Re: LEDs in parallel

John Larkin

unread,

Jan 28, 2012, 2:58:15 PM1/28/12

to

On Thu, 26 Jan 2012 17:50:24 -0600, John Fields
<jfi...@austininstruments.com> wrote:

>On Wed, 25 Jan 2012 18:44:44 -0800, John Larkin
><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>
>>On Wed, 25 Jan 2012 19:26:26 -0600, John Fields
>><jfi...@austininstruments.com> wrote:
>>
>>>On Tue, 24 Jan 2012 21:11:23 -0800, John Larkin
>>><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>>
>>>>On Tue, 24 Jan 2012 20:17:17 -0500, Tom Biasi <tomb...@optonline.net>
>>>>wrote:
>>>>
>>>>>On Tue, 24 Jan 2012 15:56:48 -0800 (PST), Bill Bowden
>>>>>>
>>>>>>I've seen LED flashlights with 2 white 3 volt LEDs wired directly in
>>>>>>parallel across two AA batteries.
>>>>>>
>>>>>>-Bill
>>>>>
>>>>>They rely heavily on the battery's internal resistance and luck.
>>>>
>>>>LEDs have a current:voltage slope that's not a brick wall.
>>>
>>>---
>>>Pretty close, though, once you get past the knee.
>>
>>
>>It's usually the other way around: exponential at low currents, ohmic
>>at higher currents.
>
>---
>It never really gets ohmic unless you drive the junction hard enough
>to short it,

Nonsense, unless you plan to quibble about the word "really."

> and once you get past the knee -

As noted, diodes don't have a "knee" unless you arbitrarily define
one.

where a relatively large
>voltage change results in a small current change - the slope changes
>so that for a relatively small increase in voltage you get a large
>increase in current.

No, that's backwards. Diodes, and LEDs, have current exponential on
voltage at low currents. At higher currents, the contact and bulk
resistivity start to dominate, and the voltage:current curve gets
nearly linear.

Just look at the curves on real led data sheets. The smaller parts
start to get ohmic at low currents, just a few mA. Bigger junctions
will stay exponential at higher currents, because they have less bulk
resistance.

This is a really tiny junction, so the v/i curve is a straight line at
operating currents:

http://vcclite.com/wp-content/files/VAOL-S8GT4-LED-0805-green.pdf

Bigger parts start to go ohmic at higher currents:

http://www.vishay.com/docs/81345/vlmp232.pdf

http://catalog.osram-os.com/jsp/download.jsp?rootPath=/media/&name=LA_LO_LY_E67F_Pb_free.pdf&docPath=Graphics/00057343_0.pdf&url=/media//_en/Graphics/00057343_0.pdf

Ordinary diodes do this, too. That's why diodes have some current
where their v:i curve has a zero temperature coefficient; the
exponential part has a negative TC but the bulk resistance TC is
positive. For small schottky diodes, that can be in the 10 mA
ballpark, so that can be useful.

--

John Larkin, President Highland Technology Inc
www.highlandtechnology.com jlarkin at highlandtechnology dot com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom timing and laser controllers
Photonics and fiberoptic TTL data links
VME analog, thermocouple, LVDT, synchro, tachometer
Multichannel arbitrary waveform generators

John Fields

unread,

Jan 28, 2012, 7:43:09 PM1/28/12

to

On Sat, 28 Jan 2012 11:58:15 -0800, John Larkin

---
There's no quibbling about the word "really", the quibble is about
your assertion that a diode junction is ohmic at vaguely described
qualitative "higher currents".
---

>> and once you get past the knee -
>
>As noted, diodes don't have a "knee" unless you arbitrarily define
>one.

---
"As noted"???

I don't really think an "arbitrary" definition is necessary, since the
location of the knee has been with us for decades.
---

>where a relatively large
>>voltage change results in a small current change - the slope changes
>>so that for a relatively small increase in voltage you get a large
>>increase in current.
>
>No, that's backwards. Diodes, and LEDs, have current exponential on
>voltage at low currents. At higher currents, the contact and bulk
>resistivity start to dominate, and the voltage:current curve gets
>nearly linear.

---
True enough, but "nearly linear" isn't quite the same as "ohmic", is
it?

And, it's just plain silly talk since it has very little to do with
what we're talking about, which is running LEDs from a voltage source.
Look at the V:I curve for a vanilla silicon diode at from zero volts
to where it lets, say, 1mA through the diode and you'll see that the
voltage across the diode, at that point, will be about 0.7V, mas o
menos.

Now run the voltage up to about 1.4V.

Will the current through the diode stop at 2mA?
---

>Just look at the curves on real led data sheets. The smaller parts
>start to get ohmic at low currents, just a few mA. Bigger junctions
>will stay exponential at higher currents, because they have less bulk
>resistance.
>
>This is a really tiny junction, so the v/i curve is a straight line at
>operating currents:
>
>http://vcclite.com/wp-content/files/VAOL-S8GT4-LED-0805-green.pdf

---
Surely you can't be serious.

If you examine the Forward Current vs Forward Voltage curve with some
care, you'll find that with 10mA through the LED it drops about 1.9
volts, and with 20mA through it drops about 2 volts.

Simply using Ohm's law in both cases - in order to determine the
resistance of the LED in each case - yields for the first case:

E 1.9V
R = --- = ------- = 190 ohms
I 1e-2A

and for the second:

E 2.0V
R = --- = ------- = 100 ohms
I 2e-2A

then, since an ohmic load's resistance must be constant as the current
through it varies, that LED is clearly _not_ an ohmic load.
---

>Bigger parts start to go ohmic at higher currents:
>
>http://www.vishay.com/docs/81345/vlmp232.pdf
>
>http://catalog.osram-os.com/jsp/download.jsp?rootPath=/media/&name=LA_LO_LY_E67F_Pb_free.pdf&docPath=Graphics/00057343_0.pdf&url=/media//_en/Graphics/00057343_0.pdf
>
>
>Ordinary diodes do this, too. That's why diodes have some current
>where their v:i curve has a zero temperature coefficient; the
>exponential part has a negative TC but the bulk resistance TC is
>positive. For small schottky diodes, that can be in the 10 mA
>ballpark, so that can be useful.

---
Grasping at straws?

--
JF

k...@att.bizzzzzzzzzzzz

unread,

Jan 28, 2012, 8:11:51 PM1/28/12

to

On Sat, 28 Jan 2012 18:43:09 -0600, John Fields

John, I told you there are people who believe there is a "knee" in an
exponential function. ;-)

>---
>
>>where a relatively large
>>>voltage change results in a small current change - the slope changes
>>>so that for a relatively small increase in voltage you get a large
>>>increase in current.
>>
>>No, that's backwards. Diodes, and LEDs, have current exponential on
>>voltage at low currents. At higher currents, the contact and bulk
>>resistivity start to dominate, and the voltage:current curve gets
>>nearly linear.
>
>---
>True enough, but "nearly linear" isn't quite the same as "ohmic", is
>it?

What do you think constant dI/DV means?

<...>

Jim Thompson

unread,

Jan 28, 2012, 8:14:25 PM1/28/12

to

On Sat, 28 Jan 2012 18:43:09 -0600, John Fields

John Larkin

unread,

Jan 28, 2012, 8:42:27 PM1/28/12

to

And there are people who think that the only kind of resistance is
E/I.

Some famous person once said "When all you know is Ohm's Law,
everything looks like a resistor."

Well, that was me, actually.

John Larkin

unread,

Jan 28, 2012, 8:51:38 PM1/28/12

to

On Sat, 28 Jan 2012 18:43:09 -0600, John Fields

Lots of consumer products do exactly that.

>Look at the V:I curve for a vanilla silicon diode at from zero volts
>to where it lets, say, 1mA through the diode and you'll see that the
>voltage across the diode, at that point, will be about 0.7V, mas o
>menos.

0.7 volts at 1 mA? Where can I get some of those cool silicon LEDs?

What color are they?

John

John Larkin

unread,

Jan 28, 2012, 9:31:23 PM1/28/12

to

If you have anything intelligent to say on topic, give it a try.

Jim Yanik

unread,

Jan 29, 2012, 11:43:10 AM1/29/12

to

John Fields <jfi...@austininstruments.com> wrote in
news:v4l8i79lka6tno3o5...@4ax.com:

the Harbor Freight giveaway LED flashlights have NINE LEDs in parallel,no
series resistor,and three AAA cells to power it. I have several of them.

--
Jim Yanik
jyanik
at
localnet
dot com

John Fields

unread,

Jan 29, 2012, 11:44:58 AM1/29/12

to

On Sat, 28 Jan 2012 17:51:38 -0800, John Larkin

---
Sure, and if they're designed properly, the resistance of the LED and
the internal resistance of the battery will limit the current through
the LED to safe levels.

If not, LED life will be reduced; sometimes drastically.
---

>>Look at the V:I curve for a vanilla silicon diode at from zero volts
>>to where it lets, say, 1mA through the diode and you'll see that the
>>voltage across the diode, at that point, will be about 0.7V, mas o
>>menos.
>>

>>Now run the voltage up to about 1.4V.
>>Will the current through the diode stop at 2mA?
>
>

>0.7 volts at 1 mA? Where can I get some of those cool silicon LEDs?
>What color are they?

---
More silliness, since since when is a "vanilla silicon diode" an LED?

I've got a vanilla 1N4148 on the bench with 1.0000 mA through it and
0.618 V across it, so what do you think will happen if I double the
voltage across it?

--
JF

John Larkin

unread,

Jan 29, 2012, 12:47:03 PM1/29/12

to

On Sun, 29 Jan 2012 10:43:10 -0600, Jim Yanik <jya...@abuse.gov>
wrote:

It would be more efficient (uniform brightness vs battery voltage,
longer battery life) if it used a switcher, but it's cheap, so it
doesn't.

If the LEDs are purchased in quantity, all from the same wafers, they
will share current pretty well without added resistance. Thermal
runaway is not an issue once they get into their ohmic ranges, which
most LEDs do at their design operating current.

The idea that you can't parallel diodes is silly. Motorola sold high
current paralleled-diode rectifier arrays decades ago.

My Name Is Tzu How Do You Do

unread,

Jan 29, 2012, 12:59:03 PM1/29/12

to

Yes, but they were thermally bonded together or the spec for their use
in parallel required thermal equality in operation.

John Larkin

unread,

Jan 29, 2012, 1:18:28 PM1/29/12

to

On Sun, 29 Jan 2012 10:44:58 -0600, John Fields

Something like 350 mA. But there's a lot of variation between
different manufacturers' parts.

If it's on a PC board, with a bit of pad+trace to heatsink the leads,
it should survive.

At that voltage, it will be well into its ohmic range, namely the
current:voltage line will be straight, not exponential. That should be
close to the zero TC point too.

Without the series resistance component, pure exponential, increasing
the diode voltage from 0.6 to 1.2 would increase the current by a
factor of about 10^10.

If you paralleled a bunch of 1N4148s from the same reel, and ran them
at, say, 100 mA or so, I'd expect pretty good current sharing.

John Larkin

unread,

Jan 29, 2012, 1:21:48 PM1/29/12

to

The ones that I'm thinking of were soldered to a common heat sink
plate. But they needn't be. At high currents, diodes can have zero TCs
and, in effect, have internal ballasting resistors.

My Name Is Tzu How Do You Do

unread,

Jan 29, 2012, 1:53:39 PM1/29/12

to

On Sun, 29 Jan 2012 10:21:48 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>The ones that I'm thinking of were soldered to a common heat sink
>plate.

NO SHIT, SHERLOCK!

John Fields

unread,

Jan 29, 2012, 1:56:21 PM1/29/12

to

On Sun, 29 Jan 2012 09:47:03 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>On Sun, 29 Jan 2012 10:43:10 -0600, Jim Yanik <jya...@abuse.gov>
>wrote:

>>the Harbor Freight giveaway LED flashlights have NINE LEDs in parallel,no
>>series resistor,and three AAA cells to power it. I have several of them.
>
>It would be more efficient (uniform brightness vs battery voltage,
>longer battery life) if it used a switcher, but it's cheap, so it
>doesn't.
>
>If the LEDs are purchased in quantity, all from the same wafers, they
>will share current pretty well without added resistance.

---
"Pretty well"?

Is that a new technical term?
---

>Thermal
>runaway is not an issue once they get into their ohmic ranges, which
>most LEDs do at their design operating current.
>
>The idea that you can't parallel diodes is silly. Motorola sold high
>current paralleled-diode rectifier arrays decades ago.

---
Apples and oranges, plus they took great pains to match the forward
voltages of the individual rectifiers in the array and to keep them
isothermal.

From Motorola's 1980 Silicon Rectifier Data Manual:

news:r94bi79k5pqkbc58l...@4ax.com

--
JF

John Larkin

unread,

Jan 29, 2012, 2:26:41 PM1/29/12

to

On Sun, 29 Jan 2012 12:56:21 -0600, John Fields
<jfi...@austininstruments.com> wrote:

>On Sun, 29 Jan 2012 09:47:03 -0800, John Larkin
><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>
>>On Sun, 29 Jan 2012 10:43:10 -0600, Jim Yanik <jya...@abuse.gov>
>>wrote:
>
>>>the Harbor Freight giveaway LED flashlights have NINE LEDs in parallel,no
>>>series resistor,and three AAA cells to power it. I have several of them.
>>
>>It would be more efficient (uniform brightness vs battery voltage,
>>longer battery life) if it used a switcher, but it's cheap, so it
>>doesn't.
>>
>>If the LEDs are purchased in quantity, all from the same wafers, they
>>will share current pretty well without added resistance.
>
>---
>"Pretty well"?
>
>Is that a new technical term?

It means "well enough to be useful."

You might try being useful yourself, instead of just whining all the
time.

My Name Is Tzu How Do You Do

unread,

Jan 29, 2012, 2:33:46 PM1/29/12

to

On Sun, 29 Jan 2012 11:26:41 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>You might try being useful yourself, instead of just whining all the
>time.
>

He is. He told you Motorola performed the matching and culling at the
time of manufacture. It has nothing to do with "coming off the same
die", or anything else. It has to do with REAL performance AFTER
manufacture. And they culled together matched sets based on REAL
performance numbers.

You are apparently in the dark a bit more than you think you are.

John Larkin

unread,

Jan 29, 2012, 2:44:28 PM1/29/12

to

On Sun, 29 Jan 2012 11:33:46 -0800, My Name Is Tzu How Do You Do
<T...@hereforlongtime.org> wrote:

>On Sun, 29 Jan 2012 11:26:41 -0800, John Larkin
><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>
>>You might try being useful yourself, instead of just whining all the
>>time.
>>
>
> He is. He told you Motorola performed the matching and culling at the
>time of manufacture.

Hardly insider information. When else would they possibly do it?

It has nothing to do with "coming off the same
>die", or anything else.

It sure does. Parts off the same die tend to be very similar. More so
nowadays, since silicon processing has improved a lot in the last few
decades.

I find that small parts, like LEDs, tend to be very repeatable,
especially ones off the same reel. We once made a deal with Mouser to
sample us ten parts each off a bunch of different reels of rectifiers,
with the understanding that they would test the samples and they would
then sell us the reels that we liked, for Grehkov (drift step-recovery
diode) use. That worked great. Parts from different reels were very
different, but we fould one reel that had all good snappers.

It has to do with REAL performance AFTER
>manufacture. And they culled together matched sets based on REAL
>performance numbers.

Duh.

>
> You are apparently in the dark a bit more than you think you are.

Well, I do this for a living, and build stuff that works.

John Larkin

unread,

Jan 29, 2012, 2:46:00 PM1/29/12

to

On Sun, 29 Jan 2012 11:44:28 -0800, John Larkin

<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>On Sun, 29 Jan 2012 11:33:46 -0800, My Name Is Tzu How Do You Do
><T...@hereforlongtime.org> wrote:
>
>>On Sun, 29 Jan 2012 11:26:41 -0800, John Larkin
>><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>
>>>You might try being useful yourself, instead of just whining all the
>>>time.
>>>
>>
>> He is. He told you Motorola performed the matching and culling at the
>>time of manufacture.
>
>Hardly insider information. When else would they possibly do it?
>
>
> It has nothing to do with "coming off the same
>>die", or anything else.
>
>It sure does. Parts off the same die tend to be very similar. More so
>nowadays, since silicon processing has improved a lot in the last few
>decades.
>
>I find that small parts, like LEDs, tend to be very repeatable,
>especially ones off the same reel. We once made a deal with Mouser to
>sample us ten parts each off a bunch of different reels of rectifiers,
>with the understanding that they would test the samples and they would

no, ^^^^ we did the testing

John Fields

unread,

Jan 29, 2012, 2:53:22 PM1/29/12

to

On Sun, 29 Jan 2012 10:18:28 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>On Sun, 29 Jan 2012 10:44:58 -0600, John Fields
><jfi...@austininstruments.com> wrote:
>
>>
>>I've got a vanilla 1N4148 on the bench with 1.0000 mA through it and
>>0.618 V across it, so what do you think will happen if I double the
>>voltage across it?
>
>Something like 350 mA. But there's a lot of variation between
>different manufacturers' parts.

---
Snipped irrelevant pontificating.

--
JF

Fred Bloggs

unread,

Jan 29, 2012, 2:47:39 PM1/29/12

to

On Jan 29, 12:47 pm, John Larkin

<jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
> On Sun, 29 Jan 2012 10:43:10 -0600, Jim Yanik <jya...@abuse.gov>
> wrote:
>
>
>
>
>

> >John Fields <jfie...@austininstruments.com> wrote in

There's a reason why those cheap white LED flashlights are made by
battery manufacturers, but who can complain about the price, they're
almost free. There's also a reason why the cheapies are all powered by
AAAs and button cell batteries and that is the relatively high battery
internal resistances. The white LED luminous output versus drive
current is practically stationary at its peak which is very convenient
for design purposes, and design point, if your battery impedance is
fairly well characterized. A variation of +/-15 % does not make much
of visible difference. They also don't seem all that concerned about
driving the LED with up to 4x maximum continuous IF rating, that
current does not last long enough to damage anything. The story
changes if you move to a more serious power source, the current is
almost always regulated.
Please view in a fixed-width font such as Courier.

.
.
.
.
. From load line:
.
.
.
.
. V - V
. BATT LED
. I = ------------
. LED R + R
. BATT LED
.
.
.
. from which:
.
.
.
. dI d(R + R ) dR
. LED BATT LED BATT
. ----- = (-) -------------- ~ (-) ------
. I R + R R
. LED BATT LED BATT
.
.
.
. single digit percentage sensitivity
.

John Fields

unread,

Jan 29, 2012, 3:12:00 PM1/29/12

to

On Sun, 29 Jan 2012 11:26:41 -0800, John Larkin

<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>On Sun, 29 Jan 2012 12:56:21 -0600, John Fields
><jfi...@austininstruments.com> wrote:
>
>>On Sun, 29 Jan 2012 09:47:03 -0800, John Larkin
>><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>
>>>On Sun, 29 Jan 2012 10:43:10 -0600, Jim Yanik <jya...@abuse.gov>
>>>wrote:
>>
>>>>the Harbor Freight giveaway LED flashlights have NINE LEDs in parallel,no
>>>>series resistor,and three AAA cells to power it. I have several of them.
>>>
>>>It would be more efficient (uniform brightness vs battery voltage,
>>>longer battery life) if it used a switcher, but it's cheap, so it
>>>doesn't.
>>>
>>>If the LEDs are purchased in quantity, all from the same wafers, they
>>>will share current pretty well without added resistance.
>>
>>---
>>"Pretty well"?
>>
>>Is that a new technical term?
>
>It means "well enough to be useful."
>
>You might try being useful yourself, instead of just whining all the
>time.

---
As usual, you try to trivialize criticism by belittling it.

As far as being useful goes, you really ought to take a look at all
the crap you post before you make disparaging remarks.

--
JF

John Larkin

unread,

Jan 29, 2012, 3:18:40 PM1/29/12

to

On Sun, 29 Jan 2012 13:53:22 -0600, John Fields

<jfi...@austininstruments.com> wrote:

>On Sun, 29 Jan 2012 10:18:28 -0800, John Larkin
><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>
>>On Sun, 29 Jan 2012 10:44:58 -0600, John Fields
>><jfi...@austininstruments.com> wrote:
>>
>>>
>>>I've got a vanilla 1N4148 on the bench with 1.0000 mA through it and
>>>0.618 V across it, so what do you think will happen if I double the
>>>voltage across it?
>>
>>Something like 350 mA. But there's a lot of variation between
>>different manufacturers' parts.

What did you measure?

>
>---
>Snipped irrelevant pontificating.

Electronics is irrelevant to you?

The LT Spice model for the 1N4148 has an Rs value of 0.568 ohms, which
looks a tad low to me.

John Larkin

unread,

Jan 29, 2012, 3:37:30 PM1/29/12

to

On Sun, 29 Jan 2012 14:12:00 -0600, John Fields

What current did you measure on the 1N4148?

My Name Is Tzu How Do You Do

unread,

Jan 29, 2012, 4:18:13 PM1/29/12

to

On Sun, 29 Jan 2012 11:44:28 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>
>Well, I do this for a living, and build stuff that works.

Then why take a stupid stance that they do not need to be thermally
coupled?

John Fields

unread,

Jan 29, 2012, 4:33:40 PM1/29/12

to

On Sun, 29 Jan 2012 12:18:40 -0800, John Larkin

<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>On Sun, 29 Jan 2012 13:53:22 -0600, John Fields
><jfi...@austininstruments.com> wrote:
>
>>On Sun, 29 Jan 2012 10:18:28 -0800, John Larkin
>><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>
>>>On Sun, 29 Jan 2012 10:44:58 -0600, John Fields
>>><jfi...@austininstruments.com> wrote:
>>>
>>>>
>>>>I've got a vanilla 1N4148 on the bench with 1.0000 mA through it and
>>>>0.618 V across it, so what do you think will happen if I double the
>>>>voltage across it?
>>>
>>>Something like 350 mA. But there's a lot of variation between
>>>different manufacturers' parts.
>
>What did you measure?

---

712mA.
---

>>---
>>Snipped irrelevant pontificating.
>
>Electronics is irrelevant to you?

---
No, irrelevant pontificating about it is.
---

>The LT Spice model for the 1N4148 has an Rs value of 0.568 ohms, which
>looks a tad low to me.

---
Assuming 750mA puts it into the linear region, then Rs = Vf/If ~ 1.7
ohms, so I agree.

How's that for a first? ;)

--
JF

John Fields

unread,

Jan 29, 2012, 4:35:34 PM1/29/12

to

On Sun, 29 Jan 2012 12:37:30 -0800, John Larkin

---
712mA.

--
JF

John Larkin

unread,

Jan 29, 2012, 5:15:53 PM1/29/12

to

I explained that. At low currents, diodes have negative TCs and can go
into thermal runaway if paralleled (and, obviously, poorly heat sunk.)
At high currents, they become ohmic and have more positive TCs, so
there's no thermal runaway mechanism.

That's why people sell flashlights that have multiple LEDs and no
resistors. Because it works.

And multiple diodes soldered to a common heatsunk baseplate really
aren't "thermally coupled" very well... not much better than if they
were separate diodes on the same heat sink.

Whether diodes can be usefully paralleled is an engineering issue that
has to be evaluated for specific cases. But a hard rule to never
parallel diodes, because you heard it somewhere, doesn't make sense.

Thermal runaway is a bigger issue for transistors.

John Larkin

unread,

Jan 29, 2012, 5:27:45 PM1/29/12

to

On Sun, 29 Jan 2012 15:33:40 -0600, John Fields

That's the static resistance. A more useful value is dV/dI, the slope
of the curve in the linear region. That's the Spice 0.568 value. The
value from several 1N4148 data sheets is more like 0.7 ohms.

>
>How's that for a first? ;)

LT Spice shows 0.64 amps at 1.236 volts. Most diode data sheets show
about half that current.

John Fields

unread,

Jan 29, 2012, 7:10:41 PM1/29/12

to

On Sun, 29 Jan 2012 14:15:53 -0800, John Larkin

<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>On Sun, 29 Jan 2012 13:18:13 -0800, My Name Is Tzu How Do You Do
><T...@hereforlongtime.org> wrote:
>
>>On Sun, 29 Jan 2012 11:44:28 -0800, John Larkin
>><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>
>>>
>>>Well, I do this for a living, and build stuff that works.

---
Yeah, but it's mostly paint by numbers.
---

>> Then why take a stupid stance that they do not need to be thermally
>>coupled?
>
>I explained that. At low currents, diodes have negative TCs and can go
>into thermal runaway if paralleled (and, obviously, poorly heat sunk.)

---
If that's true, and the array is driven by a voltage source, thermal
runaway at low current is kind of an oxymoron if increasing the
current through the diode tends to heat it up and drive its tempco,
and its resistance, more positive.

Such being the case, the diode's parallel neighbors will respond
independently and would be harmed by having to conform to a common
thermal base.
---

>That's why people sell flashlights that have multiple LEDs and no
>resistors. Because it works.

---
But you haven't really explained _why_ it works, you merely
pontificate.
---

>And multiple diodes soldered to a common heatsunk baseplate really
>aren't "thermally coupled" very well... not much better than if they
>were separate diodes on the same heat sink.

---
Finally, you've lost your mind.
---

>Whether diodes can be usefully paralleled is an engineering issue that
>has to be evaluated for specific cases. But a hard rule to never
>parallel diodes, because you heard it somewhere, doesn't make sense.

---
Thanks for the clue.
---

>Thermal runaway is a bigger issue for transistors.

---
That issue was resolved long ago, don't you know?

--
JF

John Fields

unread,

Feb 2, 2012, 8:38:06 PM2/2/12

to

On Sun, 29 Jan 2012 14:27:45 -0800, John Larkin

---
Since there will always be a disparity between minority and majority
carriers in the mix, and charge will be shared between them, there can
never be a linear region.

Except, perhaps, for the fleeting moment when the TC across the
junction equals zero.
---

>That's the Spice 0.568 value. The
>value from several 1N4148 data sheets is more like 0.7 ohms.

---
Maximum, at If(nom) or, perhaps, at If(max).

Do you have some real data to support your position or are you just
looking to blame LTspice for your failings?
---

>>How's that for a first? ;)
>
>LT Spice shows 0.64 amps at 1.236 volts. Most diode data sheets show
>about half that current.

---
Most 1N4148 data sheets show the drop across the diode with up to a
couple of hundred milliamps through it, so what are you talking about?
--
JF

John Larkin

unread,

Feb 2, 2012, 10:49:41 PM2/2/12

to

On Thu, 02 Feb 2012 19:38:06 -0600, John Fields

You don't actually care about diodes, you just want something to whine
about.

AllInTheChi

unread,

Feb 2, 2012, 11:16:11 PM2/2/12

to

On Thu, 02 Feb 2012 19:49:41 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>
>You don't actually care about diodes, you just want something to whine
>about.

And you guys call *me* a troll?

I have PCBs with 30 LEDs and limit resistors on it. Powered by a 12V
6W dongle.

I posted a hi res pic of them on flickr.

josephkk

unread,

Feb 2, 2012, 11:28:15 PM2/2/12

to

On Sat, 28 Jan 2012 17:42:27 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>
>>>I don't really think an "arbitrary" definition is necessary, since the
>>>location of the knee has been with us for decades.
>>

>>John, I told you there are people who believe there is a "knee" in an
>>exponential function. ;-)
>
>And there are people who think that the only kind of resistance is
>E/I.
>
>Some famous person once said "When all you know is Ohm's Law,
>everything looks like a resistor."
>
>Well, that was me, actually.
>

Switch famous to infamous and the description of yourself is all too apt.

?-)

John Larkin

unread,

Feb 3, 2012, 12:09:10 AM2/3/12

to

After almost a week of brooding, that's all you can come up with?

John Larkin

unread,

Feb 3, 2012, 12:10:07 AM2/3/12

to

You connected LEDs and resistors?

Jim Thompson

unread,

Feb 3, 2012, 7:24:52 AM2/3/12

to

On Thu, 02 Feb 2012 20:28:15 -0800, josephkk <joseph_...@sbcglobal.net>
wrote:

AllInTheChi

unread,

Feb 3, 2012, 8:27:55 AM2/3/12

to

On Thu, 02 Feb 2012 21:10:07 -0800, John Larkin

<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>On Thu, 02 Feb 2012 20:16:11 -0800, AllInTheChi
><Hos...@magicregion.org> wrote:
>
>>On Thu, 02 Feb 2012 19:49:41 -0800, John Larkin
>><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>
>>>
>>>You don't actually care about diodes, you just want something to whine
>>>about.
>>
>> And you guys call *me* a troll?
>>
>> I have PCBs with 30 LEDs and limit resistors on it. Powered by a 12V
>>6W dongle.
>>
>> I posted a hi res pic of them on flickr.
>
>You connected LEDs and resistors?

36 LEDs and 12 resistors.

12 sets at one dongle per set of two sticks. That's only 3 Watts each
stick. Unless the dongles lie or are being overtaxed.

I am quite sure I could put more than two sticks on a single dongle,
but I didn't want to sit down and configure an LVDC distribution panel.
They work fine in pairs, and the power strip is a 'panel' for the
dongles. Same juice ends up getting used either way, though there are
likely a bit more losses on the many dongles method. Making fewer dongle
fight harder is probably a slight bit more efficient. (splitting it on
the DC side).

John Larkin

unread,

Feb 3, 2012, 1:39:57 PM2/3/12

to

On Fri, 03 Feb 2012 05:24:52 -0700, Jim Thompson
<To-Email-Use-Th...@On-My-Web-Site.com> wrote:

>On Thu, 02 Feb 2012 20:28:15 -0800, josephkk <joseph_...@sbcglobal.net>
>wrote:
>
>>On Sat, 28 Jan 2012 17:42:27 -0800, John Larkin
>><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>
>>>
>>>>>I don't really think an "arbitrary" definition is necessary, since the
>>>>>location of the knee has been with us for decades.
>>>>
>>>>John, I told you there are people who believe there is a "knee" in an
>>>>exponential function. ;-)
>>>
>>>And there are people who think that the only kind of resistance is
>>>E/I.
>>>
>>>Some famous person once said "When all you know is Ohm's Law,
>>>everything looks like a resistor."
>>>
>>>Well, that was me, actually.
>>>
>>
>>Switch famous to infamous and the description of yourself is all too apt.
>>
>>?-)
>
>Arguing with Larkin is like arguing with a liberal, they're all mouth and no
>math.
>
> ...Jim Thompson

How come you never post anything with content?

**********************************

John Larkin, President
Highland Technology, Inc

jlarkin at highlandtechnology dot com

http://www.highlandtechnology.com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators

Custom laser controllers

Photonics and fiberoptic TTL data links

VME thermocouple, LVDT, synchro acquisition and simulation

Jim Thompson

unread,

Feb 3, 2012, 1:50:09 PM2/3/12

to

On Fri, 03 Feb 2012 10:39:57 -0800, John Larkin
<jla...@highlandtechnology.com> wrote:

>On Fri, 03 Feb 2012 05:24:52 -0700, Jim Thompson
><To-Email-Use-Th...@On-My-Web-Site.com> wrote:
>
>>On Thu, 02 Feb 2012 20:28:15 -0800, josephkk <joseph_...@sbcglobal.net>
>>wrote:
>>
>>>On Sat, 28 Jan 2012 17:42:27 -0800, John Larkin
>>><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>>
>>>>
>>>>>>I don't really think an "arbitrary" definition is necessary, since the
>>>>>>location of the knee has been with us for decades.
>>>>>
>>>>>John, I told you there are people who believe there is a "knee" in an
>>>>>exponential function. ;-)
>>>>
>>>>And there are people who think that the only kind of resistance is
>>>>E/I.
>>>>
>>>>Some famous person once said "When all you know is Ohm's Law,
>>>>everything looks like a resistor."
>>>>
>>>>Well, that was me, actually.
>>>>
>>>
>>>Switch famous to infamous and the description of yourself is all too apt.
>>>
>>>?-)
>>
>>Arguing with Larkin is like arguing with a liberal, they're all mouth and no
>>math.
>>
>> ...Jim Thompson
>
>
>How come you never post anything with content?
>

I knew you were going to say that ;-) However, when I _do_ publish technical
content, it is complete, and _works_ >:-}

Didn't anyone ever teach you about separating the message body from your
signature with dash-dash-space ??

See... http://linux.sgms-centre.com/misc/netiquette.php

>
>**********************************
>
>John Larkin, President
>Highland Technology, Inc
>
>jlarkin at highlandtechnology dot com
>http://www.highlandtechnology.com
>
>Precision electronic instrumentation
>Picosecond-resolution Digital Delay and Pulse generators
>Custom laser controllers
>Photonics and fiberoptic TTL data links
>VME thermocouple, LVDT, synchro acquisition and simulation

John Fields

unread,

Feb 3, 2012, 6:58:39 PM2/3/12

to

On Fri, 03 Feb 2012 10:39:57 -0800, John Larkin
<jla...@highlandtechnology.com> wrote:

>On Fri, 03 Feb 2012 05:24:52 -0700, Jim Thompson
><To-Email-Use-Th...@On-My-Web-Site.com> wrote:
>
>>On Thu, 02 Feb 2012 20:28:15 -0800, josephkk <joseph_...@sbcglobal.net>
>>wrote:
>>
>>>On Sat, 28 Jan 2012 17:42:27 -0800, John Larkin
>>><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>>
>>>>
>>>>>>I don't really think an "arbitrary" definition is necessary, since the
>>>>>>location of the knee has been with us for decades.
>>>>>
>>>>>John, I told you there are people who believe there is a "knee" in an
>>>>>exponential function. ;-)
>>>>
>>>>And there are people who think that the only kind of resistance is
>>>>E/I.
>>>>
>>>>Some famous person once said "When all you know is Ohm's Law,
>>>>everything looks like a resistor."
>>>>
>>>>Well, that was me, actually.
>>>>
>>>
>>>Switch famous to infamous and the description of yourself is all too apt.
>>>
>>>?-)
>>
>>Arguing with Larkin is like arguing with a liberal, they're all mouth and no
>>math.
>>
>> ...Jim Thompson
>
>
>How come you never post anything with content?

---
He often does, whether or not you're capable of understanding the
content.

--
JF

John Larkin

unread,

Feb 3, 2012, 7:09:58 PM2/3/12

to

Well, that was a great technical contribution too.

John Fields

unread,

Feb 3, 2012, 7:59:44 PM2/3/12

to

On Thu, 02 Feb 2012 19:49:41 -0800, John Larkin

---
Nonsense.

Whenever I checkmate you, and you realize you've lost, you sidestep,
pretending you have moves left which you don't, in order to prolong
the knell of your death bell and hope for a miracle to save you.

--
JF

John Fields

unread,

Feb 3, 2012, 8:23:51 PM2/3/12

to

On Fri, 03 Feb 2012 16:09:58 -0800, John Larkin

---
Sorry you misunderstood, but that was a non-technical comment directed
at your reticence to admit to error.

--
JF

John Larkin

unread,

Feb 3, 2012, 8:32:24 PM2/3/12

to

On Fri, 03 Feb 2012 18:59:44 -0600, John Fields

Checkmate? Save me? How pathetic.

You are raving again, in really corny stilted prose, and it's not
about diodes.

To 500 mA:

http://www.nxp.com/documents/data_sheet/1N4148_1N4448.pdf

To 800 mA:

http://www.fairchildsemi.com/ds/1N/1N4148.pdf

~~800 mA:

http://www.vishay.com/docs/81857/1n4148.pdf

which you could have found yourself in a couple of minutes if you were
interested in diodes. But you'd rather whine and strut and cluck.

**********************************

John Larkin, President
Highland Technology, Inc

jlarkin at highlandtechnology dot com

http://www.highlandtechnology.com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators

Custom laser controllers

Photonics and fiberoptic TTL data links

DarkMatter

unread,

Feb 3, 2012, 8:40:11 PM2/3/12

to

On Fri, 03 Feb 2012 10:39:57 -0800, John Larkin
<jla...@highlandtechnology.com> wrote:

>

>How come you never post anything with content?

Is that what you call the crap you post?

Numer0 Un0

unread,

Feb 3, 2012, 8:44:36 PM2/3/12

to

On Fri, 03 Feb 2012 16:09:58 -0800, John Larkin
<jla...@highlandtechnology.com> wrote:

>Well, that was a great technical contribution too.

Hey, idiot. YOU are UNqualified to make a valid assessment about
anyone else.

You got that, boy?

k...@att.bizzzzzzzzzzzz

unread,

Feb 3, 2012, 8:51:20 PM2/3/12

to

On Thu, 02 Feb 2012 21:09:10 -0800, John Larkin

<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>On Thu, 02 Feb 2012 20:28:15 -0800, josephkk
><joseph_...@sbcglobal.net> wrote:
>
>>On Sat, 28 Jan 2012 17:42:27 -0800, John Larkin
>><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>
>>>
>>>>>I don't really think an "arbitrary" definition is necessary, since the
>>>>>location of the knee has been with us for decades.
>>>>
>>>>John, I told you there are people who believe there is a "knee" in an
>>>>exponential function. ;-)
>>>
>>>And there are people who think that the only kind of resistance is
>>>E/I.
>>>
>>>Some famous person once said "When all you know is Ohm's Law,
>>>everything looks like a resistor."
>>>
>>>Well, that was me, actually.
>>>
>>
>>Switch famous to infamous and the description of yourself is all too apt.
>>
>>?-)
>
>After almost a week of brooding, that's all you can come up with?

Week? He's been on the rag for years.

John Larkin

unread,

Feb 3, 2012, 8:59:27 PM2/3/12

to

Your new nym: NumeroZero.

--

Hattori Hanzo

unread,

Feb 3, 2012, 9:06:55 PM2/3/12

to

On Fri, 03 Feb 2012 17:32:24 -0800, John Larkin
<jla...@highlandtechnology.com> wrote:

>Checkmate? Save me? How pathetic.

No different than you with "word salad", you fucking total retard, AND
total PATHETIC hypocrite!

In fact, his usage is quite common. Just not so much among brainless
bloodlines in their little rat races. You are a member of one such
bloodline.

John Larkin

unread,

Feb 3, 2012, 9:12:17 PM2/3/12

to

On Fri, 03 Feb 2012 18:06:55 -0800, Hattori Hanzo
<Outint...@billsbackyard.org> wrote:

>On Fri, 03 Feb 2012 17:32:24 -0800, John Larkin
><jla...@highlandtechnology.com> wrote:
>
>>Checkmate? Save me? How pathetic.
>
> No different than you with "word salad", you fucking total retard, AND
>total PATHETIC hypocrite!
>
> In fact, his usage is quite common.

Way, way too common.

--

Hattori Hanzo

unread,

Feb 3, 2012, 9:43:42 PM2/3/12

to

On Fri, 03 Feb 2012 18:12:17 -0800, John Larkin
<jla...@highlandtechnology.com> wrote:

>Way, way too common.

You are way too uncommonly stupid for the position you claim to be in.

John Larkin

unread,

Feb 3, 2012, 9:50:34 PM2/3/12

to

On Fri, 03 Feb 2012 11:50:09 -0700, Jim Thompson

The alternator switcher thing that you posted was neither.

--

John Larkin, President Highland Technology Inc
www.highlandtechnology.com jlarkin at highlandtechnology dot com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators

Custom timing and laser controllers

Photonics and fiberoptic TTL data links

John Larkin

unread,

Feb 3, 2012, 10:17:51 PM2/3/12

to

I was talking about diode series resistance, part of the issue of
paralleling LEDs, and JF got into his usual clucking and whining mode.

Do you own any LED flashlights? Open one up and see if there are any
resistors.

--

John Larkin, President Highland Technology Inc
www.highlandtechnology.com jlarkin at highlandtechnology dot com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators

Custom timing and laser controllers

Photonics and fiberoptic TTL data links

Hattori Hanzo

unread,

Feb 4, 2012, 6:23:03 AM2/4/12

to

On Fri, 03 Feb 2012 19:17:51 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>
>Do you own any LED flashlights? Open one up and see if there are any
>resistors.
>

Mine are not the dumb versions. The switch, and control circuit is in
the tail. That means that it takes the 4.5 volts in and correctly
controls the feed to a single, super-bright LED with 3 distinct modes.

It is so well designed that when your batteries begin to avalanche,
the second mode no longer functions precisely as designed, signaling
the operator to use the first mode or third mode, but to also prepare to
replace or recharge your cells. The 'nose' end of the battery pack gets
applied directly to the LED. The circuit and equivalent limit resistance
presented to the virtual series circuit and LED and power source resides
in the tail of the lamp and is electronic. Bike riders use similar LED
drivers units in headlights that flash (tail lights too).

I know more about LED flashlights and bike lights than a "see if there
are any resistors" dope like you ever will.

John Larkin

unread,

Feb 4, 2012, 1:37:19 PM2/4/12

to

Cool. How does it actually work?

Hattori Hanzo

unread,

Feb 4, 2012, 1:54:57 PM2/4/12

to

On Sat, 04 Feb 2012 10:37:19 -0800, John Larkin

<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>On Sat, 04 Feb 2012 03:23:03 -0800, Hattori Hanzo
><Outint...@billsbackyard.org> wrote:
>
>>On Fri, 03 Feb 2012 19:17:51 -0800, John Larkin
>><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>
>>>
>>>Do you own any LED flashlights? Open one up and see if there are any
>>>resistors.
>>>
>>
>> Mine are not the dumb versions. The switch, and control circuit is in
>>the tail. That means that it takes the 4.5 volts in and correctly
>>controls the feed to a single, super-bright LED with 3 distinct modes.
>>
>> It is so well designed that when your batteries begin to avalanche,
>>the second mode no longer functions precisely as designed, signaling
>>the operator to use the first mode or third mode, but to also prepare to
>>replace or recharge your cells. The 'nose' end of the battery pack gets
>>applied directly to the LED. The circuit and equivalent limit resistance
>>presented to the virtual series circuit and LED and power source resides
>>in the tail of the lamp and is electronic. Bike riders use similar LED
>>drivers units in headlights that flash (tail lights too).
>>
>> I know more about LED flashlights and bike lights than a "see if there
>>are any resistors" dope like you ever will.
>
>Cool. How does it actually work?

It is beyond your grasp, resistor boy.

Go read your degreaser manual.

John Larkin

unread,

Feb 4, 2012, 2:16:56 PM2/4/12

to

On Sat, 04 Feb 2012 10:54:57 -0800, Hattori Hanzo

<Outint...@billsbackyard.org> wrote:

>On Sat, 04 Feb 2012 10:37:19 -0800, John Larkin
><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>
>>On Sat, 04 Feb 2012 03:23:03 -0800, Hattori Hanzo
>><Outint...@billsbackyard.org> wrote:
>>
>>>On Fri, 03 Feb 2012 19:17:51 -0800, John Larkin
>>><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>>
>>>>
>>>>Do you own any LED flashlights? Open one up and see if there are any
>>>>resistors.
>>>>
>>>
>>> Mine are not the dumb versions. The switch, and control circuit is in
>>>the tail. That means that it takes the 4.5 volts in and correctly
>>>controls the feed to a single, super-bright LED with 3 distinct modes.
>>>
>>> It is so well designed that when your batteries begin to avalanche,
>>>the second mode no longer functions precisely as designed, signaling
>>>the operator to use the first mode or third mode, but to also prepare to
>>>replace or recharge your cells. The 'nose' end of the battery pack gets
>>>applied directly to the LED. The circuit and equivalent limit resistance
>>>presented to the virtual series circuit and LED and power source resides
>>>in the tail of the lamp and is electronic. Bike riders use similar LED
>>>drivers units in headlights that flash (tail lights too).
>>>
>>> I know more about LED flashlights and bike lights than a "see if there
>>>are any resistors" dope like you ever will.
>>
>>Cool. How does it actually work?
>
> It is beyond your grasp, resistor boy.

I think all you know about that flashlight is how to push the button
on the end.

John S

unread,

Feb 4, 2012, 2:45:03 PM2/4/12

to

You give him way too much credit, John.

Hattori Hanzo

unread,

Feb 4, 2012, 2:46:35 PM2/4/12

to

On Sat, 04 Feb 2012 11:16:56 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>I think all you know about that

degreaser is that your employees use it, not you. The reason is because
even that simple device is beyond your understanding.

So you are not going to understand the drive circuit in my LED lamp.

Hell, it took you nearly a decade to figure out how to add a pathetic,
self aggrandizing sig to your pathetic, abusive posts.

Michael A. Terrell

unread,

Feb 4, 2012, 2:56:10 PM2/4/12

to

John Larkin wrote:
>
> I think all you know about that flashlight is how to push the button
> on the end.

He knows how to push your buttons. :(

--
You can't have a sense of humor, if you have no sense.

Jamie

unread,

Feb 4, 2012, 5:18:04 PM2/4/12

to

In other words, you don't know?

Jamie

John Fields

unread,

Feb 4, 2012, 8:03:59 PM2/4/12

to

---
Actually, I did peruse two of those sites and found that with the
voltage drop across the diode, at the currents you mentioned, both
cases resulted in real-world power dissipation in excess of the limits
allowed.

In addition, the Fairchild site which, in figure 5, clearly plots the
ordinate in millivolts, is wrong.

Apparently not having noticed that, I suspect you'll now make excuses
and bitch on forever about how unimportant accuracy is.

--
JF

John Larkin

unread,

Feb 4, 2012, 9:08:32 PM2/4/12

to

On Sat, 04 Feb 2012 19:03:59 -0600, John Fields

Now you are weaseling.

>
>In addition, the Fairchild site which, in figure 5, clearly plots the
>ordinate in millivolts, is wrong.
>
>Apparently not having noticed that, I suspect you'll now make excuses
>and bitch on forever about how unimportant accuracy is.

It's not a matter of "accuracy", it's an obvious typo. The axis is
obviously volts.

--

John Larkin, President Highland Technology Inc

www.highlandtechnology.com jlarkin at highlandtechnology dot com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators

Custom timing and laser controllers

Photonics and fiberoptic TTL data links

John Fields

unread,

Feb 5, 2012, 11:24:56 AM2/5/12

to

On Sat, 04 Feb 2012 18:08:32 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>On Sat, 04 Feb 2012 19:03:59 -0600, John Fields
><jfi...@austininstruments.com> wrote:
>

>>Actually, I did peruse two of those sites and found that with the
>>voltage drop across the diode, at the currents you mentioned, both
>>cases resulted in real-world power dissipation in excess of the limits
>>allowed.
>
>Now you are weaseling.

---
Nope, just stating a fact.
---

>>In addition, the Fairchild site which, in figure 5, clearly plots the
>>ordinate in millivolts, is wrong.
>>
>>Apparently not having noticed that, I suspect you'll now make excuses
>>and bitch on forever about how unimportant accuracy is.
>
>It's not a matter of "accuracy", it's an obvious typo. The axis is
>obviously volts.

---
If the axis was correctly annotated, would that not be more accurate
than if it were annotated incorrectly?

--
JF

John Larkin

unread,

Feb 5, 2012, 12:38:32 PM2/5/12

to

On Sun, 05 Feb 2012 10:24:56 -0600, John Fields

<jfi...@austininstruments.com> wrote:

>On Sat, 04 Feb 2012 18:08:32 -0800, John Larkin
><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>
>>On Sat, 04 Feb 2012 19:03:59 -0600, John Fields
>><jfi...@austininstruments.com> wrote:
>>
>
>
>>>Actually, I did peruse two of those sites and found that with the
>>>voltage drop across the diode, at the currents you mentioned, both
>>>cases resulted in real-world power dissipation in excess of the limits
>>>allowed.
>>
>>Now you are weaseling.
>
>---
>Nope, just stating a fact.
>---
>
>>>In addition, the Fairchild site which, in figure 5, clearly plots the
>>>ordinate in millivolts, is wrong.
>>>
>>>Apparently not having noticed that, I suspect you'll now make excuses
>>>and bitch on forever about how unimportant accuracy is.
>>
>>It's not a matter of "accuracy", it's an obvious typo. The axis is
>>obviously volts.
>
>---
>If the axis was correctly annotated, would that not be more accurate
>than if it were annotated incorrectly?

The y-axis on that fig 5 graph is obviously volts, and the "m" is
obviously a typo. Given that, there's no accuracy problem. My job is
to design stuff, not freeze in my tracks at the slightest excuse.

If you look at a smattering of 1N4148 data sheets, there's a huge
spread of I-V curves. So I certainly wouldn't design a circuit that
depends on the high-current I-V curve without making sure we'd
purchase only one vendors's parts, and even then it would be risky. I
wouldn't depend on their capacitance or reverse leakage behavior,
either. 1N4148 is a very sloppy part.

All of which is aside from the issue of paralleling LEDs, which lots
of people seem to do.

John Larkin

unread,

Feb 5, 2012, 12:44:44 PM2/5/12

to

On Sat, 04 Feb 2012 11:46:35 -0800, Hattori Hanzo
<Outint...@billsbackyard.org> wrote:

>On Sat, 04 Feb 2012 11:16:56 -0800, John Larkin
><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>
>>I think all you know about that
>
>degreaser is that your employees use it, not you. The reason is because
>even that simple device is beyond your understanding.
>
> So you are not going to understand the drive circuit in my LED lamp.
>

Cool, let's talk about that:

OK, you have your super multimode Wal-Mart LED flashlight, maybe with
optional laser pointer. The question is, how does it regulate the LED
current? Inductive boost? Capacitive charge pump? Resistor? Linear or
PWM current-regulator IC? Raw battery connected to LEDs?

You could take it apart and see what's inside. Or you could search the
web for controller ICs. Or, more fun, you could learn some things
about the issue without opening the flashlight at all.

Any ideas how?

John Larkin

unread,

Feb 5, 2012, 12:47:40 PM2/5/12

to

I don't know either. He should tell us how his flashlight regulates
current. Or, if he doesn't know (which he doesn't) he could do a
little work to find out, and tell us. Which he won't.

Bart!

unread,

Feb 5, 2012, 1:43:46 PM2/5/12

to

On Sun, 05 Feb 2012 09:44:44 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>
>
>OK, you have your super multimode Wal-Mart LED flashlight,

What is this another attempt at a social insult?

Fuck off and DIE, John Larkin. You are a sad excuse for a civil,
intelligent man. That may well mean that you qualify for none of the
three titles. Likely even. You are worse, in fact, than Ross Perot could
have ever been. And his remarks were not racist. It was just a poor
choice of words. You actually were trying to be insulting. Big
difference.

It means that a real, intelligent man like Ross has more honor and
character than an insulting, pathetic, self aggrandizing piece of shit
like you ever will.

k...@att.bizzzzzzzzzzzz

unread,

Feb 5, 2012, 1:49:11 PM2/5/12

to

On Sun, 05 Feb 2012 10:43:46 -0800, Bart! <B@rt_The_Sheriff_Is_A_Nig***!.org>
wrote:

>On Sun, 05 Feb 2012 09:44:44 -0800, John Larkin
><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>
>>
>>
>>OK, you have your super multimode Wal-Mart LED flashlight,
>
> What is this another attempt at a social insult?

Interesting that you would take a slight of a *flashlight* as such a personal
insult....

> Fuck off and DIE, John Larkin. You are a sad excuse for a civil,
>intelligent man. That may well mean that you qualify for none of the
>three titles. Likely even. You are worse, in fact, than Ross Perot could
>have ever been. And his remarks were not racist. It was just a poor
>choice of words. You actually were trying to be insulting. Big
>difference.
>
> It means that a real, intelligent man like Ross has more honor and
>character than an insulting, pathetic, self aggrandizing piece of shit
>like you ever will.

...it's like he was insulting your manhood. You certainly respond as such. I
wonder what Freud would have to say?

Bart!

unread,

Feb 5, 2012, 1:51:43 PM2/5/12

to

On Sun, 05 Feb 2012 09:44:44 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>The question is, how does it regulate the LED
>current?

A simple ditz like you wouldn't know.

> Inductive boost? Capacitive charge pump? Resistor? Linear or
>PWM current-regulator IC? Raw battery connected to LEDs?

Think about cost, idiot.

>You could take it apart and see what's inside.

Here we go with retarded primer boy, on his retarded primer boy soap
box... again.

> Or you could search the
>web for controller ICs. Or, more fun, you could learn some things
>about the issue without opening the flashlight at all.
>
>Any ideas how?

It's a 555 timer circuit, you fucking retard. Even when it appears to
be "full on",it is really being pulsed. It is detectable at night.

A fucking retarded punk like you deserves no further information.

Bart!

unread,

Feb 5, 2012, 1:57:00 PM2/5/12

to

On Sun, 05 Feb 2012 09:47:40 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>I don't know either.

The first time you have spoken the truth in this group in quite some
time.

> He should tell us how his flashlight regulates
>current.

I would, but only if it is MY desire, you pathetic, cringing little
milksop.

> Or, if he doesn't know (which he doesn't)

Back to the lying horseshit behavior already, I see. There was a time
when a man would simply drop you where you stand for pathetic, uncivil
behavior like that. That is where the Colt motto came from.

> he could do a
>little work to find out,

You are an idiot.

> and tell us. Which he won't.

For all you know, utilization of LEDs is what will make me rich.

Fact is, you don't know a goddamned thing, you pathetic, characterless,
honorless piece of shit.

Bart!

unread,

Feb 5, 2012, 2:02:54 PM2/5/12

to

On Sun, 05 Feb 2012 13:49:11 -0500, "k...@att.bizzzzzzzzzzzz"
<k...@att.bizzzzzzzzzzzz> wrote:

>Interesting that you would take a slight of a *flashlight* as such a personal
>insult....

You are too retarded to know what I was referring to as being an
insult, much less an insult to whom, which is obvious since you didn't
catch the remainder of the reference.

Essentially, you are almost as big an idiot as he is, except yours is a
senility thing. Before that, it was a stupid fucktard thing. His is a
sub-human thing.

whit3rd

unread,

Feb 5, 2012, 2:13:57 PM2/5/12

to

On Sunday, January 29, 2012 9:47:03 AM UTC-8, John Larkin wrote:

[about cheap LED flashlights, resistor-less]

> If the LEDs are purchased in quantity, all from the same wafers, they
> will share current pretty well without added resistance. Thermal
> runaway is not an issue once they get into their ohmic ranges, which
> most LEDs do at their design operating current.
>
> The idea that you can't parallel diodes is silly. Motorola sold high
> current paralleled-diode rectifier arrays decades ago.

But, Motorola's rectifier app notes clearly indicate that series chokes
as ballast elements were the preferred way to parallel high current
rectifiers. The flashlights work because the internal resistance of
the (usually 3 AAA) cells is part of the circuit: you can burn this type
of flashlight out by using a low-Z cell (AA, or AAA NiCd, for instance).

The flashlights that DO use DC/DC conversion (white LED out, single
AA cell for power) haven't got any voltage reference, they just impedance-
transform. A true current-regulated switchmode supply is considered too
expensive for this market.

Jamie

unread,

Feb 5, 2012, 2:22:16 PM2/5/12

to

Bart! wrote:

> On Sun, 05 Feb 2012 09:47:40 -0800, John Larkin
> <jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>
>
>>I don't know either.
>
>
>
> The first time you have spoken the truth in this group in quite some
> time.

You mean it is the first time you have understood comments made from
JL in a long time?

>
>>He should tell us how his flashlight regulates
>>current.
>
>
> I would, but only if it is MY desire, you pathetic, cringing little
> milksop.

I am sure it would be your desire to do so, if you only knew, and then
most likely would be a case where we wouldn't be able to shut you up!

>
>>Or, if he doesn't know (which he doesn't)
>
>
>
> Back to the lying horseshit behavior already, I see. There was a time
> when a man would simply drop you where you stand for pathetic, uncivil
> behavior like that. That is where the Colt motto came from.
>

And there were and still is, times where the pen is mightier than the sword.

>>he could do a
>>little work to find out,
>
>
> You are an idiot.

Yes, how shameful of him to make such rude remarks about you doing work!

>
>>and tell us. Which he won't.
>
>
>
> For all you know, utilization of LEDs is what will make me rich.

Well, there is a sucker born in every crowd. I am sure with your
skills you'll be able to find some. Most people tend to hang with their
own kind.

> Fact is, you don't know a goddamned thing, you pathetic, characterless,
> honorless piece of shit.

He may appear to you to be character less in your eyes, but I believe
that would be due to lack of time on his part, wasting such traits to
pamper the likes of you.

Have a good day, I am sure i'll be next on the list.

Jamie

Jamie

unread,

Feb 5, 2012, 2:24:46 PM2/5/12

to

It's truly amazing what woman are pushing out these days.

Jamie

Bart!

unread,

Feb 5, 2012, 2:31:33 PM2/5/12

to

On Sun, 05 Feb 2012 14:22:16 -0500, Jamie
<jamie_ka1lpa_not_v...@charter.net> wrote:

> I am sure it would be your desire to do so, if you only knew, and then
>most likely would be a case where we wouldn't be able to shut you up!

You're an idiot. Go away, little boy.

Bart!

unread,

Feb 5, 2012, 2:33:45 PM2/5/12

to

On Sun, 05 Feb 2012 14:22:16 -0500, Jamie
<jamie_ka1lpa_not_v...@charter.net> wrote:

> Well, there is a sucker born in every crowd. I am sure with your
>skills you'll be able to find some. Most people tend to hang with their
>own kind.

Which is why we see you "hanging with" John, particularly in the midst
of his dishonorable moments, which is nearly always.

You are pathetic. A coattail hanging twit, if there ever was one.

John Larkin

unread,

Feb 5, 2012, 2:38:34 PM2/5/12

to

On Sun, 05 Feb 2012 10:43:46 -0800, Bart!
<B@rt_The_Sheriff_Is_A_Nig***!.org> wrote:

>On Sun, 05 Feb 2012 09:44:44 -0800, John Larkin
><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>
>>
>>
>>OK, you have your super multimode Wal-Mart LED flashlight,
>
> What is this another attempt at a social insult?

Hey, I like Wal-Mart. Too bad there are none nearby. We go to the ones
in Sacramanto or Reno when we're in the neighborhood.

>
> Fuck off and DIE, John Larkin. You are a sad excuse for a civil,
>intelligent man. That may well mean that you qualify for none of the
>three titles. Likely even. You are worse, in fact, than Ross Perot could
>have ever been. And his remarks were not racist. It was just a poor
>choice of words. You actually were trying to be insulting. Big
>difference.
>
> It means that a real, intelligent man like Ross has more honor and
>character than an insulting, pathetic, self aggrandizing piece of shit
>like you ever will.

Which means that you don't know how your flashlight works, and you're
not interested anyhow.

John Larkin

unread,

Feb 5, 2012, 2:40:53 PM2/5/12

to

That still doesn't say how the current is controlled. Or what stores
the multiple operating states.

John Larkin

unread,

Feb 5, 2012, 2:44:22 PM2/5/12

to

On Sun, 05 Feb 2012 10:57:00 -0800, Bart!
<B@rt_The_Sheriff_Is_A_Nig***!.org> wrote:

>On Sun, 05 Feb 2012 09:47:40 -0800, John Larkin
><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>
>>I don't know either.
>
>
> The first time you have spoken the truth in this group in quite some
>time.
>
>> He should tell us how his flashlight regulates
>>current.
>
> I would, but only if it is MY desire, you pathetic, cringing little
>milksop.
>
>> Or, if he doesn't know (which he doesn't)
>
>
> Back to the lying horseshit behavior already, I see. There was a time
>when a man would simply drop you where you stand for pathetic, uncivil
>behavior like that. That is where the Colt motto came from.
>
>> he could do a
>>little work to find out,
>
> You are an idiot.
>
>> and tell us. Which he won't.
>
>
> For all you know, utilization of LEDs is what will make me rich.

Then start by understanding how they work. How does your high-end
flashlight regulate LED current?

Jim Thompson

unread,

Feb 5, 2012, 3:28:58 PM2/5/12

to

On Sun, 05 Feb 2012 09:38:32 -0800, John Larkin

Phil Hobbs

unread,

Feb 5, 2012, 3:31:32 PM2/5/12

to

Nah, some of my favourite transistors are bipolar.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

hobbs at electrooptical dot net
http://electrooptical.net

k...@att.bizzzzzzzzzzzz

unread,

Feb 5, 2012, 3:51:06 PM2/5/12

to

On Sun, 05 Feb 2012 14:22:16 -0500, Jamie
<jamie_ka1lpa_not_v...@charter.net> wrote:

>Bart! wrote:
>
>> On Sun, 05 Feb 2012 09:47:40 -0800, John Larkin
>> <jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>
>>
>>>I don't know either.
>>
>>
>>
>> The first time you have spoken the truth in this group in quite some
>> time.
>
> You mean it is the first time you have understood comments made from
>JL in a long time?

DimBulb didn't this time, either.

>>>He should tell us how his flashlight regulates
>>>current.
>>
>>
>> I would, but only if it is MY desire, you pathetic, cringing little
>> milksop.
>
> I am sure it would be your desire to do so, if you only knew, and then
>most likely would be a case where we wouldn't be able to shut you up!

When has knowledge ever had anything to do with AlwaysWrong's screeching on?

>>>Or, if he doesn't know (which he doesn't)
>>
>>
>>
>> Back to the lying horseshit behavior already, I see. There was a time
>> when a man would simply drop you where you stand for pathetic, uncivil
>> behavior like that. That is where the Colt motto came from.
>>
>
>
>And there were and still is, times where the pen is mightier than the sword.

Certainly mightier than Nymbecile's limp-wristed threats.

>>>he could do a
>>>little work to find out,
>>
>>
>> You are an idiot.
>
>Yes, how shameful of him to make such rude remarks about you doing work!

But not unexpected. It's the thought of a little knowledge that's really got
his back hair in a bun.

>>>and tell us. Which he won't.
>>
>>
>>
>> For all you know, utilization of LEDs is what will make me rich.
>
> Well, there is a sucker born in every crowd. I am sure with your
>skills you'll be able to find some. Most people tend to hang with their
>own kind.

Come on. Dimbulb couldn't calculate the resistance of an LED's ballast
resistor. He's called DimBulb for a reason.

>> Fact is, you don't know a goddamned thing, you pathetic, characterless,
>> honorless piece of shit.
>
> He may appear to you to be character less in your eyes, but I believe
>that would be due to lack of time on his part, wasting such traits to
>pamper the likes of you.
>
>
> Have a good day, I am sure i'll be next on the list.
>

Back of the line. Back of the line.

Jamie

unread,

Feb 5, 2012, 4:07:09 PM2/5/12

to

You're correct to assume so, I hang with the best!

At least you got something correct, for a change.

Jamie

Jim Thompson

unread,

Feb 5, 2012, 4:06:30 PM2/5/12

to

Mine, too >:-}

I love to cook with wine. Sometimes I even put it in the food.

Jamie

unread,

Feb 5, 2012, 4:26:38 PM2/5/12

to

:)

k...@att.bizzzzzzzzzzzz

unread,

Feb 5, 2012, 4:28:41 PM2/5/12

to

On Sun, 05 Feb 2012 15:31:32 -0500, Phil Hobbs
<pcdhSpamM...@electrooptical.net> wrote:

Me too. The unipolar ones are just, well, weird.

John Larkin

unread,

Feb 5, 2012, 4:55:38 PM2/5/12

to

On Sun, 05 Feb 2012 10:57:00 -0800, Bart!
<B@rt_The_Sheriff_Is_A_Nig***!.org> wrote:

>On Sun, 05 Feb 2012 09:47:40 -0800, John Larkin
><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>
>>I don't know either.
>
>
> The first time you have spoken the truth in this group in quite some
>time.
>
>> He should tell us how his flashlight regulates
>>current.
>
> I would, but only if it is MY desire, you pathetic, cringing little
>milksop.
>
>> Or, if he doesn't know (which he doesn't)
>
>
> Back to the lying horseshit behavior already, I see. There was a time
>when a man would simply drop you where you stand for pathetic, uncivil
>behavior like that. That is where the Colt motto came from.

Which Colt motto? There seem to be a number of them.

Phil Hobbs

unread,

Feb 5, 2012, 5:04:17 PM2/5/12

to

John Larkin wrote:
>
> On Sun, 05 Feb 2012 10:57:00 -0800, Bart!
> <B@rt_The_Sheriff_Is_A_Nig***!.org> wrote:
>
> >On Sun, 05 Feb 2012 09:47:40 -0800, John Larkin
> ><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
> >
> >>I don't know either.
> >
> >
> > The first time you have spoken the truth in this group in quite some
> >time.
> >
> >> He should tell us how his flashlight regulates
> >>current.
> >
> > I would, but only if it is MY desire, you pathetic, cringing little
> >milksop.
> >
> >> Or, if he doesn't know (which he doesn't)
> >
> >
> > Back to the lying horseshit behavior already, I see. There was a time
> >when a man would simply drop you where you stand for pathetic, uncivil
> >behavior like that. That is where the Colt motto came from.
>
> Which Colt motto? There seem to be a number of them.

I assume "It works every time!" (*)

Cheers

Phil Hobbs

(*)
http://blogs.riverfronttimes.com/dailyrft/2009/06/warning_it_works_every_time_colt_45_ads_now_come_with_disclaimer.php

John Larkin

unread,

Feb 5, 2012, 5:08:18 PM2/5/12

to

So far, life has been a huge amount of fun. But what does that have to
do with diodes?

Jim Thompson

unread,

Feb 5, 2012, 5:39:42 PM2/5/12

to

On Sun, 05 Feb 2012 14:08:18 -0800, John Larkin

The same as talking amps of current in a 1N4148.

The first sign of senility is persistently trying to be an asshole

The second sign of senility is touting your company's wonderful
circuit designs as your own, while posting amateur crap on S.E.D

The third sign is acting like Polly Prissypants :-)

josephkk

unread,

Feb 5, 2012, 6:13:38 PM2/5/12

to

On Thu, 02 Feb 2012 21:09:10 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>On Thu, 02 Feb 2012 20:28:15 -0800, josephkk
><joseph_...@sbcglobal.net> wrote:
>
>>On Sat, 28 Jan 2012 17:42:27 -0800, John Larkin
>><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>
>>>
>>>>>I don't really think an "arbitrary" definition is necessary, since the
>>>>>location of the knee has been with us for decades.
>>>>
>>>>John, I told you there are people who believe there is a "knee" in an
>>>>exponential function. ;-)
>>>
>>>And there are people who think that the only kind of resistance is
>>>E/I.
>>>
>>>Some famous person once said "When all you know is Ohm's Law,
>>>everything looks like a resistor."
>>>
>>>Well, that was me, actually.
>>>
>>
>>Switch famous to infamous and the description of yourself is all too apt.
>>
>>?-)
>
>After almost a week of brooding, that's all you can come up with?

Naw, it is all you are worth. But it is fun to rag on your vainglory.

?-)

John Larkin

unread,

Feb 5, 2012, 6:22:14 PM2/5/12

to

On Sun, 05 Feb 2012 15:39:42 -0700, Jim Thompson

The datasheets I linked to go to 500 and 800 mA. Presumably those were
pulsed measurements, but don't say so.

But what is a 1N4148 anyhow? I suppose JEDEC specs exist, but I've
never seen one. And if JEDEC defines the part, where's the
enforcement? What keeps some Chinese chop shop from shipping anything
they care to call a 1N4148?

We have a number of cases where we have stock numbers for, say, a
generic 2N7002, and another stock number for a Fairchild 2N7002,
because sometimes it matters. It's safer to use a house number than a
1N or 2N number, if performance really matters.

I remember the classic 2N3055 fiasco, when people started selling tiny
epitaxial parts in aluminum cans as "2N3055", and they tended to blow
up.

Hmmm, I don't seem to use many 1N4148s. They appear on only about 10
of our 800 BOMs. The MELF version on even fewer. We do use a ton of
BAV types.

Anybody know how, say, a BAV99 is defined and enforced?

Jim Thompson

unread,

Feb 5, 2012, 6:50:10 PM2/5/12

to

On Sun, 05 Feb 2012 15:22:14 -0800, John Larkin

Unfortunately you do indeed need to evaluate each vendor's version...
in the mid '70's, at GenRad, I blacklisted Motorola's version of the
LM324 as absolute crap.

John Fields

unread,

Feb 5, 2012, 6:56:08 PM2/5/12

to

On Sun, 05 Feb 2012 13:28:58 -0700, Jim Thompson

<To-Email-Use-Th...@On-My-Web-Site.com> wrote:

>On Sun, 05 Feb 2012 09:38:32 -0800, John Larkin
><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>
>>On Sun, 05 Feb 2012 10:24:56 -0600, John Fields
>><jfi...@austininstruments.com> wrote:

>>>---
>>>If the axis was correctly annotated, would that not be more accurate
>>>than if it were annotated incorrectly?
>>
>>The y-axis on that fig 5 graph is obviously volts, and the "m" is
>>obviously a typo. Given that, there's no accuracy problem.

---
Of course there is, since "obvious" is relative and the annotation is
incorrect and requires disambiguation from the reader before the
proper value can be discerned.
---

>>My job is to design stuff, not freeze in my tracks at the slightest excuse.

---
If you enjoy designing stuff, then that's what you should do, instead
of freezing in your tracks any time you're corrected and feel forced
by your insecurities to mount an offensive defense.
---

>>If you look at a smattering of 1N4148 data sheets, there's a huge
>>spread of I-V curves.

---
I think that devices which are advertised as 1N4148's have to adhere
to the limits of the JEDEC spec for 1N4148's, so if one stays within
the recommended limits, then the spec's for devices which claim to be
1N4148's will be equivalent.
---

>>So I certainly wouldn't design a circuit that
>>depends on the high-current I-V curve without making sure we'd
>>purchase only one vendors's parts, and even then it would be risky.

---
Well, 1N4148's _aren't_ 1N4001's, so your trepidation is justified.
---

>>I wouldn't depend on their capacitance or reverse leakage behavior,
>>either.

---
Why not?

As a servoed varactor or a voltage variable resistor (in either
forward or reverse mode) it might just be the perfect part for the
application, for much less than pennies per unit.
---

>>1N4148 is a very sloppy part.

---
A poor workman blames his tools.
---

>>All of which is aside from the issue of paralleling LEDs, which lots
>>of people seem to do.

---
Sure they do, but the issue isn't about paralleling LEDs, it's about
paralleling LEDs without the use of current limiting resistors.

So far, your position has been that, even though LEDs have a negative
temperature coefficient of resistance at low currents, the tempco goes
positive at higher currents and can, thus, protect the LED array from
burnout from a voltage source.

What current are you talking about where the tempco goes positive?
---

>It must be hell to be manic depressive.

---
Heaven or Hell, as I understand it.

--
JF

Jim Thompson

unread,

Feb 5, 2012, 7:00:08 PM2/5/12

to

On Sun, 05 Feb 2012 17:56:08 -0600, John Fields

Which part are we being subjected to ?:-)

FigureItOut

unread,

Feb 5, 2012, 7:06:09 PM2/5/12

to

On Sun, 05 Feb 2012 15:22:14 -0800, John Larkin
<jjla...@highNOTlandTHIStechnologyPART.com> wrote:

>
>The datasheets I linked to go to 500 and 800 mA. Presumably those were
>pulsed measurements, but don't say so.
>
>But what is a 1N4148 anyhow? I suppose JEDEC specs exist, but I've
>never seen one. And if JEDEC defines the part, where's the
>enforcement? What keeps some Chinese chop shop from shipping anything
>they care to call a 1N4148?

Ummm... Performance.

>
>We have a number of cases where we have stock numbers for, say, a
>generic 2N7002, and another stock number for a Fairchild 2N7002,
>because sometimes it matters. It's safer to use a house number than a
>1N or 2N number, if performance really matters.

You need to wonder?

>
>I remember the classic 2N3055 fiasco, when people started selling tiny
>epitaxial parts in aluminum cans as "2N3055", and they tended to blow
>up.

That was a completely different animal. Money was to be made from that
differential. It is a bit harder to make the smallest form factor glass
diode (from the axial days). They have to fit inside the package and they
have to work. It is a pretty tight window.

>Hmmm, I don't seem to use many 1N4148s.

It, and the specs it had are one of the most commonly used diodes there
is. It is a "small signal" purpose device, so all you surface mount and
even integrated diodes would follow its spec to a degree as the switch
from discreet TTL to integrated devices came to being. Of course the
elements are far smaller now.

> They appear on only about 10
>of our 800 BOMs.

But the effect they had on the industry made its way into nearly
everything you have.

> The MELF version on even fewer. We do use a ton of
>BAV types.

Is it the 1N4148?

>Anybody know how, say, a BAV99 is defined and enforced?

It is certain that the original maker released a spec and the contract
fabs follow(ed) it. To easy to shoot oneself in the foot these days.

I doubt there is so much actual fakery going on as there used to be.
Nowadays, they sell you outright failures that had zero cost attached.

Don't see many chips being reverse engineered anymoreeither.By the time
they chop off the lid and analyze it, it is obsolete and the next great
chip is out. It costs far too much to risk so many assets trying to
reverse engineer a chip.

But a diode isn't a chip.

But a 1N4148 isn't that hard to make and get right either.

John Fields

unread,

Feb 5, 2012, 7:08:01 PM2/5/12

to

On Sun, 05 Feb 2012 14:08:18 -0800, John Larkin

---
At the end, you can't go back.

--
JF

John Larkin

unread,

Feb 5, 2012, 7:45:33 PM2/5/12

to

On Sun, 05 Feb 2012 15:13:38 -0800, josephkk

<joseph_...@sbcglobal.net> wrote:

>On Thu, 02 Feb 2012 21:09:10 -0800, John Larkin
><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>
>>On Thu, 02 Feb 2012 20:28:15 -0800, josephkk
>><joseph_...@sbcglobal.net> wrote:
>>
>>>On Sat, 28 Jan 2012 17:42:27 -0800, John Larkin
>>><jjla...@highNOTlandTHIStechnologyPART.com> wrote:
>>>
>>>>
>>>>>>I don't really think an "arbitrary" definition is necessary, since the
>>>>>>location of the knee has been with us for decades.
>>>>>
>>>>>John, I told you there are people who believe there is a "knee" in an
>>>>>exponential function. ;-)
>>>>
>>>>And there are people who think that the only kind of resistance is
>>>>E/I.
>>>>
>>>>Some famous person once said "When all you know is Ohm's Law,
>>>>everything looks like a resistor."
>>>>
>>>>Well, that was me, actually.
>>>>
>>>
>>>Switch famous to infamous and the description of yourself is all too apt.
>>>
>>>?-)
>>
>>After almost a week of brooding, that's all you can come up with?
>
>Naw, it is all you are worth. But it is fun to rag on your vainglory.
>
>?-)

That took three days?