halogen filament resistance at switch on

[matt]

I have to replace a battery supply for an array of 12v halogen lamps
(21 x 35watts=735 watts) with a 230 vac mains energised supply unit. I
have to use the existing installation, so require a regulated dc supply
with an adjustable output (about +2volts for voltage drop). The problem
I am having is selecting a supply unit as the switch on surge current
when the filaments are cold is likely to activate the supply protection
circuitry. I contacted Lambda supplies and they requested a current
profile of the halogen lamp at switch on in order to recommend a
suitable supply. Any help in sorting this out would be welcome.

Cheers

Matt

[Douglas G. Cummins]

Without actually profiling your bulbs, a good rule of thumb is that the
inrush current for a cold incandescent bulb is 10x the nominal current
at the applied voltage for a few milliseconds, quickly dropping down to
nominal current within 100 milliseconds (dependent on wattage of the
bulb).

The simplest thing to do is get a power supply with a soft-start, slowly
ramping the voltage applied to the bulbs (within a second or two would
be fine), rather than an instantaneous start. Most DC power supplies
these days can easily do this. Lambda doesn't need a current profile to
achieve this.

Post by matt
I have to replace a battery supply for an array of 12v halogen lamps
(21 x 35watts=735 watts) with a 230 vac mains energised supply unit. I
have to use the existing installation, so require a regulated dc supply
with an adjustable output (about +2volts for voltage drop). The problem
I am having is selecting a supply unit as the switch on surge current
when the filaments are cold is likely to activate the supply protection
circuitry. I contacted Lambda supplies and they requested a current
profile of the halogen lamp at switch on in order to recommend a
suitable supply. Any help in sorting this out would be welcome.
Cheers
Matt

[Victor Roberts]

On 12 Apr 2006 16:45:35 -0700, "matt"

Post by matt
I have to replace a battery supply for an array of 12v halogen lamps
(21 x 35watts=735 watts) with a 230 vac mains energised supply unit. I
have to use the existing installation, so require a regulated dc supply
with an adjustable output (about +2volts for voltage drop). The problem
I am having is selecting a supply unit as the switch on surge current
when the filaments are cold is likely to activate the supply protection
circuitry. I contacted Lambda supplies and they requested a current
profile of the halogen lamp at switch on in order to recommend a
suitable supply. Any help in sorting this out would be welcome.

I don't have a link for the profile, but you can get a good
idea of the cold resistance by going to any table of
tungsten resistance vs. temperature and calculating the
resistance ratio of tungsten for T=3000K vs T=300K. Then
apply this factor to the normal operating current.

You will probably need a supply with an active current
limit, one that can switch from voltage regulation to safe
current limiting without having a fit. If you use one of
these, the current vs. time profile on a fixed voltage
source is not operative anyway.
--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

[m***@worldnet.att.net]

Post by Victor Roberts
On 12 Apr 2006 16:45:35 -0700, "matt"

I contacted Lambda supplies and they requested a current profile of
the halogen lamp at switch on in order to recommend a suitable supply.

I don't have a link for the profile, but you can get a good
idea of the cold resistance by going to any table of
tungsten resistance vs. temperature and calculating the
resistance ratio of tungsten for T=3000K vs T=300K. Then
apply this factor to the normal operating current.

I agree with Douglas' stated 10x factor and Victor's implied 10x factor.
It's not the same kind of lamp, but I found the following on an 1157
automotive lamp:

The "book" specs are:

bright filament: 12.8V, 26.9W, 2.10A (R=V/I=12.8/2.1=6.1 ohms)
dim filament: 14.0V, 8.3W, 0.59A (R=24 ohms)

Actually measuring the lamp with my DMM gave 0.5 ohms for the bright
filament (hot = 12.2x cold) and 2.1 ohms for the dim filament (hot =
11.4x cold). By Ohm's Law, this means the bright filament is 328W and
the dim filament is 93W. If this were true, the tail light lenses on a
new car would probably melt the first time the driver stepped on the
brake.

When I connected the lamp to a "13.8 V" linear power supply I have, I get

bright filament 13.3V, 2.19A (29.1W, 6.07 ohms)
dim filament 13.5V, 0.57A (7.70W, 23.7 ohms)

These numbers are much closer to the published spec.

Another non-lighting application I am working with requires charging an
~8200 uF filter capacitor at the input of a power supply to around 300
VDC when power is switched on. Initially, the cap appears as a dead
short, causing a large inrush current. The standard way to avoid welding
relay contacts together in this application is to first charge the
capacitor through a resistor for a second or two, then switch to a direct
connection.

You might be able to implement the above with one component by using a
negative temperature coefficient thermistor. This will have a
resistance of a few ohms cold, decreasing to a fraction of an ohm hot -
the opposite of what tungsten does. Mouser sells one (527-CL30) that is
rated at 8 amps for US$1.40 in small quantities; I realize this isn't
enough current for your application, but perhaps you can find a beefier
one or use a couple in parallel.

Matt Roberds

[Victor Roberts]

Post by m***@worldnet.att.net

Post by Victor Roberts
On 12 Apr 2006 16:45:35 -0700, "matt"

I contacted Lambda supplies and they requested a current profile of
the halogen lamp at switch on in order to recommend a suitable supply.

I don't have a link for the profile, but you can get a good
idea of the cold resistance by going to any table of
tungsten resistance vs. temperature and calculating the
resistance ratio of tungsten for T=3000K vs T=300K. Then
apply this factor to the normal operating current.

I agree with Douglas' stated 10x factor and Victor's implied 10x factor.
It's not the same kind of lamp, but I found the following on an 1157

I never implied a 10x factor for the resistance. I gave a
10X factor for the temperature and suggested that the OP
look up the resistance change for this temperature change.

The resistivity of tungsten at 3000K is not 10 times its
resistivity at 300K, but somewhat greater than 16 times the
resistivity at 300K.
--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

[m***@worldnet.att.net]

Post by m***@worldnet.att.net

I don't have a link for the profile, but you can get a good idea
of the cold resistance by going to any table of tungsten resistance
vs. temperature and calculating the resistance ratio of tungsten
for T=3000K vs T=300K.

I agree with Douglas' stated 10x factor and Victor's implied 10x
factor. It's not the same kind of lamp, but I found the following

I never implied a 10x factor for the resistance. I gave a 10X factor
for the temperature and suggested that the OP look up the resistance
change for this temperature change.

You've caught me being lazy. I knew it wasn't linear but I figured it
was close enough. This time I actually went and did my homework and
found that the 3000 K resistivity is indeed a bit over 16 times the
300 K resistivity.

Going back to the 1157 lamp I measured, the bright filament's resistance
went up about 12.2x over cold, which implies it was running at something
in the 2300-2400 K range. The dim filament resistance went up about
11.4x, or 2200-2300 K. (The original poster was talking about a tungsten
halogen lamp, which would run closer to 3000 K; an 1157 is probably gas-
filled but no halogens.)

Matt Roberds

[Z***@pidham.vispa.com]

Just make sure your DC PSU has a current-limiting protection, NOT
foldback.

Jon

[matt]

Thanks for the advice folks

regards

Matt

[Charles]

Post by matt
I have to replace a battery supply for an array of 12v halogen lamps
(21 x 35watts=735 watts) with a 230 vac mains energised supply unit. I
have to use the existing installation, so require a regulated dc supply
with an adjustable output (about +2volts for voltage drop). The problem
I am having is selecting a supply unit as the switch on surge current
when the filaments are cold is likely to activate the supply protection
circuitry. I contacted Lambda supplies and they requested a current
profile of the halogen lamp at switch on in order to recommend a
suitable supply. Any help in sorting this out would be welcome.
Cheers
Matt

Okay, I'm a dumb guy, but why not just get an ohmmeter and measure the
resistance? The bulbs will be about as cold as when you first switch
them on.

[Don Klipstein]

Post by Charles
Okay, I'm a dumb guy, but why not just get an ohmmeter and measure the
resistance? The bulbs will be about as cold as when you first switch
them on.

I have tried that, and not always gotten good readings, especially when
the shell contact of the base is aluminum.
If you try this, play around for good contact. I would assume that the
lowest reading that can be sustained, especially if it can be repeated,
will be the correct one.

- Don Klipstein (***@misty.com)