[ASK] Incandescent lamps in series for load testing.

[BravoV]

Scored a really cheap, used AC/DC converter recently (photo attached), and I was only allowed to do on site test at the output voltage but unloaded with DMM, and it was spot on.

As its output max out at 36 Volt at 46.5 Amp, I have problem finding the suitable load at 36 V in order to load test it, and I mean quick & cheap way at constant load only. I do have many of car's 12Volt incandescent lamps, just lets assume they're all same model and spec.

The question is, can I use three of these lamps (12 V x 3 = 36 V) "in series" to load test this PSU ?

All I need is to test whether its capable of running say at about 30 plus Amps for about 6 hours, and also planning to measure it's output quality with scope while on load, and do some simple temperature measurements.

Also if there are any other tips for testing or verifying this particular PSU, I'm all ears.

Thanks in advance.

[alm]

Since incandescent bulbs have a positive temperature coefficient, I would expect this to work fairly well as long as you use lamps with the same voltage and current ratings.

An alternative might be this (although 6 hours might require a very large bucket or replacing the water multiple times):

[amyk]

36V * 46.5A = 1674W = 1.674kJ/s which is coincidentally almost exactly 0.4 of the specific heat constant of water, 4.184kJ/kg/K. It'll raise the temperature of 1L of water 0.4 degrees per second, or 200 seconds for an 80 degree rise (assuming 20C to boiling). 10L will take 2000s, 100L will take 20000s (5:33:20). Alternatively, it will take 6h to boil 108L of water at that power.

[oldway]

A very cheap solution is to use inox MIG welding wire to make power resistors.
A 0.8mm diameter wire can take up to about 5A at free air....

For 12V, I use about 1M of wire.

In order to wind it, it is first necessary to anneal it by elevating his temperature up to red hot.

[MrAl]

Hello,

I've used power resistors in distilled water, and it works pretty well at low voltages.  Never tried it at high voltage like 120v.

I've also used light bulbs and there is one caution.
light bulbs have a very large dynamic resistance range.  They can start out at around 1/10 the resistance of what they are when hot.  So a 12v bulb that draws 1 amp when hot (12 watts) could draw 10 amps when first turned on, then quickly heat up.  Trouble is inverters may trip on overcurrent before the bulbs heat up when you get nearer to full load.  To get around this you might be able to switch them on in banks, one at a time.  For example, of you wanted to test up to 360 watts and had 12v 1 amp bulbs, that would be 36 watts per string, so you'd need 10 strings of 3 bulbs each.  Turning them on one at a time would probably prevent the inverter from going into over current limit, or at least maybe the last 5 strings should be turned on one at a time.

They also sell load cones that have nichrome wire wound around a ceramic cone.  They screw into a standard 120vac light bulb base.  The wattage varies up to 1000 watts each, but that's at 120v volts, so each one may only be about 100 watts at 12v, maybe 300 watts or so at 36 volts.  They are made to take the heat even when they get red hot, which is at 120vac, so they can run for hours but they will heat up the room fast because after all they will be putting out a lot of heat.  A typical room electric space heater is 1000 watts or a little more, so go from there.

[Brumby]

Simple answer is Yes.

If you want to test 36V @ 30A, then you are talking 1.08KW.  That's a lot of light.

If you want to use the classic 21W/5W automotive stop/tail light (using both filaments) - you will need

42

of them ... 14 parallel strings of 3 globes in series (with the filaments of each globe in parallel).  The numbers put that at 1.092KW - but with a few losses thrown in, it should be pretty close to your target.

[BrianHG]

I would use 11 * (3 halogen spot light 12v 50 watt bulbs in serial).  Dirt cheap, BUT, having 33 bulbs on all in 1 area will be like a space heater...

The only annoying thing is wiring them without sockets is a pain because wire wont solder to the pins too well...

[amyk]

I would use 11 * (3 halogen spot light 12v 50 watt bulbs in serial).  Dirt cheap, BUT, having 33 bulbs on all in 1 area will be like a space heater...

You could put them outside around the house...

[MrAl]

Hi,

That would be one heck of a night light :-)

Biggest bulb i had in the past was a 300 watter.  I used it mostly as a load for testing stuff.

[BravoV]

Thank you all for the replies !   

Also appreciation to @amyk for the heat calculation, now I have a picture how much water that I'm going to deal with, and also MrAl for the incand bulb cold resistance, must think of sequential turn on at the arrays of bulbs in parallel. 

As using cooled wire in water, this need to be fully attended even with plenty of water, and also I need to find a suitable place to test it, no way going to turn my little puny lab into steam bathroom.

If I'm going to use water as Mike does, will copper wire from unwind from broken transformer is enough ?

Have to find one that suitable for shorting it in 36 Volt @ +30 Amps ...

[MrAl]

Hi again,

Watch out when you do the load immersion in water thing.  If the surface area of the part being used for the load is too small it can cause local heating even in the water.  This will cause the water to boil near the location of the wire even though the rest of the water is not boiling.  Depends highly on how much power is present over the given surface area.  Lowering the current will stop it.

Power resistors work pretty well, but there could be problems if it is pushed too far because of internal hot spots.  A check on the resistance would tell if the resistance is changing too much as it heats up.  Keeping it within reason would be wise.

[alm]

Some stirring should help against local hotspots, though of course the thermal resistance from resistive element to water will ultimately limit the maximum power dissipated by the resistor.

[Brumby]

I have a 27R load resistor.  Lots of power dissipation capacity, especially in water.

Same model as this:

[Kjelt]

I would use 11 * (3 halogen spot light 12v 50 watt bulbs in serial).  Dirt cheap, BUT, having 33 bulbs on all in 1 area will be like a space heater...
The only annoying thing is wiring them without sockets is a pain because wire wont solder to the pins too well...

Ah but they are the exact fit for a what we in Holland call a kroonsteen (crownstone) and the Germans call Europaklemmen (european clamps) but I do not know or can find the english term.
They are these rows of plastick 1:1 connectors for mains with a small screw inside and you break off the amount you need, they are in multiple sizes and distances, some do fit these bulbs quite nicely.

Due to the heat you do need the porcelain version as in the second picture.

[Ian.M]

2x parallel 230V 18" 1KW ceramic wirewound bar fire elements, with 6 equally spaced narrow stainless Jubilee clips on each wired alternately to either end of the element to convert each to 7 paralleled sections of approx 1/7 the resistance would also be in the ball-park for a 36V 46A load.   Use heat resistant, or even bare wires, bolted to a hole drilled in the tail of each clip and alternate the tail directions so the paralleling wires can be on opposite sides of the element.    The connections are chunkier and the elements less fragile so its probably easier than trying to hook up 33 50W halogens without the proper sockets.   Build it on a shiny metal sheet and you only need high temperature insulation for the element support brackets - through bolt ordinary steel L brackets with over-size holes, on mica sheet with mica washers (+ a steel washer on top) under the bolt heads and pack between the bolt and the oversize hole with natural fiber string (which may carbonise) or glassfibre string to keep the bolt centered.  You've still got that pesky 1.67KW of heat output, but at least its dry radiant heat, not steam so if you can point it out of an open window, you can run your load all day.

Adjust one bar at a time with the other out of circuit so you are running somewhere near 50% load. If the load is too high, disconnect one end terminal and if still too high, move the last clip towards the disconnected end. If its too low close up the Jubilee clip spacing towards one end till the hot resistance is slightly too low (measure it by powering 4 sections and measure the current into the center pair), then move the last clip back towards the other end to fine trim it.   

[Delta]

"Ah but they are the exact fit for a what we in Holland call a kroonsteen (crownstone) and the Germans call Europaklemmen (european clamps) but I do not know or can find the english term."

The British sparky slang terms for those is "choccie blocks", as they look like a chocolate bar.

[Ian.M]

Yes but tightening a porcelain screw terminal block on the pins of a glass pinch GU5.3 base risks cracking the bulb and with 33 to do the odds are against you. 

If one wanted to go down the bulb route, personally, I'd pick up a job lot of cheap pigtailed GU5.3 sockets off EBAY, mount them all on a 7.5 cm grid to a polished metal baseplate (must be polished because dichroic reflector bulbs put most of the heat out the back), and wire them all in series at the back with each node brought out to one way of a porcelain screw terminal block (or a regular one on a thermally insulating standoff).  By daisy chain (up to the wire ampacity) jumpering 1 of N ways of the terminal block alternately to either supply terminal, you can reconfigure it as parallel strings of N bulbs to suit any voltage that's multiple of 12V, or leave part of the array only connected at one end to reduce the total load.

[BravoV]

2x parallel 230V 18" 1KW ceramic wirewound bar fire elements, with 6 equally spaced narrow stainless Jubilee clips on each wired alternately to either end of the element to convert each to 7 paralleled sections of approx 1/7 the resistance would also be in the ball-park for a 36V 46A load.   Use heat resistant, or even bare wires, bolted to a hole drilled in the tail of each clip and alternate the tail directions so the paralleling wires can be on opposite sides of the element.    The connections are chunkier and the elements less fragile so its probably easier than trying to hook up 33 50W halogens without the proper sockets.   Build it on a shiny metal sheet and you only need high temperature insulation for the element support brackets - through bolt ordinary steel L brackets with over-size holes, on mica sheet with mica washers (+ a steel washer on top) under the bolt heads and pack between the bolt and the oversize hole with natural fiber string (which may carbonise) or glassfibre string to keep the bolt centered.  You've still got that pesky 1.67KW of heat output, but at least its dry radiant heat, not steam so if you can point it out of an open window, you can run your load all day.

Adjust one bar at a time with the other out of circuit so you are running somewhere near 50% load. If the load is too high, disconnect one end terminal and if still too high, move the last clip towards the disconnected end. If its too low close up the Jubilee clip spacing towards one end till the hot resistance is slightly too low (measure it by powering 4 sections and measure the current into the center pair), then move the last clip back towards the other end to fine trim it.

Thank you, you've made me think of heating element for iron.

Here locally, the replacement part for iron are abundant and cheap, most are rated about 300 Watt at 220 Volt, and usually priced just about $ 0.5 per piece. Fyi, here mains is 220 Volt.

They look very similar to this, and comes with different variety of connection terminals.

(random photo taken from Google search)


Of course they will be submerged into the water or flowing water, I've been thinking to use a big bucket and just put the water hose in it and let the water run during the test period.

Now, more questions :

1. Can I use these in 5 or maybe 6 of in parallel ?

2. Safety concern, as I'm going to do this in a wet environment, lets say I've secured and isolated "ALL" mains 220 Volt wiring at this power brick, "AND" with earth connection at the brick is also properly done. Also the unit is relatively far from the water tank. Is it safe to expose the 36 Volt DC ? With the possibilities that my body may get connected accidentally with it, either directly or thru the water.

3. Do these type of iron heater element has similarity to incandescent bulb at the cold and hot resistance as Mr.Al mentioned ?


Also just fyi, this is the specification of the converter :

[Beamin]

Since incandescent bulbs have a positive temperature coefficient, I would expect this to work fairly well as long as you use lamps with the same voltage and current ratings.

An alternative might be this (although 6 hours might require a very large bucket or replacing the water multiple times):

AS soon as I read the first post this was exactly what came to mind: An English voice saying "Right..." then a big bucket full of water and a jumble of wire. Stop reading my thoughts.

[Beamin]

36V * 46.5A = 1674W = 1.674kJ/s which is coincidentally almost exactly 0.4 of the specific heat constant of water, 4.184kJ/kg/K. It'll raise the temperature of 1L of water 0.4 degrees per second, or 200 seconds for an 80 degree rise (assuming 20C to boiling). 10L will take 2000s, 100L will take 20000s (5:33:20). Alternatively, it will take 6h to boil 108L of water at that power.

Won't the 200 second rise just bring the water to 100'C but to get to boiling you need to put in much more energy (8 or 10 times IIRC) as its a state change to go from 100'C water to 100'C steam? Would be a fun experiment to try.

[Ian.M]

A1. The iron's hot element resistance is about 160R.  The outer sheath surrounds the mineral powder insulation and is normally grounded, so no you cant tap the element part way to parallel section of it.  You'd need about 200 elements paralleled for your load, so unless a pallet of them falls of the back of a lorry in front of you, they are no good to you.  If you do have a pallet-full they'd dissipate about 8W each @36V, probably low enough for fan assisted air cooling.

A2. Its under 50V, with a floating output and 3KV RMS isolation between its primary secondary and secondary sides, so it *probably* can be categorised as SELV, and is fairly safe.  However if you ground either side of the output it becomes far more dangerous - consider the scenario of you tripping and ending up on your knees in a puddle with your arm in the tank. 36V DC diagonally across you body with large area abrasions in contact with wet conductive surfaces.  As little as 42V DC can be lethal with enough low resistance contact area so I'd treat a wet area 36V system with *EXTREME*  respect.

A3. Got an iron in the house?  Measure its cold element resistance (pin to pin on the mains plug) then calculate its nominal hot resistance from its rated wattage and supply voltage.
I would expect some change but far less than the 10:1 ratio typical of incandescent bulbs.