electrical characteristics of tungsten

Thread Starter

paul

Joined Jan 28, 2007
1
I am carring out an experiment using a 12 volt 5wtt bulb. When I apply 6,9 and 12 volts to the bulb it draws a current of 0.56, 0.697 and 0.825 amps respectively. To me it seems as though the filament resistance in the lamp is nonlinear.

Could you explain why this is so. That is if my assupmtions are correct.

Thanks
 

Gadget

Joined Jan 10, 2006
614
The hotter the filiment is, the higher the resistance. It's a well known property of bulbs and heater elements and can be usefull.
 

thingmaker3

Joined May 16, 2005
5,083
It is this non-linearity which allows ordinary light bulbs to be used as current regulators in simple electroplating and anodizing rigs. I've seen one EDM that uses the same principle.
 

sparky 1

Joined Nov 3, 2018
756
There is a pdf: AN 263 sine wave generation techniques, page 5 shows the bulb called #327 I believe these were sold to aircraft manufacturers, possibly instrumentation light. (possibly 3W 28V. note makes reference to HP's famous audio generator designers. So the small light bulb was used to help fix a sine wave generator thermal drift error.
The bulb's has a positive temperature coefficient (tungsten). As a component combined with the right negative temperature coefficient properties in a circuit the result is a net of zero. So for materials science people tungsten can improve drift error in older analog type instrumentation. The article shows a later method for historical accuracy.
 

OBW0549

Joined Mar 2, 2015
3,566
Yes, your results show exactly that the filament resistance is non-linear. Voltage is not directly proportional to the current.
Ummm... no. Calling the resistance "non-linear" is a gross misapplication of the term.

The apparent non-linearity is entirely a result of the filament's self-heating combined with tungsten's temperature coefficient of resistance, and would disappear completely if the tungsten were held at a constant temperature (such as by being bonded to a large heat sink), or if its resistance were measured by applying voltage in pulses too short to cause significant heating.

To say that the tungsten filament's resistance is "non-linear" is just plain wrong.

And by the way: the person you were replying to hasn't been around here since the day he posted his OP more than 13 years ago; you are literally talking to thin air. Always check the dates on a post before replying to it. And also check to see if the poster is still active on the site (you can get this information by hovering your mouse pointer over the person's username alongside their post).
 

vu2nan

Joined Sep 11, 2014
345
Temperature coefficient of resistance of Tungsten is 0.0004 / Ω / ° C.

That of Nichrome 0.0045 / Ω / ° C.

Nichrome wire is used in heating elements.
 

OBW0549

Joined Mar 2, 2015
3,566
There is a pdf: AN 263 sine wave generation techniques, page 5 shows the bulb called #327 I believe these were sold to aircraft manufacturers, possibly instrumentation light. (possibly 3W 28V. note makes reference to HP's famous audio generator designers. So the small light bulb was used to help fix a sine wave generator thermal drift error.
That lamp has absolutely nothing to do with any "thermal drift error." NOTHING. If you think it does, you really don't understand how that circuit works.

Here's the circuit:
weinbridge.png
A Wein bridge oscillator has two feedback networks: one positive, and one negative.

The positive feedback network is the frequency-determining network, comprising the two 100kΩ resistors and 10nF capacitors. With equal resistors and equal capacitors, it has zero degrees phase shift at the resonant frequency, along with a 3:1 attenuation from the op amp's output back to its (+) input.

The negative feedback network consists of the 430Ω resistor and the #327 lamp. Its job is to set the gain of the amplifier at 3, to exactly compensate for the attenuation of the frequency-determining network. If the gain is less than three, even by an infinitesimal amount, oscillations will either die out or will never start in the first place; if it is greater than 3, oscillations will build and build until the op amp's output saturates, producing a distorted output.

What the lamp does, and this is critical, is to regulate the amplitude of the oscillations by varying its resistance and therefore the gain of the circuit. If the oscillations become too strong, the lamp filament heats up more and the resistance increases, lowering the gain; and if the oscillations become too weak the resistance lowers, raising the gain. The net result is to keep the oscillations at a constant amplitude.

Again: the lamp has absolutely NOTHING to do with "fixing" any sort of "thermal drift error."

(By the way, the #327 bulb is a 28 volt, 0.04 amp flange base bulb available from Digi-Key, Amazon, Mouser, and many other distributors. There's nothing special about it.)

The bulb's has a positive temperature coefficient (tungsten). As a component combined with the right negative temperature coefficient properties in a circuit the result is a net of zero.
This is just plain nonsense.

So for materials science people tungsten can improve drift error in older analog type instrumentation.
This is absolute, utter nonsense, too.
 

sparky 1

Joined Nov 3, 2018
756
Thank you OBW, I stated also for historical accuracy. Reiterating or looking for further clarification. The oscillations tend to build up and the gain exceeds 3. It is this amplitude rise that is being compensated?
Also for accuracy, on that same national semi AN 263 there is another approach to sine generation that's not uses a UTC ls-52 transformer. I think it is 8000 ohm to many smaller value secondaries. I wonder if this method was switching different secondaries rather than trying to use just one secondary?
 

OBW0549

Joined Mar 2, 2015
3,566
Reiterating or looking for further clarification. The oscillations tend to build up and the gain exceeds 3. It is this amplitude rise that is being compensated?
When the amplitude of the oscillations increases, the lamp heats up; this increases its resistance, which lowers the gain of the amplifier. This causes the amplitude of the oscillations to drop back down. Conversely, if the oscillation amplitude falls below the desired level the lamp cools and its resistance decreases. This increases the gain of the amplifier and the oscillations build back up. The system stabilises with the amplitude at a constant level. Very simple, nothing mysterious.

Also for accuracy, on that same national semi AN 263 there is another approach to sine generation that's not uses a UTC ls-52 transformer. I think it is 8000 ohm to many smaller value secondaries. I wonder if this method was switching different secondaries rather than trying to use just one secondary?
I have no idea what you're talking about when you say "different secondaries"; I see only one secondary and one primary. There is no switching going on.

NOTE: Section 4 on page 5, "High Voltage AC Calibrator," mistakenly refers the reader to Figure 2 (the basic Wein bridge circuit) when it should have said Figure 5; the latter circuit is the one that section is describing. (In case that wasn't obvious)
 
Last edited:

sparky 1

Joined Nov 3, 2018
756
I am at work now, limited on access. The UTC LS-52 I found matched with that model in a PDF Allied Radio catalog dated 1955. The 15W transformer was shielded. Most likely having taps. I was curious that the application note showed the team might have chose to use this transformer were interested in narrow range of low frequency possibly selecting by range? Some of the older low frequency generators had slot of select buttons. Possibly the cost of shipping was not the most important criteria for the design. Possibly electronic enthusist could either build or pay for an affordable low noise generator in that era? The extra shielded Arduino nano might radiate square pulses if the electronic enthusist was not carful.
 

OBW0549

Joined Mar 2, 2015
3,566
I am at work now, limited on access. The UTC LS-52 I found matched with that model in a PDF Allied Radio catalog dated 1955. The 15W transformer was shielded. Most likely having taps. I was curious that the application note showed the team might have chose to use this transformer were interested in narrow range of low frequency possibly selecting by range? Some of the older low frequency generators had slot of select buttons. Possibly the cost of shipping was not the most important criteria for the design. Possibly electronic enthusist could either build or pay for an affordable low noise generator in that era? The extra shielded Arduino nano might radiate square pulses if the electronic enthusist was not carful.
One thing to keep in mind while you're asking "why?" is that the circuit shown in Fig. 5 is an absolutely ancient design. I recall using the LM3900 quad Norton op amp back in the early 1970's, a half-century ago, but I don't recall seeing it used anywhere in at least 30 years. The LH0002 was first introduced shortly after the Roman conquest of Brittania (or was it around the time of the Battle of Hastings? I forget...) and I don't think it's even made anymore. So there's not much point in wondering why the design is the way it is-- it's ancient history, using obsolete components and methods.

In any case, I'm done here. Good luck with whatever it is you're trying to do.

Peace.
 
Top