• Please review our updated Terms and Rules here

Driving Emitters in a TTL Gate

ziloo

Veteran Member
Joined
Feb 7, 2006
Messages
990
Location
in the basement
Hello Folks,

In a situation when two TTL logic gates, namely TTL-Out and
TTL-In are connected, and the TTL-Out logic is HIGH....
how can we have 40 uA of current flowing into the TTL-In gate
when the inlet on the TTL-In gate is an emitter?

ziloo :mrgreen:
 


Alright...it is leaking....but where does it go?
The emitter is forward-biased!

ziloo :mrgreen:


Actually, it is reverse biased or at least close to. You need to be more specific as to what condition it is measured at. Are you talking with the input at 5V or at a minimum logic one or what?
Dwight
 
A transistor has 2 junctions. They are physically close to each other. They are so close that if you inject a current forward bias into the base emitter junction, holes or electrons ( depending on PNP or NPN ) get pulled by the reverse biased base to collector, into the depletion region of the collector to base. Once in that region, they can't go back, against the voltage, and become collector current.
Surprisingly, if you turned the transistor upside down and wired the emitter to the collector circuit and the collector to the emitter circuit, it would still work the same but with reduced current gain ( as long as the collector circuit voltage was less than about 6.8V ). If the voltage was too high, the emitter to base would zener.
In order to analyze the typical TTL input you'd have to understand how this happen. When the input voltage is high, of a TTL, the emitter is working like a collector and the collector is biasing the next transistor on, as though it were an emitter. In a regular TTL load there would be about 1.6ma flowing forward bias from the input transistors base, forward biased through the collector junction. As I said, when the biasing of the transistor is reversed, there is some gain making the emitter look like a collector, supplying 40 ua on top of the 1.6ma into the following transistors base lead, turning it on.
I hope that helps a little.
To get your terms right, Forward bias is current flowing, reverse bias is less current flowing. When you turn a transistor on, the collector current is flowing through what otherwise would have been a reversed biased junction except for the stray current carriers from the base, cause by the forward biased base to emitter. You can swap emitter and base leads and still have transistor action, just not as much gain.
I should mention, the emitter base junction is highly doped, making it zener when reverse biased at about 6.8V, for a typical transistor.
Dwight
 
The bottom picture (lower case) shows the current flowing towards the emitter
in the second TTL gate. Where does the flow go.....?

View attachment 52212


ziloo :mrgreen:

You are missing the transistor following the input transistor.
The current is ( not electrons, in physics current flows from positive to negative. Ben Franklin got it wrong ) flows from the top transistor of the output through the emitter of the input transistor through the collector of the input transistor , along with the forward biased base current of the input transistor through the base emitter of the transistor that you don't show. With the input transistor in the reversed operation, it has a gain of about 0.025. Not a particularly useful gain but it is still some gain. The emitter is working like a collector and the collector is working like an emitter as a poor gain transistor.
Dwight
 
I'm not following, but evidently Dwight is, thank heavens. Draw a box around everything but the power supply. The sum of the currents going into the PS is the same as the currents flowing the reverse direction. Simplify what you're thinking with Norton equivalents.

Reverse operation of BJT's isn't common, but it does have its uses. I've seen BJT's operated in collector-to-collector or emitter-to-emitter configurations between the power rails. Germanium BJTs seem to be preferred over silicon.
 
He was right that the emitter is reverse biased. What he doesn't see is that the entire transistor is biased like the emitter is acting as a collector ( poor gain of 0.025 ) but, still, some gain from the current flowing through the base and the forward biased collector. It is still working like a transistor and injecting current carriers into the reverse biased junctions depletion region. It becomes current in the reverse bias emitter-base.
Dwight
 
I'm not following, but evidently Dwight is, thank heavens. Draw a box around everything but the power supply. The sum of the currents going into the PS is the same as the currents flowing the reverse direction. Simplify what you're thinking with Norton equivalents.

Reverse operation of BJT's isn't common, but it does have its uses. I've seen BJT's operated in collector-to-collector or emitter-to-emitter configurations between the power rails. Germanium BJTs seem to be preferred over silicon.

In the case of the early germanium transistors, there was little difference between the emitter and the collector. They basically took a thin piece of germanium doped as the collector and emitter and made it thinner where they attached the base that was diffused into what was left of the thin path from emitter to collector.
In silicon, first they start with material, only doped with the collector material. They then dope the base with enough material to make it the opposite of the collector ( just by over powering the previous doping ). They then diffuse the emitter. This is so heavily doped that it makes a vary poor transistor when the roles are reversed but still has some transistor action ( 40 ua with about 1.6 ma base current ).
Dwight
 
Thank you both for your comments!!

One thing that I am not still clear about is that when the TTL-Out is
logically High at a voltage of about 2.4V, there is still (5V-2.4V=2.6V)
voltage difference between the emitter and the base of the input transistor on
TTL-In. In this situation the emitter of the input transistor is not in "Reverse Bias"!!!
Or.....I don't quite understand the meaning of "Reverse Bias".....

ziloo :mrgreen:
 
Thank you both for your comments!!

One thing that I am not still clear about is that when the TTL-Out is
logically High at a voltage of about 2.4V, there is still (5V-2.4V=2.6V)
voltage difference between the emitter and the base of the input transistor on
TTL-In. In this situation the emitter of the input transistor is not in "Reverse Bias"!!!
Or.....I don't quite understand the meaning of "Reverse Bias".....

ziloo :mrgreen:

RB.JPG
 
Thank you both for your comments!!

One thing that I am not still clear about is that when the TTL-Out is
logically High at a voltage of about 2.4V, there is still (5V-2.4V=2.6V)
voltage difference between the emitter and the base of the input transistor on
TTL-In. In this situation the emitter of the input transistor is not in "Reverse Bias"!!!
Or.....I don't quite understand the meaning of "Reverse Bias".....

ziloo :mrgreen:

That was an incorrect statement based on the limited information I had at the time. You didn't state the conditions so I was incorrect.
The Base lead of the input transistor is 2 diode drops from ground ( from the collector of the input transistor and base-emitter of the transistor connected to the collector of the input transistor ( that is missing in your pictures ). 2 diode drops is about 1.3V. That is 2.4V - 1.3 = 1.1V of negative bias or reverse bias.
Dwight
 
Thank you again Dwight :D !!

I read your comments several times and I got a much better
picture of TTL innards. I have to do a lot more readings still....

ziloo :mrgreen:
 
Dwight, on the page that I pointed to, the author claims a 12dB gain from a "flipped" 2n5088, which is a bit surprising. In the "normal" position using the same circuit, he says that he gets 25dB gain. He also explains the difference in diffused- vs. alloy-junction transistors.

I used to have a 2n21A point-contact transistor in my collection--it got lost somewhere along the way. I wonder how well it would work in "flipped" mode. Probably about as well as it did in normal mode, given the variations in P-C samples.
 
Dwight, on the page that I pointed to, the author claims a 12dB gain from a "flipped" 2n5088, which is a bit surprising. In the "normal" position using the same circuit, he says that he gets 25dB gain. He also explains the difference in diffused- vs. alloy-junction transistors.

I used to have a 2n21A point-contact transistor in my collection--it got lost somewhere along the way. I wonder how well it would work in "flipped" mode. Probably about as well as it did in normal mode, given the variations in P-C samples.

My memory of the germanium transistor was wrong. It starts with the base material that is thinned and then the emitter and collector are diffused into the base. This is somewhat shown in the picture on the page Chuck pointed to. It doesn't show the thinning that was required to get the diffusion of the two sides close enough to get gain.
Dwight
 
I was reading the article about reversing the transistor and one thing caught my eye. He mentioned that when clipping it made a lot of noise. It just occurred to me why. You'll recall that I mentioned that the reverse emitter-base voltage should be below around 6.8V. This is the typical zener voltage of the highly doped emitter to base. It is interesting that this voltage is right at the point that the diode changes from two different types of zener action. It goes from a knee like curve at the breakdown to an avalanche type breakdown ( actually 6.2V ). This breakdown in the zener region is actually quite noisy. In a home project I made years ago, I used a reverse biased emitter-base junction as a white noise source for an ocean wave sound generator. In the service, I needed an extra white noise generator for looking at RF bandwidth. For this I used a 6.2V diode ( as I recall ). I suspect using a 9V supply he was getting into the region of this noise.
Just an interesting side note. Should you ever need a noise source.
Dwight
 
The reverse-biased PN junction has been used as a shot noise generation for decades. In some applications, the noise is used to drive a counter to create a true random-number generator.

Zener noise is different from avalanche noise, even though a "zener" diode is involved in both. Forward-biased PN junctions also have noise characteristic.

Of course, a vacuum or gas diode can also be used. You can even use a resistor.

The bigger issue to me is avoiding noise. :)
 
You'll recall that I mentioned that the reverse emitter-base voltage should be below around 6.8V. This is the typical zener voltage of the highly doped emitter to base.
Dwight

There are some interesting things about this. Zener diodes in the range of 7 to 9V tend to have a small positive temperature coefficient, about equal to the negative tempco of a forward biased transistor B-E junction which is about -2.1mV/degC. In circuitry such as early alternator and some dynamo regulators, the B-E junction of the regulator's input transistor, which is used like a comparator reference voltage, is temperature compensated, with a series zener of a voltage in this range. In fact one company, Bosch, used an actual transistor with a reversed B-E junction as the zener (circuit attached) shows the collector not connected, for their dynamo regulator.

Another interesting thing, is that if you relentlessly zener the B-E junction of a transistor, it slowly degrades its hfe. There were some research papers done by Motorola on the theoretical mechanisms on how this comes about. One circuit, prone to do this is a typical multivibrator, when its power supply is over 7V if no protection diodes are added in the base circuit. Over time the transistors get so degraded, it won't start.
 

Attachments

  • REG.jpg
    REG.jpg
    90.2 KB · Views: 1
Last edited:
Back
Top