I don't have a 12V 45W lamp, but I attached a photo catching the ON
transient for a 12V 12W lamp like this:


12V 12W
| ---[DUT]---+--------+
| | |
+12VDC--O [0.1R] [HP56402B]
| |
GND-------------------+--------+



and for a 120V 25W lamp like this:


120V 25W
| ---[DUT]---+--------+
| | |
+120DC--O [10R] [HP56402B]
| |
GND-------------------+--------+

I've also attached some photos using AC excitation of 60w, 100W, and
200W lamps so you can get a feel for the decay time variation with lamp
power rating.

I don't recall the value of the sensing resistor, but knowing that the
inrush surge current is pretty close to 10 times the lamp's rating
should make it easy to figure out if you need it.

On heat sinking, since the part can handle 35 amps all day long with a
case temp of 100C and your worst case surge is going to be some 40 amps
decaying down to about 4 in some tens of milliseconds, I'd say you don't
need it at all.

JF

On Fri, 04 Dec 2009 08:19:45 -0600, John Fields
wrote:

I don't have a 12V 45W lamp, but I attached a photo catching the ON
transient for a 12V 12W lamp like this:


12V 12W
| ---[DUT]---+--------+
| | |
+12VDC--O [0.1R] [HP56402B]
| |
GND-------------------+--------+



and for a 120V 25W lamp like this:


120V 25W
| ---[DUT]---+--------+
| | |
+120DC--O [10R] [HP56402B]
| |
GND-------------------+--------+

I've also attached some photos using AC excitation of 60w, 100W, and
200W lamps so you can get a feel for the decay time variation with lamp
power rating.

I don't recall the value of the sensing resistor, but knowing that the
inrush surge current is pretty close to 10 times the lamp's rating
should make it easy to figure out if you need it.

On heat sinking, since the part can handle 35 amps all day long with a
case temp of 100C and your worst case surge is going to be some 40 amps
decaying down to about 4 in some tens of milliseconds, I'd say you don't
need it at all.

JF

You don't mention the minimum time enforced between scope shots. These
things take quite a while to cool to ambient. This has caught many a
design verification short, as the 'first' shot is not easy to
duplicate, without patience.

Simple low power measurement of lamp resistance might give a better
idea of what to expect as a valid recorded starting pulse and the
likely effects of sensing resistor, source or switch impedance.

RL

On Sat, 05 Dec 2009 00:35:44 -0500, legg wrote:

On Fri, 04 Dec 2009 08:19:45 -0600, John Fields
wrote:

I don't have a 12V 45W lamp, but I attached a photo catching the ON
transient for a 12V 12W lamp like this:


12V 12W
| ---[DUT]---+--------+
| | |
+12VDC--O [0.1R] [HP56402B]
| |
GND-------------------+--------+



and for a 120V 25W lamp like this:


120V 25W
| ---[DUT]---+--------+
| | |
+120DC--O [10R] [HP56402B]
| |
GND-------------------+--------+

I've also attached some photos using AC excitation of 60w, 100W, and
200W lamps so you can get a feel for the decay time variation with lamp
power rating.

I don't recall the value of the sensing resistor, but knowing that the
inrush surge current is pretty close to 10 times the lamp's rating
should make it easy to figure out if you need it.

On heat sinking, since the part can handle 35 amps all day long with a
case temp of 100C and your worst case surge is going to be some 40 amps
decaying down to about 4 in some tens of milliseconds, I'd say you don't
need it at all.

JF

You don't mention the minimum time enforced between scope shots. These
things take quite a while to cool to ambient. This has caught many a
design verification short, as the 'first' shot is not easy to
duplicate, without patience.

---
I waited a couple of minutes between shots, took a lot of shots and
posted the shot with the highest spike.
---

Simple low power measurement of lamp resistance might give a better
idea of what to expect as a valid recorded starting pulse and the
likely effects of sensing resistor, source or switch impedance.

---
Yes, of course, and that's been addressed in the thread on sed. My
purpose in posting these shots was to post what the OP wanted to see,
which along with the turn-on current surge, was the decay time to normal
operating current.

Just as an aside, one of these days I'll get around to doing if for
common incandescent lamps, but for now I've attached R derived from E
and I data for a 100W 120VAC lamp.

The 10 ohms with zero volts across the lamp was measured using an
ohmmeter.

I think I'll also run the current shots again, but this time with a
TRIAC triggered at the peak of the AC cycle and kept on until the
current settles.

That way I should get a good shot the first time instead of having to do
a bunch of random shots with waits in between to cool.


JF