Small SCRs with 1mA or less in holding current?

[Joerg]

Gents,

Does anyone know of a non-boutiqueish (meaning available) SCR that can
keep its juices flowing with less than 1mA hold current? The usual ones
are around 5mA, like this:

http://www.onsemi.com/PowerSolutions/product.do?id=NYC0102B

Ok, I can make one out of transistors but real estate is critical. Needs
to be able to withstand 60V or more when open. Max current is miniscule,
probably 5-10x the hold current. Small package, SOT23 or less.

 

[Joerg]

Come on! You can't get two trannies in there?

Nope. Because it would be two transistors and two resistors.

 

[Spehro Pefhany]

Gents,

Does anyone know of a non-boutiqueish (meaning available) SCR that can
keep its juices flowing with less than 1mA hold current? The usual ones
are around 5mA, like this:

http://www.onsemi.com/PowerSolutions/product.do?id=NYC0102B

Ok, I can make one out of transistors but real estate is critical. Needs
to be able to withstand 60V or more when open. Max current is miniscule,
probably 5-10x the hold current. Small package, SOT23 or less.

That stuff's not so available generally. Even PUTs never caught on in
SMT.

A tiny dual NPN/PNP pair with integrated resistors would be ideal, but
they are not very popular with such a high voltage rating.


Best regards,
Spehro Pefhany

 

[Jamie]

Gents,

Does anyone know of a non-boutiqueish (meaning available) SCR that can
keep its juices flowing with less than 1mA hold current? The usual ones
are around 5mA, like this:

http://www.onsemi.com/PowerSolutions/product.do?id=NYC0102B

Ok, I can make one out of transistors but real estate is critical. Needs
to be able to withstand 60V or more when open. Max current is miniscule,
probably 5-10x the hold current. Small package, SOT23 or less.

Is it possible to simply add an R for extra load to maintain it at
lower currents beyond the burden R ?

Jamie

 

[Spehro Pefhany]

does something like this: http://www.onsemi.com/pub_link/Collateral/EMF18XV6T5-D.PDF

get close?

-Lasse

Or this one, plus a resistor or two.

http://www.onsemi.com/pub_link/Collateral/BC846BPDW1T1-D.PDF

About 4 cents ea. @ 1 reel, pretty reasonable.


Best regards,
Spehro Pefhany

 

[[email protected]]

Two resistors? for what? if you think you need base resistors, there
are "pre-biased" transistors available in the same package.

Yeah, but they're too expensive for Joerg. ;-)

 

[Joerg]

does something like this: http://www.onsemi.com/pub_link/Collateral/EMF18XV6T5-D.PDF

get close?

If I have to I'll do it with transistors. But even if I use a dual pack
it still needs a couple of resistors to make it an SCR. I was hoping
some company had a scaled down SCR. There ought to be enough of a market
for that.

 

[Joerg]

The best PNP/NPN biased pairs seem to be 50V rating:
Rohm IMD10AT108, IMD16AT108 in SC-74-6 package; UMD6NTR in their UM-6
package; EMD6T2R in their EMT-6 package; IMD8AT108, IMD6AT108, IMD1AT108
in SC-74-6 package.

That 50V would be a bit of a white-knuckle ride.

 

[Joerg]

Is it possible to simply add an R for extra load to maintain it at
lower currents beyond the burden R ?

It's higher current one would need, not lower, in order to hold the SCR.
The smallest SCRs I could find still need around 5mA and that causes too
much battery drain in my application.

This is where I envy IC designers. They just sketch it onto their CAD
and it's done. Us board level guys can only have what's on the big
buffet table. But when it comes to big inductors they lose and we win

 

[Joerg]

many dual packs with resistors build in

Yes, although nearly always with resistors in series with the base as
well. It can work but they are usually for logic stuff and cannot take
60V. At least the ones I have seen. I'll have to check out the PBLS
series from NPX some more. The NPN in there is only 50V but the PNP is
60V, and that should be enough.

http://www.nxp.com/documents/data_sheet/PBLS6005D.pdf

However, only the NPN in there comes with a BE resistor.

 

[Joerg]

Yeah, but they're too expensive for Joerg. ;-)

No, the concern is marginal breakdown voltage. Believe it or not, cost
(within resonable limits) is not an issue on this project. Real estate is.

 

[Fred Abse]

Does anyone know of a non-boutiqueish (meaning available) SCR that can
keep its juices flowing with less than 1mA hold current? The usual ones
are around 5mA

C106. Typical 0.19mA @ 25 Celsius?

Max is 3mA, so you'd have to select, which probably means you won't like
it.

Pushing the envelope a bit? I've always used C106 when I needed a
sensitive SCR, but not quite as sensitive as you appear to want.

 

[Joerg]

C106. Typical 0.19mA @ 25 Celsius?

Max is 3mA, so you'd have to select, which probably means you won't like
it.

Nope, don't like it

Must be guaranteed.

Pushing the envelope a bit? I've always used C106 when I needed a
sensitive SCR, but not quite as sensitive as you appear to want.

There are some nice latching apps for SCRs and the only reason I never
used any of those ideas was the hold current. Looks like the same this
time :-(

 

[Fred Bartoli]

Joerg a écrit :

No, the concern is marginal breakdown voltage. Believe it or not, cost
(within resonable limits) is not an issue on this project. Real estate is.

Diodes' SOT363 HBDM60V600W or MMDT5451 and 2 x 0402/0201 resistors ?

 

[Fred Abse]

Nope, don't like it

Must be guaranteed.


There are some nice latching apps for SCRs and the only reason I never
used any of those ideas was the hold current. Looks like the same this
time :-(

There used to be some nice SCS made, that would work, but I bet they're
long defunct.

 

[Jamie]

It's higher current one would need, not lower, in order to hold the SCR.
The smallest SCRs I could find still need around 5mA and that causes too
much battery drain in my application.

The idea was to add R to your load to give a minimum
additive current. This would only been noticed when the SCR is on with
external load.

What is the highest V and I that is to be around this?

Jamie

 

[Spehro Pefhany]

C106. Typical 0.19mA @ 25 Celsius?

Max is 3mA, so you'd have to select, which probably means you won't like
it.

Pushing the envelope a bit? I've always used C106 when I needed a
sensitive SCR, but not quite as sensitive as you appear to want.

I've used C106's back around the dawn of time- are they actually
available in small SMT packages? The GE ones used to come in a
horrible package with wide (0.053" max) stamped leads on 0.1" pitch
eg. C106B1 but that's been replaced with some nicer packages.

Oh, and sensitivity goes hand-in-glove with dv/dt problems so high
volume applications usually don't want too much sensitivity.


Best regards,
Spehro Pefhany

 

[Fred Abse]

Oh, and sensitivity goes hand-in-glove with dv/dt problems

That used to be a good reason for using SCS.

 

[Joerg]

I've used C106's back around the dawn of time- are they actually
available in small SMT packages? The GE ones used to come in a
horrible package with wide (0.053" max) stamped leads on 0.1" pitch
eg. C106B1 but that's been replaced with some nicer packages.

Oh, and sensitivity goes hand-in-glove with dv/dt problems so high
volume applications usually don't want too much sensitivity.

Actually in this case dv/dt is even desired

It's supposed to come on when a voltage rail comes on and then remain
on, unless a uC deliberately "extinguishes" it but briefly shunting a
FET across the SCR.

I just can't have several mA wasting away all day long.

 

[Joerg]

The idea was to add R to your load to give a minimum
additive current. This would only been noticed when the SCR is on with
external load.

That would deplete the battery. There are periods of time where the load
draws very little and I can't have the SCR path burn off power that
whole time.

What is the highest V and I that is to be around this?

V can be up to 60V and I should be <1mA.