Re: Amplify 1.5v DC to 5v DC?

From: Jonathan Kirwan (jkirwan_at_easystreet.com)
Date: 10/25/04 

Date: Mon, 25 Oct 2004 20:48:37 GMT

On 25 Oct 2004 04:58:15 -0700, "mark.mcgee@csfb.com" <mark.mcgee@csfb.com>
wrote:

>Is it possible in PSPICE to generate a schematic from the NETLIST?

It *would* be easy for the software to just drop down the parts with the
indicated connections, but I'm not sure how it would also decide what "looks
good" to your eye. It might even just lay the parts on top of each other in a
kind of "rat's nest" and you'd have to tease it apart, visually.

>Anyway, I think I understand how they work - just the transistor
>numbering to understand. From my own schematic, I think the sequence
>is c-b-e.

Yes, Q1 is "node 3" as collector, "node 1" as base, and "node 2" as emitter,
with the transistor being a 2N3904.

>Working with that, I get input in to base of Q1,

Yes, from V2 as the pulse generator.

>resistor R1 at emitter, and R2 at collector.

Yes. Of course, you didn't mentioned where the other end of any of these went,
but I can assume you can figure it out from what you've already said.

Please note that node 0 is always "ground".

>That's not what I was modelling.

Okay.

>I thought we were discussing providing an input to the emitter? That's
>what I was attempting to model.

Okay. I figured there was a disconnect between us going on.

>I'm fine understanding a cct, which provides a signal in to the
>transistor base, but this will invert, won't it? I thought there was a
>proposal for a non-inverting, single transistor amplifier for my data
>lines, which took an input in to the emitter?

My mistake, then. With all the various options floating around, I guess I
didn't know *what* you were simulating. Here is what I imagine, now:

              +5
              |
              / R1
              \ 47k
              /
              |
              +---Vout
              |
        R2 |/c Q1
+1.5--/\/\--| 2N3904
       220k |>e
              |
             pulse
             input (1.5V pulses)
              |
             gnd

or,

Q1 Vout 2 IN 2N3904
R1 V5 Vout 47k
R2 2 V1.5 220k
V1 V1.5 0 1.5
V2 V5 0 5
V3 IN 0 PULSE(0 1.5 0 1n 1n 625u 1.25m)

That is from KM's post, with values I suggested.

>What difference does it make, having R1 in the cct (in your netlist
>above) then?

First off, I had thought you were curious about that circuit because it is used
as one part of that more complex circuit you asked about. Not because I
believed you'd actually use it, since it doesn't do what you want. It inverts
the input, for one thing, and I already know you don't want that. Second, it
doesn't provide anything like a 0V to 5V output span and I know you want that.

That said, in that example the emitter resistor is required because the input
pulsing is directly coupled to the base. You cannot reasonably expect to have
1.5V across the base-emitter diode without absolutely HUGE base currents
flowing. Placing a resistor on the emitter leg allows the emitter to "float"
with the base. It would simply be "bad news" without it. Of course, you could
use a base resistor but then that's a different topology.

>Also, why the preference for a NPN-PNP transistor pair for a
>non-inverting cct (in your previous posts), and not just a collector of
>one NPN feeding in to the base of another NPN?

There were several examples provided, but all of them would just leave the
signal inverted if the output transistor were replaced with an NPN. Plus, the
output levels would be 'way off,' as well. Perhaps, if you were to expose your
own thinking on this question, I (or someone) could help you see better why this
is so. It's hard to guess what you are thinking, when all you do is suggest a
part change. Try writing more about why you think it may work okay.

Jon

Quantcast