assuming Vbe = 0.6 V, is 0.6 mW for a total transistor dissipation of 0.8 mW. Note that the power associated with the base current is 80% of the total power.

If you look at that curve you'll see that even if you use 1% (Ib = 0.1 mA) you would (typically) still have a Vce of under 0.1 V. But this IS for typical transistors, probably at room temperature, and you are near the knee that marks the transition from active to saturation, so this is probably pushing it closer than you might like (though, in most cases, even if you are up the knee far enough to give you a quarter volt you are still at a Vce that has historically been considered the saturation voltage for back-of-the-envelope stuff). So anything over 2% will almost certainly work and 5% gives you a nice, comfortable margin.

So let's look at the 5% case more closely. The Vce appears to be about 40 mV, so Pce has doubled to 0.4 mW. But the Pbe is halved to 0.3 mW giving a total dissipation of 0.7 mW.

The absolute dissipation in this region is usually sufficiently low that finding the optimal point isn't worth the effort, especially since the optimal point varies with temperature and from transistor to transistor.

 

The rule of thumb is that for hard saturation, the base current should be 10% of the collector current. But this rule comes from the 60's, when the average transistor had much less gain, and the gain dropped off rapidly at higher collector currents. Today's parts are much better, so for a small signal transistor with a typical gain of 100, I start with 5% base current.

ak

The rule may come from the 60's but so does the 2N2222.

It was first announced by Motorola in 1962.

 

The rule may come from the 60's but so does the 2N2222.

It was first announced by Motorola in 1962.

Was it really? Doesn't time fly when you're having fun.

 

The rule may come from the 60's but so does the 2N2222.

It was first announced by Motorola in 1962.

Queue the comments about why anyone would use an old obsolete part when more modern transistors are available

2N2222/2N2907 and 2N3904/2N3906 are still my workhorses...

 

Those are not obsolete parts. Just look how many manufacturers still make them.

Why use them? A good percentage of online tutorials are a rehash of old designs and data. Old designs use old parts.

I too use the 2N2222 in some form or another. Why change after 45 years?

 

I too use the 2N2222 in some form or another. Why change after 45 years?

Most of the parts I use were designed 30+ years ago...

 

The rule may come from the 60's but so does the 2N2222.
It was first announced by Motorola in 1962.

True, but it was an exceptional part at the time, the main reason it imprinted on an entire generation of designers. Also, the 2n3055 is of the same vintage and had a worst case gain of 20.

ak

 

See

 

Whenever this part comes up I would be remiss if I failed to mention the time I was called in to fix the SSN 774 Virginia after the first sea trials to replace some 2N2222's that failed for excessive sat voltage.

Seems the Mil Version Jan parts do not have a spec for sat voltage, and normal production runs can get quite high. We had to screen for this as the build was qualified and hence locked in stone.

 

Was the fix needed because the systems actually failed to perform, or was the failure purely because there were transistors that failed to meet the requirements?

Do you recall what the saturation voltage specs were?

Did you bin them exactly against that spec, or did you set the selection criteria more stringent to give a margin of safety (even though, presumably, the spec'ed sat voltage included a margin of safety in the design as well)?

 

The problem was someone chose these transistors as switching elements for a rather large rotary style relay that drew significant current, I don't remember how much exactly. When the sat voltage went up the transistors would fail and the crew had to manually adjust water in the hot well tanks that hold the working water for the nuclear reactor.

Since there was no spec to test against I made one up, probably 0.1 or 0.2 volts at the expected current. Rigged up a test box and off we went.

Anything failing would be trash as the parts met the full spec.

 

Whenever this part comes up I would be remiss if I failed to mention the time I was called in to fix the SSN 774 Virginia after the first sea trials to replace some 2N2222's that failed for excessive sat voltage.

Seems the Mil Version Jan parts do not have a spec for sat voltage, and normal production runs can get quite high. We had to screen for this as the build was qualified and hence locked in stone.

<Off Topic>
I still have my T-Shirt and a ball point pen from when we shipped the first Virginia Class CRDM (Control Rod Drive Motor) shipset. The pen and shirt are dated 2000. Come May 1st I will have been retired 4 years but Navy Power was very good to me. </Off Topic>

The 2N2222 was a workhorse transistor in countless military hardware over all the years, including panels of dozens of 327 lamps driven by 2N2222 transistors. Yeah, 1962 sounds about right, I was about 12 years old and they still produce them.

Ron