Whats the trick to getting a Darlington pair to turn off quicker?

[Teece]

Hi,

I am having trouble with the "low going edge" of a squarewave using a
Darlington pair that I am using to replace a single transistor. Since this
is an existing design, and I am trying to save the layout, etc, I don't have
complete freedom to change a lot of things around.

I am using the NPN Darlington as a Common Collector input buffer. Since it
is an integrated component (two transistors in a SOT-23 package) I don't
have the ability to add a resistor to the emitter of the "first" transistor
so as to bleed off charge during the low going edge of a square wave.

What would work best? Would reducing the base current help? In a Darlington
pair, it is already incredibly low.

Thanks
Tom

 

[Jim Thompson]

Hi,

I am having trouble with the "low going edge" of a squarewave using a
Darlington pair that I am using to replace a single transistor. Since this
is an existing design, and I am trying to save the layout, etc, I don't have
complete freedom to change a lot of things around.

I am using the NPN Darlington as a Common Collector input buffer. Since it
is an integrated component (two transistors in a SOT-23 package) I don't
have the ability to add a resistor to the emitter of the "first" transistor
so as to bleed off charge during the low going edge of a square wave.

What would work best? Would reducing the base current help? In a Darlington
pair, it is already incredibly low.

Thanks
Tom

Replace the Darlington with one that has the resistor built-in ?

...Jim Thompson

 

[Tim Wescott]

Hi,

I am having trouble with the "low going edge" of a squarewave using a
Darlington pair that I am using to replace a single transistor. Since this
is an existing design, and I am trying to save the layout, etc, I don't have
complete freedom to change a lot of things around.

I am using the NPN Darlington as a Common Collector input buffer. Since it
is an integrated component (two transistors in a SOT-23 package) I don't
have the ability to add a resistor to the emitter of the "first" transistor
so as to bleed off charge during the low going edge of a square wave.

What would work best? Would reducing the base current help? In a Darlington
pair, it is already incredibly low.

Thanks
Tom

Reducing base current probably won't help in a reliable way -- you
wouldn't have much control over the base current into the second
transistor; if you could get it right at temperature 'A' it wouldn't
work at all at temperature 'B'.

I assume you're using a Darlington for better current gain. Zetex has
some super-Hfe transistors that may do, and may turn off better.
Otherwise start a search for faster Darlingtons!

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Posting from Google? See http://cfaj.freeshell.org/google/

"Applied Control Theory for Embedded Systems" came out in April.
See details at http://www.wescottdesign.com/actfes/actfes.html

 

[Martin Riddle]

Hi,

I am having trouble with the "low going edge" of a squarewave using a
Darlington pair that I am using to replace a single transistor. Since this
is an existing design, and I am trying to save the layout, etc, I don't
have complete freedom to change a lot of things around.

I am using the NPN Darlington as a Common Collector input buffer. Since
it is an integrated component (two transistors in a SOT-23 package) I
don't have the ability to add a resistor to the emitter of the "first"
transistor so as to bleed off charge during the low going edge of a square
wave.

What would work best? Would reducing the base current help? In a
Darlington pair, it is already incredibly low.

Thanks
Tom

How about a negative bias on the base? Insert a diode in the Emitter if you
can afford it.

Cheers

 

[[email protected]]

Reducing base current probably won't help in a reliable way -- you
wouldn't have much control over the base current into the second
transistor; if you could get it right at temperature 'A' it wouldn't
work at all at temperature 'B'.

I assume you're using a Darlington for better current gain. Zetex has
some super-Hfe transistors that may do, and may turn off better.
Otherwise start a search for faster Darlingtons!

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

I like the Zetex solution. Good stuff.

 

[Paul Probert]

Hi,

I am having trouble with the "low going edge" of a squarewave using a
Darlington pair that I am using to replace a single transistor. Since this
is an existing design, and I am trying to save the layout, etc, I don't have
complete freedom to change a lot of things around.

I am using the NPN Darlington as a Common Collector input buffer. Since it
is an integrated component (two transistors in a SOT-23 package) I don't
have the ability to add a resistor to the emitter of the "first" transistor
so as to bleed off charge during the low going edge of a square wave.

What would work best? Would reducing the base current help? In a Darlington
pair, it is already incredibly low.

Thanks
Tom

Tom,
Back before MOSFETS the way to speed up the turnoff was like this:
|
|
-------->|---------------
| |
| |
| ------
| / |
| | |
| | /
--------->|-----------| |
| | |
| |->--|
Z |
Z |->-
Z |
| |
_______ ______
/ / / / / /
The idea is that when the collector drops to very low voltage the top
diode conducts and takes away the excess base drive, keeping the
transistors from getting too saturated. If need be you can put two
diodes in series for the bottom diode (especially if this is a single
transistor and not a darlington).
Oops, I just noticed you said common collector. No matter, just adapt
the above topology.

Paul Probert
University of Wisconsin

 

[Jim Thompson]

Tom,
Back before MOSFETS the way to speed up the turnoff was like this:
|
|
-------->|---------------
| |
| |
| ------
| / |
| | |
| | /
--------->|-----------| |
| | |
| |->--|
Z |
Z |->-
Z |
| |
_______ ______
/ / / / / /
The idea is that when the collector drops to very low voltage the top
diode conducts and takes away the excess base drive, keeping the
transistors from getting too saturated. If need be you can put two
diodes in series for the bottom diode (especially if this is a single
transistor and not a darlington).
Oops, I just noticed you said common collector. No matter, just adapt
the above topology.

Paul Probert
University of Wisconsin

Common collector can't saturate, *unless* the base can be driven above
the collector.

The problem is that, with no B-E resistor on the output device,
there's no way to clear the output base.

...Jim Thompson

 

[rebel]

Hi,

I am having trouble with the "low going edge" of a squarewave using a
Darlington pair that I am using to replace a single transistor. Since this
is an existing design, and I am trying to save the layout, etc, I don't have
complete freedom to change a lot of things around.

I am using the NPN Darlington as a Common Collector input buffer. Since it
is an integrated component (two transistors in a SOT-23 package) I don't
have the ability to add a resistor to the emitter of the "first" transistor
so as to bleed off charge during the low going edge of a square wave.

What would work best? Would reducing the base current help? In a Darlington
pair, it is already incredibly low.

It would help if we knew why you moved from a single to a Darlington.

 

[Robert Baer]

Hi,

I am having trouble with the "low going edge" of a squarewave using a
Darlington pair that I am using to replace a single transistor. Since this
is an existing design, and I am trying to save the layout, etc, I don't have
complete freedom to change a lot of things around.

I am using the NPN Darlington as a Common Collector input buffer. Since it
is an integrated component (two transistors in a SOT-23 package) I don't
have the ability to add a resistor to the emitter of the "first" transistor
so as to bleed off charge during the low going edge of a square wave.

What would work best? Would reducing the base current help? In a Darlington
pair, it is already incredibly low.

Thanks
Tom

See if it can be replaced with a "biased" part (that is to say, the
resistor is integrated into the darlington).
Better yet, get a high beta non-darlington from Zetek?

 

[BILL]

Teece escreveu:

Hi,

I am having trouble with the "low going edge" of a squarewave using a
Darlington pair that I am using to replace a single transistor. Since this
is an existing design, and I am trying to save the layout, etc, I don't have
complete freedom to change a lot of things around.

I am using the NPN Darlington as a Common Collector input buffer. Since it
is an integrated component (two transistors in a SOT-23 package) I don't
have the ability to add a resistor to the emitter of the "first" transistor
so as to bleed off charge during the low going edge of a square wave.

What would work best? Would reducing the base current help? In a Darlington
pair, it is already incredibly low.

Thanks
Tom

Tom ,

You can place a Capacitor in parallel with the resistor that connects
to the base of the first transistor and adjust experimentally.

Bill

 

[Winfield Hill]

Common collector can't saturate, *unless* the base can be driven
above the collector.

The problem is that, with no B-E resistor on the output device,
there's no way to clear the output base.

Right, and unlike many power Darlington transistors, which include
an aggressive discharge resistor, Teece's poor little sot-23 doesn't
have one. So the solution is to replace it with two discrete parts,
as you suggested earlier. Sometimes it's necessary to provide an
active base pull-down, rather than struggle with a wimpy resistor,
like this:
|
,---|>|----+------+
| | |
| |/ |
--+--/\/\--+-|>|----| |
| |\V |
| | |/
| +----|
| inverter | |\V
| | \ |/ |
'---| >O--/\/\---| gnd
| / |\V
|
gnd

Myself, I prefer an appropriately-sized MOSFET, not too big,
not too small, with an aggressive gate-driver IC.

 

[Teece]

Thank you to everyone who responded to this post.

I have been looking at the Zetex, Diodes Inc, Fairchild websites and have
found many Darlington transistors. However, I have not been able to come
across one with the embedded turn off resistor. Does anyone have a
suggestion for a part number?

Thanks again
Tom

 

[Jim Thompson]

Right, and unlike many power Darlington transistors, which include
an aggressive discharge resistor, Teece's poor little sot-23 doesn't
have one. So the solution is to replace it with two discrete parts,
as you suggested earlier. Sometimes it's necessary to provide an
active base pull-down, rather than struggle with a wimpy resistor,
like this:
|
,---|>|----+------+
| | |
| |/ |
--+--/\/\--+-|>|--X-| |
| |\V |
| | |/
| +----|
| inverter | |\V
| | \ |/ |
'---| >O--/\/\---| gnd
| / |\V
|
gnd

I'd pull down the upper driver base as well... marked with "X"

Myself, I prefer an appropriately-sized MOSFET, not too big,
not too small, with an aggressive gate-driver IC.

Same here. Darlington's are just a wee bit passé.

...Jim Thompson

 

[Chris Jones]

I'd pull down the upper driver base as well... marked with "X"


Same here. Darlington's are just a wee bit passé.

...Jim Thompson

Perhaps one of those Zetex single bipolar transistors would do, they have
very high beta and might replace a darlington in some cases, if having only
a single VBE drop is not a problem. It sounds like the OP doesn't want to
change the PCB layout at all.
http://www.zetex.com/3.0/3-3-2b.asp?rid=1

Chris

 

[Jon Elson]

Hi,

I am having trouble with the "low going edge" of a squarewave using a
Darlington pair that I am using to replace a single transistor. Since this
is an existing design, and I am trying to save the layout, etc, I don't have
complete freedom to change a lot of things around.

I am using the NPN Darlington as a Common Collector input buffer. Since it
is an integrated component (two transistors in a SOT-23 package) I don't
have the ability to add a resistor to the emitter of the "first" transistor
so as to bleed off charge during the low going edge of a square wave.

What would work best? Would reducing the base current help? In a Darlington
pair, it is already incredibly low.

Geez, why are you using such a component if concerned with speed?
See if you can find a pin-compatible N-chan FET with suitable ratings.
Then, you might have to be concerned with how to slow it DOWN, but
that is so much easier to accomplish!

Yes, reduce base current to the minimum, if you saturate the transistor
then it takes longer to turn off.

Jon

 

[Joerg]

I'd pull down the upper driver base as well... marked with "X"




Same here. Darlington's are just a wee bit passé.

Except in low VCC common collector situations where Vgs would eat away
too much headroom. Of course one could include a charge pump for more
gate drive but that all becomes quite eclectic.

 

[Fred Bartoli]

Joerg a écrit :

Except in low VCC common collector situations where Vgs would eat away
too much headroom. Of course one could include a charge pump for more
gate drive but that all becomes quite eclectic.

In low VCC situations, Darlingtons probably have a prohibitive VCEsat.