Small high temp red led?

I need a small (pref. 0603, max. 0805) red led.
(used as a voltage ref in a dense small circuit).

The tough part is the max working temp: 125°C

Any part number?

 

Isn't that close to the typical max junction temp for any
semiconductor?

D from BC

 

I think the epoxy may degrade at 125C plus whatever the rise is.

 

D from BC a écrit :

Almost (it's more 150°C), but that figures depends a lot on the package
techno. Some do 175, few others 200.

In my case it doesn't dissipate much (600/700uW), and light efficiency
vs time doesn't matter at all.

But nearly all seem to be spec'd at 85°C. I just saw some at vishay that
go up to 100°C.

 

I think you need to rethink your design.

I have been involved with cooker electronics quite a bit and we always
made sure there was cool airflow over any electronics.

 

Ambient or junction?

Not right in memory, but as you stepped out of temp for pocket devices I
would browse old or newer MIL-Speck catalogs.

HTH

Stanislaw

 

"Fred Bartoli"

Automotive.

Heres a nice list of mfgs:
<http://www.maxim-ic.com/appnotes.cfm/appnote_number/828>

Nicha has some 110c parts


Cheers

 

"Fred Bartoli"

If you only need it to survive 125C and not maintain it color, most LED on a
ceramic base will be fine. We use them to 175 and 200C all the time. They
do not maintain their color and all of them turn yellow/amber but they still
work for a while. I've never had one go out but our life times are short
(tens of hours).

 

Marra a écrit :

How do you know that?

Not everybody build cookers.

 

Mook Johnson a écrit :

Yep, I only need it to survive but with unaltered electrical parameters.
That's for some industrial sensors (hence the temp range) and can be
deeply buried into huge machines and can only be accessed at the
occasion of a 5 year maintenance operation.
Some of those sensors can operate 24h a day at high temp.
'Programmed failure' isn't an option.

 

Spehro Pefhany a écrit :

For this project I don't need light. Just the higher voltage drop the
led gives.

But SMT incandescent lamps may be useful for another project I have. I
didn't know that this existed. Thanks.

 

Spehro Pefhany a écrit :

Unfortunately space is very constrained (only 15mm diamter) and full up
to the nose. Voltage refs are, well... bulky, and way too noisy.
Today, we use half a BAV99, but I'd like to have one more volt drop...
with the same 2mV/°C tempco.

If I can't find one, we'll stick to the BAV99 or so diode.

 

You won't find them quite that small, I should think, but there *are*
SMT incandescent lamps that will be very happy at 125°C 24/7.


Best regards,
Spehro Pefhany

 

Sorry, noticed that after I hit 'send'

Could you use the IR LED in an optocoupler? A bit bigger than 0805
though (and lower Vf). Failing that, of course, there are parts such
as the LM4041...


Best regards,
Spehro Pefhany

 

The inside of a car I think can get to 60C. Dash LED's have to survive
that.

Really..an LED in an oven at 125C harms what?

Most solders melts around 180C to 190C so internal connections in the
LED should be ok..
The plastics used in SMD LED's have to handle wavesoldering so that's
probably ok. (Contact LED manufacture for plastic flowing point.)

I did a quick look (.interesting topic) and found
http://www.marktechopto.com/Enginee...EDs Application Notes/what-are-leds-lamps.cfm
Shows a conductive silver paste..
I'm guessing that's a conductive adhesive.
That might be the spoiler.
However..the plastic that was melted over the adhesive has to tolerate
wavesoldering.
So..the conductive adhesive tolerates short term heat.

D from BC

 

What about component life span at high temps?

We all know that large changes in temperature causes expansion and
contraction of materials which can cause cracking especially if one
part of the device cools down quicker. Car cylinder heads are a good
example.
Chip substrates are the same but is made worse by the different
materials on the substrate expanding/contracting at different rates.

 

HP/Agilent/ Avago do an automotive range to 110degC junction temp. As
a reference, I assume that power dissipation will be minimal.

ASMT-RR45-AQ902

http://www.avagotech.com/products/product-detail.jsp?navId=H0,C1,C6469,C7889,P103745


RL

 

Hmmm. What about using the whole BAV99? Giving double voltage drop.
Ok, with 4mV/K.

Heinz

 

On Jun 14, 8:38 am, Fred Bartoli

But any diode (diode-equation-compliant forward biased diode)
will have (at 125C temperature) 0.25 percent per C tempco.
For a 1.5V forward voltage, that's gonna be about 3.7 mV/C

The same result holds for Si diodes in series or GaAs, or whatever.

Stabistors (multiple diodes) like 1N4157 will give you a higher
forward voltage, but aren't likely to be available in surface mount.

Another kind of reference, that hasn't been mentioned yet, is a FET
current source with load resistor. A JFET in a metal can with gate-
source
shorted together is a pretty good high temp device, can give its
2 mA (or so) current into a load resistor for any reference voltage
you want. The tempco, you'll have to determine yourself.

 

can't use a bandgap reference, or zener, instead?