Why not use a SOX balast to drive flurescent lamps without the need for starters

[Deactivated-1]

Wouldn't a SOX ballast circuit create high enough voltage to start a mercury vapor GDT without heating the electrodes? it would not require a starter, or a heater anymore, probably would increase lifespan of the tube by a lot...

[jonpaul]

Bonjour, we worked in lighting and ballasts (magnetic and electronic ) since 1970s and well familiar to SOX/LPS, fluro, HPS.

The LPS are totally different specs for V, I OCV etc than fluorescent.

Generally LPS/SOX are LESS efficient than fluro.

Chek the lamp current spec.

Finally there are may types of SOX wattage and linear/single ended as well as fluro (T5/T8, length) so ballast specs differ.

See Lamps and Lighting book and Electric Discharge lamps by Dr John WEYMOUTH


Jon

[Gyro]

Wouldn't a SOX ballast circuit create high enough voltage to start a mercury vapor GDT without heating the electrodes? it would not require a starter, or a heater anymore, probably would increase lifespan of the tube by a lot...

What makes you think that it would increase the lifespan of the tube?

Mercury fluorescent lamps are hot cathode devices - the heaters are actually directly heated cathodes. As the tube works on AC, it needs a cathode at each end. There is also a surrounding strip of metal welded on at each end that act as the anodes. The starter is there to get the cathodes up to operating (electron emitting) temperature before the discharge is initiated. Once the discharge is established, the discharge current and cathode bombardment keeps them up at the same operating temperature.

If you start a Mercury fluorescent tube cold, emissive material is stripped off the cathodes and deposited on the glass, blackening the ends of the tube. This damages their electron emitting capability and considerably shortens the life of the tube. This can be seen on tubes used with cheap inverters, where the pins at each end are commoned, rather than given a heater supply.

Other types of lamp (including CCFLs) are cold cathode devices, with uncoated solid Tungsten electrodes. It's a different technology.

[Gyro]

Unfortunately you don't get the choice, the filament cathodes will heat up from the discharge and take preference over the 'cold cathodes' (anode strips) in electron emission, so you will still get the end blackening at startup. The tube that you're describing is actually an LCD backlight CCFL, which have cold cathodes and high brightness, with high frequency drive.

Mercury fluorecents do contain inert gas - Krypton for room temperature tubes and Argon for low temperature tubes, to initiate the discharge until the Mercury is fully vaporized (which happens very quickly). I believe SOX lamps include Neon in the mixture which accounts for the redish colour until they are warmed up - the Sodium takes much longer to vaporize as it needs a higher temperature.

Inductive compact fluorescent lamps do did / do actually exist, I think Philips made them. They were expensive though, and although there are no electrodes to fail, the phosphors still degrade in efficiency over time. They were a bit of a loss leader. The only other place that I can think of at the moment that uses inductive coupling is the Rubidium lamps in atomic clocks, where the ligh emission is narrow band from vapour itself rather than a phosphor.


P.S. I just looked at my 1927 Electrical Engineering book. Cold cathode mercury vapour lamps did actially exist back then (without phosphor), for industrial applications. The colour rendition was (predictably) very poor and the efficiency low.

[themadhippy]

I guess what I'm asking is why have fluorescent tubes when you could have a mercury vapor lamps that are phosphor coated

you are aware the posh name for a fluorescent tube is low pressure mercury vapour discharge lamp?

[Gyro]

well wouldn't making an inductive fluorescent lamp fixture before the invention of the first vacuum tubes nearly impossible? and even with tubes it would be really impractical.
I meant it as in why are fluorescent tubes as they are. of course nowadays making complex driving circuitry isn't a big problem, but back in the day it seemed to me that the circuit was quite overcomplicated for the fluorescent tube.

I guess what I'm asking is why have fluorescent tubes when you could have a mercury vapor lamps that are phosphor coated.

speaking of inductive and SOX lamps... were there any inductive sodium lamps made?

It doesn't pre-date the vacuum tube - the triode was 1907. People were playing with sparks, arcs and discharges in low pressure gasses well before that.

There are high pressure Mercury vapour lamps that have inner quartz envelopes (you need quartz to pass the UV to the phosphor... and the very high temperature and pressure) and an outer glass bulb coated on the inside surface with phosphor to improve the colour rendition, Philips MBF is an example. They are self starting with a standard inductive ballast - they have an auxiliary electrode close to one of the main electrodes, in series with a rod PTC thermistor. The High pressure makes them more efficient than low pressure Mercury fluorescents, but the colour rendition isn't as good and the light is much more intense out of a small bulb area, not as friendly as the spread of light of a florescent tube.

I've attached a couple of pages from my 1927 The 'Modern Electrical Engineer' book anyway. There isn't much detail, but it covers efficiency, colour rendition and has a picture.


P.S. I don't have any idea on Sodium inductive lamps (I doubt it), but I wouldn't be surprised if there have been others (like the Rubidium) for engineering or scientific purposes.

[james_s]

It does work, SOX lamps are actually electrically much more similar to fluorescent lamps than to HID. In fact due to the rarity of SOX in this part of the world, I've mostly run them on instant start T8 fluorescent ballasts, they run very nicely on this arrangement, an Advance electronic T8 twin lamp ballast will run a 35-55W SOX lamp very well using both outputs in parallel.

There are a couple of very good reasons NOT to use a SOX ballast on fluorescent tubes. First is that you get a large penalty in lamp life by instant starting a fluorescent lamp without preheating the cathodes. It blows a bit of the oxide off the cathode every time so instant starting is typically only used for lamps that are turned on only once each day.

The other reason is that SOX ballasts are traditionally leak autotransformers which are very inefficient. It's one reason they never achieved widespread adoption in North America, the lamp efficiency is very high but unless you have enough voltage to get away with a simple choke ballast the system efficiency is a lot less impressive.

[mikeselectricstuff]

Why would anyone use a fluorescent tube nowadays ?

[james_s]

Why would anyone use a fluorescent tube nowadays ?

If they already have them in service is the biggest reason. Actually the efficiency of a lot of the linear fluorescent tubes is pretty reasonable, and in my experience they're more reliable than a lot of the LED replacements. The F32T8 tubes in my garage and laundry room are some of the only non-LED lighting in my house and they're still the same lamps I installed 18 years ago when I bought the house.

[LaserSteve]

An "Instant Start" ballast is common in USA. It is similar to a SOX ballast, having a high peak voltage across the lamp for starting.

Go to lamptech.co.uk for some serious reading and lamp parameters.  It should make sense at that point.
Hint, look at the difference in operating current.between SOX and FL with hot cathode.

I spent a lot of time a few months ago working on a one-off SOX ballast for the lab at work. James does know what he's talking about.

Lamptech is one of my favorite web sites for a reason.

Steve

[james_s]

you mean those electronic "ballasts"? no I am not talking about that, but the simple ballast choke with a tap connected to a capacitor arrangement that creates high voltage to start the lamp, when it starts it shunts that circuit.

No, the're called "slimline" tubes and have a single large pin on each end. The most common size is 8', they were widely used in workshops and commercial installations. They used a magnetic autotransformer ballast not entirely unlike those used for SOX lamps.

Aside from the fact that fluorescent lamps are becoming obsolete and magnetic ballasts have been outlawed in parts of the world, what are you trying to accomplish here? Starting fluorescent lamps without supplying cathode heat is detrimental to their life.

[Wolfram]

you mean those electronic "ballasts"? no I am not talking about that, but the simple ballast choke with a tap connected to a capacitor arrangement that creates high voltage to start the lamp, when it starts it shunts that circuit.

No, the're called "slimline" tubes and have a single large pin on each end. The most common size is 8', they were widely used in workshops and commercial installations. They used a magnetic autotransformer ballast not entirely unlike those used for SOX lamps.

Aside from the fact that fluorescent lamps are becoming obsolete and magnetic ballasts have been outlawed in parts of the world, what are you trying to accomplish here? Starting fluorescent lamps without supplying cathode heat is detrimental to their life.

I'm just trying to find a reason that fluorescent lamps didn't use a circuit like that.

Some did, most didn't. Many different circuits were used, and the ones that ended up dominating were the ones that provided the best balance between cost, lamp life and efficiency.

A hot cathode has a much lower voltage drop than a cold cathode, something like 10 V compared to 50+ V of a cold cathode with an emissive coating. In a tube with an arc voltage of 100 V or less, cold cathodes would reduce the efficiency dramatically. They were used in tubes with higher voltage drop across the positive coloumn, where the added voltage drop had less of an impact on efficiency. Higher voltage drop with acceptable efficiency means thin long tubes, like CCFLs used for LCD backlighting and flatbed scanners, and cold cathode lighting installations, but the increased cost of the control gear and installation made them uncompetitive for general lighting.

Cold cathodes do have increased life span, mainly because the increased cathode area allows for more emissive coating to be used. This effect is not extremely dramatic, as this only addresses one of the failure modes of fluorescent lamps. Extra control gear losses from stepping up the voltage also reduce the system efficiency. The cost of a lamp is often much smaller than the cost of the energy it consumes during its lifetime, so a lowest-cost balance would favor giving up some lifetime for efficiency, even when accounting for the cost and labor of replacing lamps.

[MrAl]

Wouldn't a SOX ballast circuit create high enough voltage to start a mercury vapor GDT without heating the electrodes? it would not require a starter, or a heater anymore, probably would increase lifespan of the tube by a lot...

Isnt there a different way to start the lamps without heating the electrodes already?  I think they use a high voltage spike.  Once the lamp starts, the spike goes away of course and the lamp runs normally.

[Gyro]

Isnt there a different way to start the lamps without heating the electrodes already?  I think they use a high voltage spike.  Once the lamp starts, the spike goes away of course and the lamp runs normally.

... and so the loop back to the OP is complete, repeat until dizzy! 

[james_s]

Isnt there a different way to start the lamps without heating the electrodes already?  I think they use a high voltage spike.  Once the lamp starts, the spike goes away of course and the lamp runs normally.

Yes, instant start ballasts have existed for many years, they don't use a spike, they just have a high OCV and in the case of magnetic ballasts they used a leaky core to limit the current. Once the lamp strikes the voltage sags to limit the current. The disadvantage is that lamp life is reduced.

[Gyro]

For those still using fluorescent tubes, the readily available solid state drop in replacement starters are the way to go. You get optimal pre-heating and ignition with non of the flickering associated with the old gas filled bimetal electrode starters.

The 8ft (heated) tube in our garage has been short cycled on trips to the freezer for getting on for two decades with zero end blackening using one of these. Less so these days as I've finally got around to fitting a (feeble 20W) PIR LED ceiling light for those quick trips. There's zero need for fluorescent tube cold starting.

[james_s]

For those still using fluorescent tubes, the readily available solid state drop in replacement starters are the way to go. You get optimal pre-heating and ignition with non of the flickering associated with the old gas filled bimetal electrode starters.

The 8ft (heated) tube in our garage has been short cycled on trips to the freezer for getting on for two decades with zero end blackening using one of these. Less so these days as I've finally got around to fitting a (feeble 20W) PIR LED ceiling light for those quick trips. There's zero need for fluorescent tube cold starting.

That depends on where you are. Here in 120V land everything 3' and longer uses an autotransformer ballast, those have been "rapid start" since around the mid 70s with separate filament windings and no starter. For smaller tubes a choke ballast and starter were common but now electronic ballasts are mandated for everything. Personally I've always liked the old preheat start ballasts, they ran the lamps harder than the later rapid start stuff so the old fixtures were noticeably brighter.

[MrAl]

These days it is probably better to use LED lighting of some sort.  Might be more expensive but probably worth it.
I used to use fluorescent lights too long time ago.  I went to LED for everything.

One catch though is that some LED lights are not as bright as what you replace.  A good example of this is those high power halogen bulbs.  It's hard to get an LED light that can do that job and if you buy something off the shelf that says it works you find out it's not nearly as bright.

[westfw]

I haven't seen a fluorescent lamp with a "starter" in a long time, despite having "many" fluorescent lamps with "magnetic" ballasts.  Maybe they just drive the filament all the time?

[james_s]

I haven't seen a fluorescent lamp with a "starter" in a long time, despite having "many" fluorescent lamps with "magnetic" ballasts.  Maybe they just drive the filament all the time?

In 120V regions a "rapid start" ballast is used for everything over 24", this has separate winding taps to provide cathode heat. Lamps 24" and under with magnetic ballasts typically do have a starter though. In UK/Europe where 240V is the norm simple choke ballasts with starters are the norm. They have 5' tubes there too which look weird to me, it's not a size that was ever used over here outside of specialty sign lamps.