Hi: In the link below, assuming Q4 is at cutoff, is Q3 in saturation or at cutoff? Or neither?

https://sites.google.com/site/lofturj/tf3lj/AVR_HVS_Programmer.jpg

 

I'd say cutoff. :\

 

If Q4 is off, Q3 is off. R5 cuts off Q3 until Q4 allows some current to flow from the emitter to base of Q3.

 

Thank you. So we can say that the two transistors circuit (Q3, Q4) switch the regulator (7805) output off and on (controlled by the DB25 connector output signal), with a small voltage drop (aprox. 0.2v) across the collector-emitter of Q3. Is that right?

 

Thank you. So we can say that the two transistors circuit (Q3, Q4) switch the regulator (7805) output off and on (controlled by the DB25 connector output signal), with a small voltage drop (aprox. 0.2v) across the collector-emitter of Q3. Is that right?

Yep (456789)

 

And cannot the same thing be done with two NPN transistors instead of using the NPN-PNP pair?

 

Not in that circuit. You have to design a different circuit to go around reversing polarities.

 

But it would be two transistors and four resistors the same as before, true? Then why use PNP?

 

Why did I wear a grey shirt today? The green one has 2 sleeves, same as the grey one, right?

What's your point? Do you want me to read the mind of the original designer or imagine what is on the other side of the DB25 connector?

Can you design a circuit with npn transistors that is going to let Q6 receive the same signal and do its job at the same time?

 

I ask because NPN transistors are more widely used than PNP, I think. And you see that the PNP made me think a lot. That is, why make complicated when it can be simple? The DB25 connector is a PC desktop computer parallel port, that used with parallel printers. The whole circuit is just a programmer for a microcontroller (kind of a microprocessor).

 

The PNP operates with the opposite polarity of the NPN so when the NPN turns on, so does the PNP. If you had two NPNs in the same circuit configuration, then when Q4 turns OFF, Q3 would turn ON and vice versa. Think about how the relative voltages and currents are in an NPN as compared to a PNP and you will see why this is so. If NPNs and PNPs were interchangeable then there would be no need for both, would there?

 

I ask because NPN transistors are more widely used than PNP, I think. And you see that the PNP made me think a lot. That is, why make complicated when it can be simple? The DB25 connector is a PC desktop computer parallel port, that used with parallel printers. The whole circuit is just a programmer for a microcontroller (kind of a microprocessor).

Sorry, I didnot read the last sentence. We could assume the Q5-Q6 circuit has been removed. Anyways, its output, the 8-contact connector P1, is a pin of a microcontroller. Initially, it is an input (to the mcontroller). After Q1-Q2 apply the +12v, it is set by the same microcontroller to an output. That is, Q5-Q6 must provide a two-way path. How Q5-Q6 accomplish this is a mystery for me.

 

Well, there is where we have a difference of opinion. To me, transistors are just Leggos. Whichever polarity is most convenient gets installed. I can think in either polarity just as conveniently, and the weaknesses that were in the pnp type in the 1960's are no longer a problem.

I will admit that npn's are more often used, but that's only because of the tradition of using a positive voltage supply. That makes most circuits start with an npn, but it's just habit. As soon as the circuit starts taking form, it's whichever fits, and if there's no difference in parts count or price, it is only a matter of opinion whether the circuit ends up with more pnp's than npn's. People with a lot of experience don't see pnp circuits as more complicated. They are just a Leggo built up side down.

 

The PNP operates with the opposite polarity of the NPN so when the NPN turns on, so does the PNP. If you had two NPNs in the same circuit configuration, then when Q4 turns OFF, Q3 would turn ON and vice versa. Think about how the relative voltages and currents are in an NPN as compared to a PNP and you will see why this is so. If NPNs and PNPs were interchangeable then there would be no need for both, would there?

Yes, I see. Now, the input to the circuit is provided, via the DB25 connector, by a computer program. So, the program could be changed (reversing the output polarity) so that all could be done with NPN transistors.

 

Well, there is where we have a difference of opinion. To me, transistors are just Leggos. Whichever polarity is most convenient gets installed. I can think in either polarity just as conveniently, and the weaknesses that were in the pnp type in the 1960's are no longer a problem.

I will admit that npn's are more often used, but that's only because of the tradition of using a positive voltage supply. That makes most circuits start with an npn, but it's just habit. As soon as the circuit starts taking form, it's whichever fits, and if there's no difference in parts count or price, it is only a matter of opinion whether the circuit ends up with more pnp's than npn's. People with a lot of experience don't see pnp circuits as more complicated. They are just a Leggo built up side down.

As crutschow points out, the only difference would be a change of polarity. I think this was done this way (NPN-PNP) because the circuit is driven by a computer program, and this program is a standardized one, not easily modifiable. And yes, though I went four years to the university, I have little experience with electronics and PNP transistors make think a lot. Some time ago, I had simple rules which, applying them, made apply what I knew about NPN to PNP in a matter of seconds. Now, I have forgotten them.

 

the program could be changed (reversing the output polarity) so that all could be done with NPN transistors.

True. If there really was a difference in performance, price, or difficulty, it would be done that way.

 

Thank you very much.