Why are old fluorescent chokes/transformers rated at 118 Volts? They could do it at 110, 115, 120, and 125 Volts. But why 118?

My guess:
"One size fits all"
The middle of the 110..125V range is 117.5V, so rounded it becomes the 118V...

My guess:
"One size fits all"
The middle of the 110..125V range is 117.5V, so rounded it becomes the 118V...

That's a good idea for them to use the average, but they could have written 110-125 on there, and it would be less confusing.

I've seen some old electrical devices rated as 118, and someone overseas has a GE ballast made in USA, likewise rated for 236. Fun fact. Old MV ballasts had various taps ranging from 100-130 for areas with slightly different voltages.

I've seen some old electrical devices rated as 118, and someone overseas has a GE ballast made in USA, likewise rated for 236. Fun fact. Old MV ballasts had various taps ranging from 100-130 for areas with slightly different voltages.

That's really strange!
And the voltages here like to change a lot, so the taps would be meaningless.

I guess this is why lots of electronic charger adapters are rated for 100-250v, both for voltage fluctuations and used in different regions. How does your power change a lot?

I guess this is why lots of electronic charger adapters are rated for 100-250v, both for voltage fluctuations and used in different regions. How does your power change a lot?

It manly changes when high load items come on, I found that a 300 Watt incandescent lamp drops it by a couple of volts.

Is your service 200a? And how long of a run is your wire from the meter to transformer?

Is your service 200a? And how long of a run is your wire from the meter to transformer?

Yes, 200 Amps, and it's not very far, maybe 25 feet?

The taps still make sense, you'd choose the one that matches the average voltage, so that the choice would be as close as possible to the actual voltage at any time

The adapters are universal voltage not as much "for" something, but "because" of something :

 - "Linear" power regulators are basically some impedance that is put there, that does not "think" for itself. It is chosen to get the lamp current (ballast), output voltage (transformer) and such that you want, assuming that it is getting the right voltage at the input. If the input voltage varies, so will the output, there is not much the device can do

 - "Switching" power regulators are capable to put out more voltage and current than required, and there is circuitry that permanently measures the resulting output voltage/current, and all the time turns down the power conversion just before the wanted output value is exceeded. If the input voltage is high, the cotrol circuitry will regulate the conversion to get the wanted output voltage/current. If the input voltage is low, the control circuit will still regulate the conversion so that exactly the same output voltage/current is reached. As the circuit can control itself completely, there naturally is no precise requirement to the input voltage, except some absolute minimum below which the circuit can't deliver the required power even when working at full capacity, and some absolute maximum that is the limit of what components can handle. Provided that those circuits work fine allready at few 10's V, and the components limitations are around 400V DC (so about 280V AC), the rated 100..250V is quite conservative part of this range, not too close to the edges

The taps still make sense, you'd choose the one that matches the average voltage, so that the choice would be as close as possible to the actual voltage at any time

The adapters are universal voltage not as much "for" something, but "because" of something :

 - "Linear" power regulators are basically some impedance that is put there, that does not "think" for itself. It is chosen to get the lamp current (ballast), output voltage (transformer) and such that you want, assuming that it is getting the right voltage at the input. If the input voltage varies, so will the output, there is not much the device can do

 - "Switching" power regulators are capable to put out more voltage and current than required, and there is circuitry that permanently measures the resulting output voltage/current, and all the time turns down the power conversion just before the wanted output value is exceeded. If the input voltage is high, the cotrol circuitry will regulate the conversion to get the wanted output voltage/current. If the input voltage is low, the control circuit will still regulate the conversion so that exactly the same output voltage/current is reached. As the circuit can control itself completely, there naturally is no precise requirement to the input voltage, except some absolute minimum below which the circuit can't deliver the required power even when working at full capacity, and some absolute maximum that is the limit of what components can handle. Provided that those circuits work fine allready at few 10's V, and the components limitations are around 400V DC (so about 280V AC), the rated 100..250V is quite conservative part of this range, not too close to the edges

So you could run a 100-240 volt switching adapter on 24 volts AC?

Actually it might start on 24V, but would probably not be able to supply its rated output power. At 40V it might allready supply something usefull, and at 60V might even already reach full power

Also, you can power most of them with DC straight to the "AC input", they dont care as the 1st thing that happens inside is rectification of the AC to DC anyway. But the rectifier output is the peak voltage, so for example if you want to supply DC that would be equialent to 24V AC, the DC voltage would be 34V and not 24V

(the DC input thing won't work with power supplies that have 120/240V selection switch or things like higher power Electronic ballasts, as they use voltage doubler inside, and that really requires AC to work. But all the 100..240 rated stuff does not have doubler)

Actually it might start on 24V, but would probably not be able to supply its rated output power. At 40V it might allready supply something usefull, and at 60V might even already reach full power

Also, you can power most of them with DC straight to the "AC input", they dont care as the 1st thing that happens inside is rectification of the AC to DC anyway. But the rectifier output is the peak voltage, so for example if you want to supply DC that would be equialent to 24V AC, the DC voltage would be 34V and not 24V

(the DC input thing won't work with power supplies that have 120/240V selection switch or things like higher power Electronic ballasts, as they use voltage doubler inside, and that really requires AC to work. But all the 100..240 rated stuff does not have doubler)

My Lumatek ballast is 120/240 volts, but has no switch for it.

Electronic ballasts are usually required to have high power factor. And the only way to achieve really high power factor is a boost type power factor correction circuit. And if such circuit is there, it is intrinsically capable to boos from virtually any voltage on the input. Well, it requires the components to be right sized for the related current. So with real devices in the power range below few 100's W it is usually comm,on practice to size the booster components so, it operates "world-wide", so from 100V till 240 or even 277V (usually the output of that boost is just 400VDC).

However there could be problem mainly with mains power boxes with low StandBy power requirements (phone chargers, notebook supplies, PC's, TV's,...): Although they have on their front end just a rectifier (and maybe the boost PFC stage), they do not tolerate DC supply at all.
The reason is linked to the need for a combination of
Rather larger input EMC filter capacitors
And the safety requirement to have the voltage at the input prongs below safe level within few seconds upon disconnection
And consuming low stand by power.

The thing is, when you have large capacitors, which have to be discharged pretty fast, you need rather high currents for that. Traditionally that is done by just plain bleeder resistors. But when the capacitors are large, the resistance needs to be lower. And that resistance consumes power even when the downstream electronic still consumes nothing at all. Because the StandBy requirements go stricter, the power dissipation here does counts for the largest part of it, or with some cases it may even exceed the StandBy limits alone.
So a trick was invented: These bleeders are connected in series with a switch circuit, which senses the voltage and when it sees the polarity changing, it means the circuit is connected to the mains, so the resistors are kept disconnected to not consume any power. When you disconnect the plug, the charged capacitors retain just a DC voltage. And when the circuit senses the voltage does not change, it turns ON the discharging resistors. This trick then allows to overcome the power dissipation and still provide the required discharge within the required time.
Because this detection needs some time to distinguish the mains connection, the actual discharge has to happen faster, so the resistors have to be of lower value, so that means they dissipate more power. Plus because this trick allows the use of larger capacitors without the idle power consequences, the EMC design goes more frequently for larger capacitances instead of the more expensive inductors, which then increases the power requirements for the  discharge resistors.
With normal designs, these are rated to just consume the energy in the capacitors, because normally they are disconnected, so are usually rather small even when the discharge currents are rather high. With AC supply there is no problem.

But the problem arises, once you power that by DC: The discharge circuit sees just constant voltage, so it misinterprets it as the plug was disconnected and there is just the remaining charge, so it turns ON the discharge resistors. But there is not just the charged capacitors, so these resistors heat up for rather long time and usually overheat.

Electronic ballasts are usually required to have high power factor. And the only way to achieve really high power factor is a boost type power factor correction circuit. And if such circuit is there, it is intrinsically capable to boos from virtually any voltage on the input. Well, it requires the components to be right sized for the related current. So with real devices in the power range below few 100's W it is usually comm,on practice to size the booster components so, it operates "world-wide", so from 100V till 240 or even 277V (usually the output of that boost is just 400VDC).

However there could be problem mainly with mains power boxes with low StandBy power requirements (phone chargers, notebook supplies, PC's, TV's,...): Although they have on their front end just a rectifier (and maybe the boost PFC stage), they do not tolerate DC supply at all.
The reason is linked to the need for a combination of
Rather larger input EMC filter capacitors
And the safety requirement to have the voltage at the input prongs below safe level within few seconds upon disconnection
And consuming low stand by power.

The thing is, when you have large capacitors, which have to be discharged pretty fast, you need rather high currents for that. Traditionally that is done by just plain bleeder resistors. But when the capacitors are large, the resistance needs to be lower. And that resistance consumes power even when the downstream electronic still consumes nothing at all. Because the StandBy requirements go stricter, the power dissipation here does counts for the largest part of it, or with some cases it may even exceed the StandBy limits alone.
So a trick was invented: These bleeders are connected in series with a switch circuit, which senses the voltage and when it sees the polarity changing, it means the circuit is connected to the mains, so the resistors are kept disconnected to not consume any power. When you disconnect the plug, the charged capacitors retain just a DC voltage. And when the circuit senses the voltage does not change, it turns ON the discharging resistors. This trick then allows to overcome the power dissipation and still provide the required discharge within the required time.
Because this detection needs some time to distinguish the mains connection, the actual discharge has to happen faster, so the resistors have to be of lower value, so that means they dissipate more power. Plus because this trick allows the use of larger capacitors without the idle power consequences, the EMC design goes more frequently for larger capacitances instead of the more expensive inductors, which then increases the power requirements for the  discharge resistors.
With normal designs, these are rated to just consume the energy in the capacitors, because normally they are disconnected, so are usually rather small even when the discharge currents are rather high. With AC supply there is no problem.

But the problem arises, once you power that by DC: The discharge circuit sees just constant voltage, so it misinterprets it as the plug was disconnected and there is just the remaining charge, so it turns ON the discharge resistors. But there is not just the charged capacitors, so these resistors heat up for rather long time and usually overheat.

That sounds like a fire hazard, but why not incorporate resistors into the thermal design?