Was there ever a HPS retrofit bulb made for the 100W mercury vapor ballasts?

Don't know about the US 100W rating, but we did have this rather interesting example which worked on either 80W or 125W ballasts 

Don't know about the US 100W rating, but we did have this rather interesting example which worked on either 80W or 125W ballasts 

Because the 0.9..0.95A ballast is between the 0.8A and 1.15A this retrofit lamp is designed for, it should work just fine...

I guess I’ll have to look for one of those then...     Thanks guys! 

There were none that I recall.. The 100W MVs generally were not too common for street lighting except for here in the northeast US, so perhaps the lamp manufacturers did not bother owing to a potentially limited market for any such retrofit lamp.

Yeah couldn't find any for H38 ballasts in the old Sylvania and Eye literature. Eye did make a complete line of HPS retrofit lamps for European mercury lamps down to 50w though. See here

The Eye “Sunlux Ultra Ace” retrofits for 80W or 125W mercury should also work just fine.  They are also lamps based on the unsaturated vapour principle like the Sylvania SPX EcoArc, and as such function stably over quite a wide range of ballast currents.  However their regular HPS retrofits with saturated vapour will not work reliably at 100W.

The Eye “Sunlux Ultra Ace” retrofits for 80W or 125W mercury should also work just fine.  They are also lamps based on the unsaturated vapour principle like the Sylvania SPX EcoArc, and as such function stably over quite a wide range of ballast currents.  However their regular HPS retrofits with saturated vapour will not work reliably at 100W.

What about using a typical European 70w internal starter high pressure sodium lamp designed for 90v 1.0a operation? I seem to find these lamps working fine on North American 100w H38 mercury vapor HX ballasts.

I also see that the Philips SON-H 68w and Osram NAV-E plug in 68w high pressure sodium lamps for European 80w mercury vapor ballasts might be a close match for operation on a North American 100w H38 HX ballast since these lamps run at 125v 0.85a while North American 100w H38 mercury vapor lamps run at 130v 0.85a.

That is quite dangerous, even if the lamp does strike and run up, its voltage is so low that it causes the current to increase significantly.  This leads to overheating of the ballast windings, followed by deterioration of its insulation and eventual failure of the ballast.  Some years ago Philips launched an energy-saving ceramic metal halide lamp to work on mercury 80W and 125W ballasts and they achieved that by reducing the arc voltage from 115V to 105V (lamp type CDO-H).  However even that small decrease started causing massive problems in the field of ballasts burning out, and the lamps were quickly withdrawn from sale.  With a 90V sodium arc tube it would be similarly bad (although current increase is not quite so high because the sodium arc has a very different power factor than mercury or metal halide arcs).

That is quite dangerous, even if the lamp does strike and run up, its voltage is so low that it causes the current to increase significantly.  This leads to overheating of the ballast windings, followed by deterioration of its insulation and eventual failure of the ballast.  Some years ago Philips launched an energy-saving ceramic metal halide lamp to work on mercury 80W and 125W ballasts and they achieved that by reducing the arc voltage from 115V to 105V (lamp type CDO-H).  However even that small decrease started causing massive problems in the field of ballasts burning out, and the lamps were quickly withdrawn from sale.  With a 90V sodium arc tube it would be similarly bad (although current increase is not quite so high because the sodium arc has a very different power factor than mercury or metal halide arcs).

The significant factor here is the vast majority of MV ballasts are on many decades old fixtures, so long past their rated life (usually 100k hours for these devices), so are just literally "hanging on their last thread". Then even a small disruption, in the form of either higher voltage peaks from the various glowbottle or FEC ignitors, or from the little extra power dissipation (the extra 5% on current may seems small, but it is 10% of power dissipation, so often 10degC higher temperature), is then enough to make that "last thread" to finally give up.

I was once told by a lighting engineer that a MV lamp can leave an electrical fingerprint within the ballast (so to speak) after over decades of use as both operating together for a very long time. Then after installing a new lamp the ballast can fail owing to whatever differences it needs to sustain the new lamp. So yes, a weak, tired link can be broken (or maybe the same with a very used capacitor?).

I was once told by a lighting engineer that a MV lamp can leave an electrical fingerprint within the ballast (so to speak) after over decades of use as both operating together for a very long time. Then after installing a new lamp the ballast can fail owing to whatever differences it needs to sustain the new lamp. So yes, a weak, tired link can be broken (or maybe the same with a very used capacitor?).

That's an interesting concept, I haven't come across it before but do recognise the phenomenon that an old ballast will sometimes burn out shortly after having a new lamp installed.  In India that was often a problem, where magnetic ballast lives are atrocious and sometimes don't even outlast one lamp (due in part to the huge variations in mains voltage as well as generally poorer quality construction).   I don't know if the reason for the sudden failures after re-lamping might be due to mechanical effects, e.g. too much vibration or shock on an old ballast whose insulation around the windings has become brittle.  Or maybe just the sudden change in lamp voltage.  Mercury lamps are characterised by rising voltage during life, which leads to reduced loading on their ballasts as they age.  However by installing a fresh new lamp with lower voltage arc, this might be enough to increase the temperature of an already aged ballast just enough to push it over the edge to failure.

An interesting theory, fascinating reading, thanks guys 

Additionally, I have recently found out that the Japanese company known as Hitachi had developed some 85 watt high pressure sodium retrofit lamps that were intended to replace Japanese 100W mercury vapor lamps on HX ballasts and series reactor ballasts, but those are very hard to find as well.

After I found that the Japanese 100W mercury vapor lamps seem to run well on North American 100W H38 mercury vapor HX ballasts, it seems like these Hitachi 85W high pressure sodium retrofit lamps might do ok.

For reference, Japanese 100W mercury vapor lamps run at 115V, 1.0A while North American 100W H38 mercury vapor lamps run at 130V, 0.85A.

While the North American 100W H38 mercury vapor HX ballast drives the Japanese 100W mercury vapor lamps properly, the ballast draws somewhat more current than it otherwise would with the correct North American 100W H38 mercury vapor lamps whenever it is supplied with the correct line voltage and frequency.

That is quite dangerous, even if the lamp does strike and run up, its voltage is so low that it causes the current to increase significantly.  This leads to overheating of the ballast windings, followed by deterioration of its insulation and eventual failure of the ballast.  Some years ago Philips launched an energy-saving ceramic metal halide lamp to work on mercury 80W and 125W ballasts and they achieved that by reducing the arc voltage from 115V to 105V (lamp type CDO-H).  However even that small decrease started causing massive problems in the field of ballasts burning out, and the lamps were quickly withdrawn from sale.  With a 90V sodium arc tube it would be similarly bad (although current increase is not quite so high because the sodium arc has a very different power factor than mercury or metal halide arcs).

Here is a collector who has had some experiences with running European self-starting 70W high pressure sodium lamps on North American 100W H38 mercury vapor HX ballasts:

https://www.lighting-gallery.net/gallery/displayimage.php?album=search&cat=0&pos=6&pid=74928