Question about scr's

[bench]

My question is about the simulation of an scr with two transistors,
one npn, one pnp. Will it always work this way, ... because scr's
can conduct large currents, whereas with the two transistors the
whole current must flow through a base-emitter junction, isn't this
so, and are transistors capable of conducting such large base
currents ?

 

[CFoley1064]

Subject: Question about scr's
From: "bench" bench@neti.com
Date: 5/18/2004 9:34 AM Central Standard Time
Message-id: <c8d6t8$8kd$1@news.epidc.co.kr

My question is about the simulation of an scr with two transistors,
one npn, one pnp. Will it always work this way, ... because scr's
can conduct large currents, whereas with the two transistors the
whole current must flow through a base-emitter junction, isn't this
so, and are transistors capable of conducting such large base
currents ?

The modelling of an SCR as back-to-back transistors is something to give you
the idea. The dies are optimized in SCRs for the circuit action, and actual
SCR and other thyristor specs are based on die topology. Take a look at the
Teccor AN-1001, "Fundamental Characteristics of Thyristors, for a basic
description:

http://www.littelfuse.com/data/Application_Notes/AN1001.pdf

(Teccor is now part of Littelfuse, I guess.)

Good luck
Chris

 

[Mantra]

"bench" <bench@neti.com> wrote in message news:<c8d6t8$8kd$1@news.epidc.co.kr>...

My question is about the simulation of an scr with two transistors,
one npn, one pnp. Will it always work this way, ... because scr's
can conduct large currents, whereas with the two transistors the
whole current must flow through a base-emitter junction, isn't this
so, and are transistors capable of conducting such large base
currents ?

Yes, using a .subckt this is often done. The tricky part is that an
SCR involves positive feedback between the two transistors a so small
parameter (e.g. BF, etc.) shift can cause the overall behavior to
change significantly. Things like trigger currents and such. If you
can find an existing subcircuit model for your device that someone
else has debugged that will probably help to avoid premature
loss/graying of your hair. Personally I wouldn't put faith in my own
home-made SCR model without having the real thing (SCR) there and
charactering my model against actual measured data.

MM