Light bulb is series with unit under test

[001]

Hi

I see few recommendation about testing gear with Light bulb in series with line AC voltage
But nobody can tell what about power and voltage of theese "safety" lamps
Do You use this method?

[sokoloff]

For low expected current devices, this is good short circuit protection. (If dead shorted, you have. A chance to troubleshoot without blowing fuses.)

Standard (incandescent!!) light bulbs work fine. I use 60W, because that’s the most common type I have on hand. You can parallel multiples if you need 500mA or more of typical current.

[Paul Rose]

Use incandescent bulb with voltage rating same as mains.  This will act as a crude current limit for your device.   Bulb wattage depends on how much power you expect your device to draw.

Worst case, the bulb lights up full brightness, the current is limited to the normal bulb current, and the device does not smoke or catch fire.

If the wattage of the bulb is too high, you don't protect as much.  Start smaller (you won't damage the bulb).

The thinking is, if the bulb lights up bright, your test piece was drawing too much.  Unplug, investigate and fix.

I usually see this in guides on repairing old vacuum tube radios.   

[001]

You can parallel multiples if you need 500mA or more of typical current.

Sorry
Can You tell why I need more current if unit failed test?

[Ian.M]

http://www.repairfaq.org/sam/tshoot.htm#tshslt

[001]

http://www.repairfaq.org/sam/tshoot.htm#tshslt

Strange article
Why author need bulbs from 25 to 200 Watts?

[Paul Rose]

Why author need bulbs from 25 to 200 Watts?

Some devices you are testing need more power than others.

A 25 watt bulb will light up fully if you are testing a properly functioning device that needs 200 watts.  The device will be starved, because the 25 watt bulb will limit the current ( and thus the power ).

It seems like he recommends "sneaking up" on the correct size.  I suspect you would develop some intuition with this technique over time, but if you don't know what the power draw is supposed to be you can still get into trouble.  But even a 200 watt bulb will do less damage than just plugging in something that might have a short.

[mvs]

Why author need bulbs from 25 to 200 Watts?

Incandescent light bulb is like PTC fuse. If filament wire is cold it has low resistance, if it gets hot it will have 10-15x higher resistance.
200W rated bulb has lower resistance, then 25W if rated voltage and filament temperature are the same.

[001]

Why no standard industrial test units like variacs with build-in bulbs?

[sokoloff]

You can parallel multiples if you need 500mA or more of typical current.

Sorry
Can You tell why I need more current if unit failed test?

A 60W light bulb conducts 500mA at 120V (by Ohm's Law).

If your device requires 250mA, a light bulb in series will also draw 250mA, probably dropping ~30V or so and your device would see 90V at 250mA.

If you have a high draw device (say a full amp), you will need multiple lamps in order to conduct an amp of current at an acceptably low voltage drop. (A single 60W lamp will not permit a full amp to pass in steady state.)

[joseph nicholas]

Keep in mind the average household has only 20 amp wires.  All this comes down to a reasonably safe level when you are an able bodied person not wearing a pacemaker or a midget.  This is true if you dead short 120 volts to neutral through a screwdriver.  A nasty spark will explode in front of you but it will just startle you.  You will not die.

[Ian.M]

Sokoloff's explanation neglects the highly non-linear resistance of a tungsten filament.  The cold resistance is typically only a tenth of the resistance at normal operating temperature (calculated from its rated power or measured current and voltage). 

Therefore initially your 60W 12V bulb may pass significantly over 500mA without excessive voltage drop as long as the startup surge is short.   If there's  a problem with the SMPSU control chip and its over-driving the chopper transistor, the light bulb will save the bridge rectifier and the fuse, and if you worked up to the right wattage for the circuit, may even save the replacement chopper transistor.   Of course you *should* have checked the gate or base drive to the transistor while it was out of circuit, and if it looked suspicious (or the if the current sense resistor had blown) changed the control chip as well.

Why not integrated with a variac? Well its often better to put the bulb in series with the positive DC bus feed to the SMPsU transformer, so it doesn't heat up on the inrush current of the reservoir cap, so a simple non-metallic bulbholder you can swap bulbs in equipped with a shortish length of flex you can solder the ends of into the circuit is more versatile.

N.B. with many SMPSUs a bulb that will protect it against a primary side fault is too small to allow the whole device to run. You may have to disconnect most secondary side loads to reduce the consumption enough for the SMPSU to start up without tripping its UvLO.

[sokoloff]

Sokoloff's explanation neglects the highly non-linear resistance of a tungsten filament.  The cold resistance is typically only a tenth of the resistance at normal operating temperature (calculated from its rated power or measured current and voltage).

I did mean to add "in steady state" to my first sentence, but in fairness, I did directly express the non-linearity in the post (120V drop at 500mA and [guessed] 30V [rather than 60V] drop at 250mA).

That non-linearity is indeed key to why the device works pretty well for most low power devices. (The lamp has a negligible voltage drop for small currents, meaning a device that works properly and is low power will see very nearly the full line voltage and work "normally".)

[001]

is it good idea to add two lamps in series for neutral and hot wires?

[Ian.M]

Nope.   

I considered asking what your chain of thought was behind that question, but decided I *really* *DONT* want to know! 

[Doctorandus_P]

No, you do not want to put lamps in series.

MVS was right about the PTC thing, I'll elaborate:
A cold incandesent light has a relatively low resistance, which means it can pass a lot of current.
If you put it in series with an AUT (Appliance Under Test). and this appliance limits the current to a "normal" value then the lamp will not light. It dissipates only a small amount of power because the voltage drop over the lamp will also be small because of the small current and resistance.

At some point the current through the lightbulb reaches a point that it starts generating some heat. This heat increases the resistance of the filament, and with more resistance (and the same current) it dissipates more power. This is a chain reaction and your lamp will light up.

When using light bulbs as a PTC fuse you generally want to stay in the region where the filament stays cold enough to pass the current without a significant voltage drop.

Some repair shops used to have a construction with a few parallel light fittings screwed to some frame, and they could easily adjust the "fuse" current by adding / removing / exchanging some of the incandesent lamps. This is not an exact science. If the lamps light up, simply turn it off, ad an extra lamp and try again. If you need too many lamps for a small appliance and they still light up, then you know there is a short or similar.

This video from PhotonInduction is a good demonstration of this effect.
He starts with pre-heating a 20kW incandescent light with 12V and it draws about 40A and the current reduces while the filament is heating up.

[Zero999]

is it good idea to add two lamps in series for neutral and hot wires?

No, just one in series with phase is what you need.

Why no standard industrial test units like variacs with build-in bulbs?

With a variac, the voltage can be increased slowing, whilst monitoring the current and the power cut, if it rises too high.

[tautech]

001, it's a 'dim bulb tester', Google will fill you in on them.

From an old post:

My Dim Bulb Tester

Just a incandescent bulb wired in series with the DUT, and good to have a small range of bulbs.

This one is switched SPST, but in countries without a polarized main plug I would recommend a double pole switch
As is evident it was knocked together at no cost in a dead PC ATX case, using the existing IEC mains socket and a few "bits and bobs".

On power-on the bulb should momentarily flash bright, then dim after inrush currents subside.(if all is well).
If the bulb remains bright and the bulb wattage is appropriate for the DUT then "Houston, we have/still have a problem. 

If you have a DUT that has emitted the magic smoke or continually blows fuses this can be indispensable.

[001]

Nice idea
bulb in place of fuse http://www.antiqueradios.com/forums/viewtopic.php?f=1&t=306219

[bsudbrink]

Here's a picture of my dim bulb tester, almost always used with the variac also in the picture.  Note the "handy" switch.  It lets me take the bulb in and out of the circuit without yanking on the cords.

[David_AVD]

We made up a lamp unit at work years ago.  It has a switch to put a second lamp in parallel with the first and another switch to bypass the lamps.  The bypass switch allows the removal of the lamps from the circuit without replugging the DUT directly to the mains.

[Sylvi]

Hi
The other name for the light bulb test jig is "Power Limiting Safety Socket" and what it does is right in the name. A variac does not limit power per se; just voltage.

The PLSS limits power using the characteristics of incandescent lamps. When cold, they have low resistance and can pass a fair amount of current before they heat up enough to increase their internal resistance and conduct their rated operating current (which produces their rated heat output).

With a typical transformer, diodes and cap as a unit to test, you plug this into a receptacle that is in series with the lamps. At turn on, the charging current needed for the cap will be pulled ultimately through the lamp and the lamp will brighten a little bit. When the cap is charged it does not pull any more current, so the bulb dims to nothing. The voltage across the cap will be lower than without the bulnb in series, as the bulb drops a bit of mains voltage to the PT. if the unit under test has other supplies, they will all be lower but their polarity can be checked safely.

If the unit is a tube amp,say, then adding some preamp tubes will increase the load and you need a bigger bulb or two bulbs in parallel to power the tube heaters and their plate supply. As you progress through testing and gain confidence that the circuit is working, you can go to very high-wattage lamps and eventually to bypassing the bulbs and relying on just fuse protection.

My PLLS has two sockets in parallel all in parallel with a switch. To use just one lamp I unscrew the second one a little bit, or tighten it back in to use it. There is an overall on-off switch.

If you live in a 100-120V area, you can use 240V lamps for more extreme limiting. Generally 25W is okay for very low-power items, 40-60W is more typical, and 100-150W flood-lamps are useful for preliminary power output testing even of solid-state amps.

You can build a PLSS using home electrical supplies from a hardware store and just mount it on a board. Some of the biggest name audio manufacturers use exactly these units to test their amp and save $$ on fuses.

You CANNOT use any other type of lamp than incandescent for this function.