Question about common-collector circuits.

[pawihte]

I remember reading a long time ago (I think it was a magazine
article) a cautionary note about operating a common-collector BJT
with the collector tied directly to the power supply rail. It
recommended inserting a resistor between the collector and Vcc.
What I don't remember is the reason given. The only thing I can
think of is where there's a possibility of the base being driven
above Vcc, forward biasing the base-collector junction. In that
case, a resistor in series with the base would also serve the
purpose of limiting the current. Is there another factor I
haven't thought of? TIA.

 

[Jon Kirwan]

On Fri, 27 Nov 2009 14:42:22 +0530, "pawihte" <pawihte@invalid.com>
wrote:

I remember reading a long time ago (I think it was a magazine
article) a cautionary note about operating a common-collector BJT
with the collector tied directly to the power supply rail. It
recommended inserting a resistor between the collector and Vcc.
What I don't remember is the reason given. The only thing I can
think of is where there's a possibility of the base being driven
above Vcc, forward biasing the base-collector junction. In that
case, a resistor in series with the base would also serve the
purpose of limiting the current. Is there another factor I
haven't thought of? TIA.

With a high supply rail, the collector resistor might allow a
smaller/survivable Vce for the emitter follower? I'm just imagining
that the emitter voltage is Vb less a diode drop and that Vc is the
supply rail (without the resistor), so it's possible to have quite a
fair sized Vce and thus quite a lot of dissipation in the BJT itself
if there is no collector resistor. Does any of that connect with your
recollection?

Jon

 

[krw]

On Fri, 27 Nov 2009 14:42:22 +0530, "pawihte" <pawihte@invalid.com>
wrote:

I remember reading a long time ago (I think it was a magazine
article) a cautionary note about operating a common-collector BJT
with the collector tied directly to the power supply rail. It
recommended inserting a resistor between the collector and Vcc.
What I don't remember is the reason given. The only thing I can
think of is where there's a possibility of the base being driven
above Vcc, forward biasing the base-collector junction. In that
case, a resistor in series with the base would also serve the
purpose of limiting the current. Is there another factor I
haven't thought of? TIA.

It's normally used to limit the current in the circuit. If the base
voltage rises the emitter current rises, causing the collector current
to rise and the collector voltage to fall (I*R in the collector
resistor). Eventually the CB junction becomes forward biased, the
transistor saturates, "Beta" (Ic/Ib, anyway) drops, and the current is
limited at that point.

You also may need a small resistor (~50R) in the base to keep the
transistor form oscillating.

 

[John Larkin]

On Fri, 27 Nov 2009 14:42:22 +0530, "pawihte" <pawihte@invalid.com>
wrote:

I remember reading a long time ago (I think it was a magazine
article) a cautionary note about operating a common-collector BJT
with the collector tied directly to the power supply rail. It
recommended inserting a resistor between the collector and Vcc.
What I don't remember is the reason given. The only thing I can
think of is where there's a possibility of the base being driven
above Vcc, forward biasing the base-collector junction. In that
case, a resistor in series with the base would also serve the
purpose of limiting the current. Is there another factor I
haven't thought of? TIA.

I can't see any reason to do that in general, without some specific
hazard to avoid.

Emitter followers do tend to oscillate, but a series base resistor is
a better fix than a collector resistor.

John

 

[Tim Wescott]

On Fri, 27 Nov 2009 14:42:22 +0530, pawihte wrote:

I remember reading a long time ago (I think it was a magazine article) a
cautionary note about operating a common-collector BJT with the
collector tied directly to the power supply rail. It recommended
inserting a resistor between the collector and Vcc. What I don't
remember is the reason given. The only thing I can think of is where
there's a possibility of the base being driven above Vcc, forward
biasing the base-collector junction. In that case, a resistor in series
with the base would also serve the purpose of limiting the current. Is
there another factor I haven't thought of? TIA.

A base resistor would have a much less reliable current limiting action,
as the HFE of the device varies over temperature and from device to
device.

If you're only ever driving one thing and it's under your control and it
won't pull too much current with the transistor on hard, then you don't
need a current-limit resistor on the collector side.

Having a current-limit resistor on the collector side is certainly a
trick to keep in your bag-o-tricks, but I wouldn't think you'd need to
use it universally.

--
www.wescottdesign.com

 

[John Larkin]

On Fri, 27 Nov 2009 14:42:22 +0530, "pawihte" <pawihte@invalid.com>
wrote:

I remember reading a long time ago (I think it was a magazine
article) a cautionary note about operating a common-collector BJT
with the collector tied directly to the power supply rail. It
recommended inserting a resistor between the collector and Vcc.
What I don't remember is the reason given. The only thing I can
think of is where there's a possibility of the base being driven
above Vcc, forward biasing the base-collector junction. In that
case, a resistor in series with the base would also serve the
purpose of limiting the current. Is there another factor I
haven't thought of? TIA.

Possibly they were concerned about the possibility of a shorted
external load.

But there's an immense amount of nonsense in magazine articles and web
sites.

John

 

[pawihte]

Jon Kirwan wrote:

On Fri, 27 Nov 2009 14:42:22 +0530, "pawihte"
pawihte@invalid.com
wrote:

I remember reading a long time ago (I think it was a magazine
article) a cautionary note about operating a common-collector
BJT
with the collector tied directly to the power supply rail. It
recommended inserting a resistor between the collector and
Vcc.
What I don't remember is the reason given. The only thing I
can
think of is where there's a possibility of the base being
driven
above Vcc, forward biasing the base-collector junction. In
that
case, a resistor in series with the base would also serve the
purpose of limiting the current. Is there another factor I
haven't thought of? TIA.

With a high supply rail, the collector resistor might allow a
smaller/survivable Vce for the emitter follower? I'm just
imagining
that the emitter voltage is Vb less a diode drop and that Vc is
the
supply rail (without the resistor), so it's possible to have
quite a
fair sized Vce and thus quite a lot of dissipation in the BJT
itself
if there is no collector resistor. Does any of that connect
with your
recollection?

It was about emitter followers in general and not about any
specific design, and I'm fairly sure the example they gave had a
100-ohm resistor in series with the collector. I don't remember
what the emitter resistor was, but I have a strong impression
that it was a low-power circuit where the emitter resistor would
keep dissipation well within safe limits.

 

[krw]

On Fri, 27 Nov 2009 23:44:57 +0530, "pawihte" <pawihte@invalid.com>
wrote:

Tim Wescott wrote:
On Fri, 27 Nov 2009 14:42:22 +0530, pawihte wrote:

I remember reading a long time ago (I think it was a magazine
article) a cautionary note about operating a common-collector
BJT
with the collector tied directly to the power supply rail. It
recommended inserting a resistor between the collector and
Vcc. What
I don't remember is the reason given. The only thing I can
think of
is where there's a possibility of the base being driven above
Vcc,
forward biasing the base-collector junction. In that case, a
resistor in series with the base would also serve the purpose
of
limiting the current. Is there another factor I haven't
thought of?
TIA.

A base resistor would have a much less reliable current
limiting
action, as the HFE of the device varies over temperature and
from
device to device.

If you're only ever driving one thing and it's under your
control and
it won't pull too much current with the transistor on hard,
then you
don't need a current-limit resistor on the collector side.

I'm pretty sure the article was talking about low-power c-c
circuits in general where it can be assumed that there's enough
resistance on the emitter side to keep current and dissipation
within safe limits, no matter what the HFE is. The exception
would be if the base is driven higher than the collector without
an appreciable resistance on either base or collector.

Or, as mentioned earlier, you short the emitter.

Having a current-limit resistor on the collector side is
certainly a
trick to keep in your bag-o-tricks, but I wouldn't think you'd
need to
use it universally.

 

[pawihte]

Tim Wescott wrote:

On Fri, 27 Nov 2009 14:42:22 +0530, pawihte wrote:

I remember reading a long time ago (I think it was a magazine
article) a cautionary note about operating a common-collector
BJT
with the collector tied directly to the power supply rail. It
recommended inserting a resistor between the collector and
Vcc. What
I don't remember is the reason given. The only thing I can
think of
is where there's a possibility of the base being driven above
Vcc,
forward biasing the base-collector junction. In that case, a
resistor in series with the base would also serve the purpose
of
limiting the current. Is there another factor I haven't
thought of?
TIA.

A base resistor would have a much less reliable current
limiting
action, as the HFE of the device varies over temperature and
from
device to device.

If you're only ever driving one thing and it's under your
control and
it won't pull too much current with the transistor on hard,
then you
don't need a current-limit resistor on the collector side.

I'm pretty sure the article was talking about low-power c-c

circuits in general where it can be assumed that there's enough
resistance on the emitter side to keep current and dissipation
within safe limits, no matter what the HFE is. The exception
would be if the base is driven higher than the collector without
an appreciable resistance on either base or collector.

Having a current-limit resistor on the collector side is
certainly a
trick to keep in your bag-o-tricks, but I wouldn't think you'd
need to
use it universally.

 

[Tim Wescott]

On Fri, 27 Nov 2009 23:44:57 +0530, pawihte wrote:

Tim Wescott wrote:
On Fri, 27 Nov 2009 14:42:22 +0530, pawihte wrote:

I remember reading a long time ago (I think it was a magazine article)
a cautionary note about operating a common-collector BJT
with the collector tied directly to the power supply rail. It
recommended inserting a resistor between the collector and Vcc. What
I don't remember is the reason given. The only thing I can think of
is where there's a possibility of the base being driven above Vcc,
forward biasing the base-collector junction. In that case, a resistor
in series with the base would also serve the purpose of
limiting the current. Is there another factor I haven't thought of?
TIA.

A base resistor would have a much less reliable current limiting
action, as the HFE of the device varies over temperature and from
device to device.

If you're only ever driving one thing and it's under your control and
it won't pull too much current with the transistor on hard, then you
don't need a current-limit resistor on the collector side.

I'm pretty sure the article was talking about low-power c-c circuits in
general where it can be assumed that there's enough resistance on the
emitter side to keep current and dissipation within safe limits, no
matter what the HFE is. The exception would be if the base is driven
higher than the collector without an appreciable resistance on either
base or collector.

Remember that just because someone's an idiot doesn't mean they can't get
published in a trade journal. Look at me -- I'm an idiot, and I have
articles published in Embedded Systems Design.

--
www.wescottdesign.com

 

[pawihte]

John Larkin wrote:

On Fri, 27 Nov 2009 14:42:22 +0530, "pawihte"
pawihte@invalid.com
wrote:

I remember reading a long time ago (I think it was a magazine
article) a cautionary note about operating a common-collector
BJT
with the collector tied directly to the power supply rail. It
recommended inserting a resistor between the collector and
Vcc.
What I don't remember is the reason given. The only thing I
can
think of is where there's a possibility of the base being
driven
above Vcc, forward biasing the base-collector junction. In
that
case, a resistor in series with the base would also serve the
purpose of limiting the current. Is there another factor I
haven't thought of? TIA.


I can't see any reason to do that in general, without some
specific
hazard to avoid.

Emitter followers do tend to oscillate, but a series base
resistor is
a better fix than a collector resistor.

That's what I thought, to both statements. As I'm pretty sure the

article was about low-power common-collectors in general, in most
cases there would be enough resistance in series with both base
and emitter to keep currents and dissipation within safe limits.

Anyway, my question does not concern any practical design I have
in mind. The thing keeps popping up in my mind now and then, and
it bugs me that I can't remember exactly why they made that
point.

 

[Rich Grise]

On Fri, 27 Nov 2009 14:42:22 +0530, pawihte wrote:

I remember reading a long time ago (I think it was a magazine article) a
cautionary note about operating a common-collector BJT with the collector
tied directly to the power supply rail. It recommended inserting a
resistor between the collector and Vcc. What I don't remember is the
reason given. The only thing I can think of is where there's a possibility
of the base being driven above Vcc, forward biasing the base-collector
junction. In that case, a resistor in series with the base would also
serve the purpose of limiting the current. Is there another factor I
haven't thought of? TIA.

I can't think of any reason that makes sense, unless you've got a really,
really crappy power supply, in which case the PS design needs to be fixed.

Frankly, I think your writer was blowing smoke out his a$$.

Don't forget the supply pypass caps, however.

Hope This Helps!
Rich

 

[Rich Grise]

On Fri, 27 Nov 2009 10:04:27 -0800, John Larkin wrote:

On Fri, 27 Nov 2009 14:42:22 +0530, "pawihte" <pawihte@invalid.com> wrote:

I remember reading a long time ago (I think it was a magazine article) a
cautionary note about operating a common-collector BJT with the collector
tied directly to the power supply rail. It recommended inserting a
resistor between the collector and Vcc. What I don't remember is the
reason given. The only thing I can think of is where there's a
possibility of the base being driven above Vcc, forward biasing the
base-collector junction. In that case, a resistor in series with the base
would also serve the purpose of limiting the current. Is there another
factor I haven't thought of? TIA.


Possibly they were concerned about the possibility of a shorted external
load.

But there's an immense amount of nonsense in magazine articles and web
sites.

To protect it from a shorted output, the proper place is in the emitter
lead. To protect it from sharp transients on the BASE, then use a base
resistor. (or slow it down with a parallel cap.)

Cheers!
Rich

 

[Rich Grise]

On Fri, 27 Nov 2009 23:44:57 +0530, pawihte wrote:

Tim Wescott wrote:
On Fri, 27 Nov 2009 14:42:22 +0530, pawihte wrote:

I remember reading a long time ago (I think it was a magazine article)
a cautionary note about operating a common-collector BJT
with the collector tied directly to the power supply rail. It
recommended inserting a resistor between the collector and Vcc. What
I don't remember is the reason given. The only thing I can think of
is where there's a possibility of the base being driven above Vcc,
forward biasing the base-collector junction. In that case, a resistor
in series with the base would also serve the purpose of
limiting the current. Is there another factor I haven't thought of?
TIA.

A base resistor would have a much less reliable current limiting
action, as the HFE of the device varies over temperature and from
device to device.

If you're only ever driving one thing and it's under your control and
it won't pull too much current with the transistor on hard, then you
don't need a current-limit resistor on the collector side.

I'm pretty sure the article was talking about low-power c-c circuits in
general where it can be assumed that there's enough resistance on the
emitter side to keep current and dissipation within safe limits, no matter
what the HFE is. The exception would be if the base is driven higher than
the collector without an appreciable resistance on either base or
collector.

If that happens, there's something wrong with your design. And to protect

it from overvoltage on the base, use a diode reverse-biased from the base
to the Vcc rail, probably preceded by some resistance.

Cheers!
Rich

 

[Rich Grise]

On Fri, 27 Nov 2009 23:56:40 +0530, pawihte wrote:

John Larkin wrote:
On Fri, 27 Nov 2009 14:42:22 +0530, "pawihte" <pawihte@invalid.com

I remember reading a long time ago (I think it was a magazine article)
a cautionary note about operating a common-collector BJT
with the collector tied directly to the power supply rail. It
recommended inserting a resistor between the collector and Vcc.
What I don't remember is the reason given. The only thing I can
think of is where there's a possibility of the base being driven
above Vcc, forward biasing the base-collector junction. In that
case, a resistor in series with the base would also serve the purpose
of limiting the current. Is there another factor I haven't thought of?

I can't see any reason to do that in general, without some specific
hazard to avoid.

Emitter followers do tend to oscillate, but a series base resistor is
a better fix than a collector resistor.

That's what I thought, to both statements. As I'm pretty sure the article
was about low-power common-collectors in general, in most cases there
would be enough resistance in series with both base and emitter to keep
currents and dissipation within safe limits.

Anyway, my question does not concern any practical design I have in mind.
The thing keeps popping up in my mind now and then, and it bugs me that I
can't remember exactly why they made that point.

Because, as has been noted, the writer of the article was an idiot.

There could have been some valid reason, but if there were, you'd have
remembered it because it'd have been so unusual. ;-)

Cheers!
Rich

 

[krw]

On Fri, 27 Nov 2009 13:39:56 -0800, Rich Grise <richgrise@example.net>
wrote:

On Fri, 27 Nov 2009 23:44:57 +0530, pawihte wrote:

Tim Wescott wrote:
On Fri, 27 Nov 2009 14:42:22 +0530, pawihte wrote:

I remember reading a long time ago (I think it was a magazine article)
a cautionary note about operating a common-collector BJT
with the collector tied directly to the power supply rail. It
recommended inserting a resistor between the collector and Vcc. What
I don't remember is the reason given. The only thing I can think of
is where there's a possibility of the base being driven above Vcc,
forward biasing the base-collector junction. In that case, a resistor
in series with the base would also serve the purpose of
limiting the current. Is there another factor I haven't thought of?
TIA.

A base resistor would have a much less reliable current limiting
action, as the HFE of the device varies over temperature and from
device to device.

If you're only ever driving one thing and it's under your control and
it won't pull too much current with the transistor on hard, then you
don't need a current-limit resistor on the collector side.

I'm pretty sure the article was talking about low-power c-c circuits in
general where it can be assumed that there's enough resistance on the
emitter side to keep current and dissipation within safe limits, no matter
what the HFE is. The exception would be if the base is driven higher than
the collector without an appreciable resistance on either base or
collector.

If that happens, there's something wrong with your design. And to protect
it from overvoltage on the base, use a diode reverse-biased from the base
to the Vcc rail, probably preceded by some resistance.

No, a collector resistor does two things; reduces the transistor's
power dissipation and limits the emitter current. The collector
resistor is quite common, particularly if the emitter is a primary
output.

 

[John Larkin]

On Fri, 27 Nov 2009 13:37:49 -0800, Rich Grise <richgrise@example.net>
wrote:

On Fri, 27 Nov 2009 10:04:27 -0800, John Larkin wrote:
On Fri, 27 Nov 2009 14:42:22 +0530, "pawihte" <pawihte@invalid.com> wrote:

I remember reading a long time ago (I think it was a magazine article) a
cautionary note about operating a common-collector BJT with the collector
tied directly to the power supply rail. It recommended inserting a
resistor between the collector and Vcc. What I don't remember is the
reason given. The only thing I can think of is where there's a
possibility of the base being driven above Vcc, forward biasing the
base-collector junction. In that case, a resistor in series with the base
would also serve the purpose of limiting the current. Is there another
factor I haven't thought of? TIA.


Possibly they were concerned about the possibility of a shorted external
load.

But there's an immense amount of nonsense in magazine articles and web
sites.


To protect it from a shorted output, the proper place is in the emitter
lead.

That would increase the output impedance. A collector resistor acts
more like a hard current limit, with little effect up to the point of
limiting.

To protect it from sharp transients on the BASE, then use a base

resistor. (or slow it down with a parallel cap.)

A cap from base to ground is an excellent way to get an emitter
follower to oscillate at 100 MHz or so.

John