# Low on-resistance bipolar transistor - does it exist?

Discussion in 'Electronic Basics' started by John Popelish, Apr 6, 2007.

1. ### John PopelishGuest

Saturated BJTs are not as simply ohmic as mosfets are, but
they can have pretty low forward drops. I don't know what
range of voltage and current you are talking about
switching, but Zetex makes some small, very low drop
transistors. E.g. ZTX1049A
http://www.zetex.com/3.0/pdf/ZTX1049A.pdf
has a collector to emitter drop of no more than 220 mV while
carrying 4 amps of collector current and driven with only 50
mA of base current. The resistance with a similar drop at 4
amps would be 0.055 ohms. And this is a little TO-92 sized
device. Unfortunate, the resistance is not linear with
variable current. In other words, at zero collector
current, the collector to emitter voltage does not go to zero.

3. ### John PopelishGuest

This doesn't sound like a problem with the transistor, but
with the circuit design. Could you send a schematic?

4. ### John PopelishGuest

I can't say whether that is the method, or an incremental
method is used (small change in collector current causes a
small change in collector voltage drop, Ron= delta V / delta
I ).
Yes. The base current leakage into the collector produces a
minimum collector voltage drop for a given base current,
even if the external collector current goes to zero. I
think that for any given collector current, there is an
optimum base drive, for lowest collector to emitter drop.
More is not necessarily better.

5. ### Michael BlackGuest

Some of the mythology about MOSFETs is based on not thinking things through.
They have very high input impedance, which means that unloaded, high
voltages can remain high voltages when applied to the gate. There's nothing
to dissipate the high voltage. A bipolar sees the same voltage on its
base from static discharge, but the relative low impedance of the input
means the "high voltage source" gets loaded down significantly, and isn't
high enough to do damage. The voltage doesn't change, the "voltage
sensitivity of the input" doesn't change, the impedance of the input of
the device changes.

Hence, MOSFETs are more prone to static damage when lying loose. It
was a real problem in the early days (or at least every time a MOSFET
was mentioned, they mentioned being very very careful about protecting
the device), though it was relatively soon after that the manufacturers
started putting protection diodes on the gates to help protect against
this (the protective diodes would not conduct normally, but would conduct
when the voltage on the gate went above a certain level, providing a path
to discharge the high static voltage on the gate).

But the worry about static damage was with the MOSFET lying loose. Once
it was installed in a circuit, you could remove the jumper around the
leads (some devices in the early days came with such a jumper, and
if it didn't all the articles warned about putting your own jumper
on the device before you started soldering). The circuit itself
would generally protect the device, because there was a very real
path to ground from the gate in the form of a resistor or something
else that would provide a path to ground (such as a coil in the case
of radio circuitry). The circuits could still be relatively high
impedance, but even with relatively high value resistors they loaded
things down enough to dissipate the static, if it ever got to that
point in the circuit.

The real exceptions would be circuitry where a MOSFET gate is somehow
open to the world, and there's nothing to protect that input. So
something like an electrometer that depends on a really high input
impedance to do it's work and needs the gate exposed to the world
might suffer. That MOSFET as an "active antenna" in that shortwave
receiver (where the device is more to transform a really high impedance
point of the whip antenna to a lower impedance that the rest of
the receiver can use) might be vulnerable, if the manufacturer didn't
put protective diodes in the circuit. Plug in boards that have a MOSFET
input connected to the connector might be vulnerable, if there is no
pullup or pulldown resistor on the gate that provides a discharge
path for static electricity on the pin.

Michael

6. ### tempus fugitGuest

Hey all;

characteristics. I have seen low (and very low) on resistance MOSFETS (2
ohms and less) but haven't seen a similar spec in a standard BJT. In fact,
sifting through the Fairchild site, one can sort the MOSFETS based on Ron,
but the same isn't true for the BJTs. Are some BJTs lower Ron than others,
or is low Ron something that BJTs simply are not capable of? It's my
understanding that a MOSFET is more susceptible to damage from static
discharge than a BJT, and so it would seem to limit it's applications as a
switch if you weren't supposed to have any voltage at the drain before
things were powered up.

Thanks

7. ### tempus fugitGuest

Thanks for the reply John. I'm switching 12v at the collector with 5v at the
base. The 12v is coming through a 100K resistor from the previous device, so
the current is pretty small. I've found that using a standard (2N2222 type)
transistor doesn't quite switch things all the way off ( there is an LED
indicator that still glows faintly).

8. ### tempus fugitGuest

Oh, one other thing - is this how the Ron is calculated, i.e., Collector to
emitter drop divided by collector current?
If this were the case, my low current would increase Ron.

9. ### jasenGuest

you probably want to look at Vces with BJTs it serves a somewhat
equivalent purpose

Bye.
Jasen

10. ### tempus fugitGuest

Thanks for the input Michael. I actually have a MAX4662 analog switch
to use for this purpose, but once I got learning more about MOSFETS, I
decided against using it because it was possible that the 12v I'm trying to
switch may be applied to the com terminal before the power was actually on.
But you're saying that this shouldn't be a problem? If not I'd wire it in
and give it a try.

John, I have a schematic but unfortunately I don't have a scanner, so I'll
try to take a picture and post it on ABSE. The schematic is from my guitar
amp, and it is a control signal for channel switching and volume boost that
I'm trying to switch remotely.