Maker Pro
Maker Pro

High voltage open collector output w/logic inputs?

J

Joel Kolstad

Jan 1, 1970
0
Anyone know of some small (e.g., SOT-23, SC70, etc.) open collector output
driver that can sink, say, 100mA, can stand off, say, 32V, and is controlled
by a logic level (5V or 3.3V CMOS) input?

The ideal is to take the slightly wimpy outputs from a microcontroller and
directly drive a small 28V relay coil, power up some 15V op-amps, etc... and
I'm looking for a better solution than a handful of FET/BJTs and resistors.

Thanks,
---Joel Kolstad
 
J

Joerg

Jan 1, 1970
0
Hello Joel,
Anyone know of some small (e.g., SOT-23, SC70, etc.) open collector output
driver that can sink, say, 100mA, can stand off, say, 32V, and is controlled
by a logic level (5V or 3.3V CMOS) input?

The ideal is to take the slightly wimpy outputs from a microcontroller and
directly drive a small 28V relay coil, power up some 15V op-amps, etc... and
I'm looking for a better solution than a handful of FET/BJTs and resistors.

Ahem, how many channels did you expect in a SOT23 package? Why not just
use a FET that is guaranteed a desired RDSon at 3.3V or 5V?

Regards, Joerg
 
J

Jim Adney

Jan 1, 1970
0
Anyone know of some small (e.g., SOT-23, SC70, etc.) open collector output
driver that can sink, say, 100mA, can stand off, say, 32V, and is controlled
by a logic level (5V or 3.3V CMOS) input?

The ideal is to take the slightly wimpy outputs from a microcontroller and
directly drive a small 28V relay coil, power up some 15V op-amps, etc... and
I'm looking for a better solution than a handful of FET/BJTs and resistors.

Sounds like a resistor and a transistor.

-
 
S

Spehro Pefhany

Jan 1, 1970
0
Anyone know of some small (e.g., SOT-23, SC70, etc.) open collector output
driver that can sink, say, 100mA, can stand off, say, 32V, and is controlled
by a logic level (5V or 3.3V CMOS) input?

The ideal is to take the slightly wimpy outputs from a microcontroller and
directly drive a small 28V relay coil, power up some 15V op-amps, etc... and
I'm looking for a better solution than a handful of FET/BJTs and resistors.

Thanks,
---Joel Kolstad

"Digital transistor" (BJT with bias resistors built in).


Best regards,
Spehro Pefhany
 
J

Joel Kolstad

Jan 1, 1970
0
Hi Joerg,

Joerg said:
Ahem, how many channels did you expect in a SOT23 package?

1 or 2 (in, e.g., a SOT23-6).
Why not just use a FET that is guaranteed a desired RDSon at 3.3V or 5V?

Well, I'd prefer high side drive (the situation I'm thinking of is to power up
some high-power amps that require 110mA at 28V, and I'd rather have 28V moving
around a little rather than "ground"), and the (P channel enhancement mode
MOSFET) 30V Vds parts I'm using only allow an 8V Vgs. So there's already one
'FET and 2 resistors (to limit Vgs). Then, I end up with an N channel FET
driving the resistor going to the gate of the P channel FET since the
microcontroller can't stand off 28V. Thus: 2 FETs, 2 resistors. At that
point, I'd prefer to have a single part solution -- adds a little reliability,
and -- at least in small quantities -- could actually end up being cheaper
(say, $0.25 vs. $0.50).

Given that you can get complete microcontrollers in SOT23-6 packages these
days, it's surprising sometimes just what _isn't_ available!

I'll look into the BJT packages with built-in bias resistors...

---Joel
 
G

Graham W

Jan 1, 1970
0
Joel said:
Hi Joerg,



1 or 2 (in, e.g., a SOT23-6).


Well, I'd prefer high side drive (the situation I'm thinking of is to
power up some high-power amps that require 110mA at 28V, and I'd
rather have 28V moving around a little rather than "ground"), and the
(P channel enhancement mode MOSFET) 30V Vds parts I'm using only
allow an 8V Vgs. So there's already one 'FET and 2 resistors (to
limit Vgs). Then, I end up with an N channel FET driving the
resistor going to the gate of the P channel FET since the
microcontroller can't stand off 28V. Thus: 2 FETs, 2 resistors. At
that point, I'd prefer to have a single part solution -- adds a
little reliability, and -- at least in small quantities -- could
actually end up being cheaper (say, $0.25 vs. $0.50).

Given that you can get complete microcontrollers in SOT23-6 packages
these days, it's surprising sometimes just what _isn't_ available!

I'll look into the BJT packages with built-in bias resistors...

Isn't the ULN-2003A, 7 channel darlington available in the required
SMD package?
 
T

Tim Williams

Jan 1, 1970
0
Joel Kolstad said:
Well, I'd prefer high side drive (the situation I'm thinking of is to
power up some high-power amps that require 110mA at 28V
^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^

Excuse me, but exactly what scale is your mindset? I'm just confused by
your terms... when I saw your post I was expecting to see at least a hundred
volts! 28V is swatting gnats, even for SOT devices.

I suppose 2.9W is "high power" compared to like... a wristwatch... but
gee...

Tim (whose mindset is in the medium power, i.e., 10kW range)
 
J

Joel Kolstad

Jan 1, 1970
0
Tim Williams said:
Excuse me, but exactly what scale is your mindset?

It's the mindset of "low power" is, say, 6V or less (typically 3.3V or 2.5V)
and many microwatts to a few handful. "High power" is >6V and hundreds of
milliwatts to watts.
I'm just confused by
your terms... when I saw your post I was expecting to see at least a hundred
volts! 28V is swatting gnats, even for SOT devices.

I'd love to buy a bunch of the single gate logic ICs (e.g., NC7SZ38) in higher
voltages (their maximum is ~6V); alas, I've never seen such things.
Tim (whose mindset is in the medium power, i.e., 10kW range)

Yeah, I took a power electronics class in college some years ago where "low
voltage" transistors were the ones rated 600V or less!

---Joel
 
J

Joerg

Jan 1, 1970
0
Hello Joel,
Well, I'd prefer high side drive ...


Ok, didn't know that you wanted high-side. You had said sink, not source...
... (the situation I'm thinking of is to power up
some high-power amps that require 110mA at 28V, and I'd rather have 28V moving
around a little rather than "ground"), and the (P channel enhancement mode
MOSFET) 30V Vds parts I'm using only allow an 8V Vgs. So there's already one
'FET and 2 resistors (to limit Vgs). Then, I end up with an N channel FET
driving the resistor going to the gate of the P channel FET since the
microcontroller can't stand off 28V. Thus: 2 FETs, 2 resistors. At that
point, I'd prefer to have a single part solution -- adds a little reliability,
and -- at least in small quantities -- could actually end up being cheaper
(say, $0.25 vs. $0.50).

$0.50? How 'bout this: BSS123 after your gate, two resistors to rail,
BSS84 up top? Ok, considering your 110mA the BSS84 is a wee bit marginal
here but you should be able to find a p-channel that does 300mA or so
for maybe a penny more.

That scenario would cost around $0.10 in production. If you'd spring for
a fifth part (base resistor) you could use a BJT and shave a little over
a penny off of this. If power efficiency isn't a concern and you also
use a BJT up top you could get the whole chebang done for $0.06 or so.

Given that you can get complete microcontrollers in SOT23-6 packages these
days, it's surprising sometimes just what _isn't_ available!

As Graham suggested, look at the ULN series. IIRC there are some
high-side versions, maybe even in TSSOP.

Regards, Joerg
 
J

Joerg

Jan 1, 1970
0
Hello Joel,
Yeah, I took a power electronics class in college some years ago where "low
voltage" transistors were the ones rated 600V or less!

For Tim it seems he'd only consider parts high power if the road has to
be blocked and they have to be hoisted up and lowered in through the
roof using a massive crane. Anything that doesn't have a crane hook is
wimpy.

Regards, Joerg
 
J

Joerg

Jan 1, 1970
0
Hello Spehro,
"Digital transistor" (BJT with bias resistors built in).

Probably they could have sold a lot more of these if the marketeers
would have coined a catchy name. 'Digisistor', 'Logistor' or something
like that.

I haven't seen them used much, maybe for cost reasons. Imagine somebody
trying to repair a board, thinking it was a normal transistor. Click
..... kablouie.

Regards, Joerg
 
H

Hal Murray

Jan 1, 1970
0
For Tim it seems he'd only consider parts high power if the road has to
be blocked and they have to be hoisted up and lowered in through the
roof using a massive crane. Anything that doesn't have a crane hook is
wimpy.

Road?

Back in the late 60s, a friend was designing SCRs for GE. Big ones.
Their primary customer was another division of GE - the one that
made diesel electric locomotives.

What sort of control electronics are used in modern locomotives?
 
Top