I have an advance 71A2800 ballast that is overheating to the point where it can boil water. I noticed that the fixture was hot to the touch last week so I flipped it over to examine the ballast. I smelled burning transformer varnish(?) and then cut power. I wet my finger with some water and touched the ballast and the water boiled away. I measured it with my IR thermometer gun and it read at around 216° f. Is it time for the bucket light to "kick the bucket"?

Mainly your observation of the light being hotter than before (usual), suggest the ballast life is really over...

Thanks Medved. I noticed also that the bulb had trouble starting at times (flickering, weak arc, loud buzzing of ballast, etc.). The ballast before was about 170°f, so I think the ballast is about to kick it.

Is there a capacitor in the circuit ? You might want to test the capacitor as well then - I am not sure what can be the reaction of the ballast to a capacitor thats developing a short circuit

Medved : If there is a shorted turn, why doesnt it just stop working completely ? (immediate big damage from the shorted turn blowing open at high current etc ?)

Nope it's just a hx ballast without a capacitor. If it is a shorted windin in the autotransformer part then it would just die.

Medved : If there is a shorted turn, why doesnt it just stop working completely ? (immediate big damage from the shorted turn blowing open at high current etc ?)

Because even the shortened turn has some resistance. Don't forget when one turn shorts out in a 100 turn choke, you get a 100:1 transformer with the resistance of the wire in the 1 turn as it's load. That leads to high losses (so power dissipation, as well as lower quality factor - hence the lamp troubles), but the primary current (1/100 of the shorted turn current) very rarely it exceeds the primary current rating, so no overcurrent protection will react.
In fact the dissipation of just single turn shorted is usually less than the overall normal power dissipation, but the problem is, it is all concentrated on that single turn, so it overheats only locally.
But the local overheating causes the insulation of the neighboring turns to fail as well, so more turns become shorted, so the dissilation gradually increases in a runaway fashion.
It is not immediate, the insulation needs time toi decompose to a state where it allows the wires to short circuit, so the ballast may live with an unnoticed shorted turn for weeks. But once it starts to heat up considerably, it means it is not only single turn anymore and the accelerating destruction is on it;'s way, so better to disconnect and replace that thing...

But the current in that 1 turn is on the order of 100x the current in the others, isnt that sufficient to get it hot enough to immediately burn all isolation around, or blow the turn itself open quickly and have more damage from the resulting arc triggering breakdown of allready heated isolation of nearby turns etc ?

Update: I removed the ballast and examined it. The paper winding cover is pretty crispy and has turned black in the area where the heat rose to the top of the fixture. I know where the short started. It was a little nick in the winding paper cover. I noticed it when I got the ballast and it looks to have been there for some years before it was taken out of service, since the nick wasn't fresh looking. I then peeled (chipped) the crispy paper away and saw that the area indeed had missing varnish and has been cooked by many years of use and heat from being enclosed. I dug out a bit more paper and noticed the surrounding area was a darker color, like it has burnt. I took the paper off a non crispy area of the ballast and saw that the windings were lighter and shinier than the crispy ones.

But the current in that 1 turn is on the order of 100x the current in the others, isnt that sufficient to get it hot enough to immediately burn all isolation around, or blow the turn itself open quickly and have more damage from the resulting arc triggering breakdown of allready heated isolation of nearby turns etc ?

Yes, the current on that turn is really high.
The quick burning and blowing out the turn or similar things, all are caused only by the high temperature and not directly by the high current or dissipation.
Because the turms are lying each on top of each other, the heat is carried away by the material of the other ones, so the consequence is just a hotter operation of the shortened turn. So the only consequence is faster degradation of the enamel and varnish rthere.
Therefore it may stay like that really for a long time, till the degradation causes more turns to be shortened as well.
So if the first short was caused by some mechanical damage (the dent,...), it may still operate for years with that turn. When the ballast is really in cold environment (remote ballast in an aluminum cover,...), it may be even for decades.
Only when more turns get shorted (as the enamel have degraded over the years), the generated heat becomes higher, so the destruction gets accelerated (higher heat, already quite some heat wear on the surrounding wires,...). Once it starts to smell, it becomes matter of hours, when it starts to smoke it becomes matter of minutes.

And if that happen to a series choke ballast, it reault into increasing lamp current, so the lamp overheating. And in the last minute the overheating may be so severe, the arctube may explode. Even with the MV. That is the reason, why the fixture design has to be able to contain the arctube explosion with any lamp type.
The only exception is, when the explosion absorbtion barrier is within the lamp itself (the open rated MH's).
But all other high pressure lamps do need to count with the explosion eventuality. In the past, lamp explosion of an outdoor lantern was not considered as a big issue.

Quote from: Ash on July 07, 2015, 02:00:22 AM

But the current in that 1 turn is on the order of 100x the current in the others, isnt that sufficient to get it hot enough to immediately burn all isolation around, or blow the turn itself open quickly and have more damage from the resulting arc triggering breakdown of allready heated isolation of nearby turns etc ?

Yes, the current on that turn is really high.
The quick burning and blowing out the turn or similar things, all are caused only by the high temperature and not directly by the high current or dissipation.
Because the turms are lying each on top of each other, the heat is carried away by the material of the other ones, so the consequence is just a hotter operation of the shortened turn. So the only consequence is faster degradation of the enamel and varnish rthere.
Therefore it may stay like that really for a long time, till the degradation causes more turns to be shortened as well.
So if the first short was caused by some mechanical damage (the dent,...), it may still operate for years with that turn. When the ballast is really in cold environment (remote ballast in an aluminum cover,...), it may be even for decades.
Only when more turns get shorted (as the enamel have degraded over the years), the generated heat becomes higher, so the destruction gets accelerated (higher heat, already quite some heat wear on the surrounding wires,...). Once it starts to smell, it becomes matter of hours, when it starts to smoke it becomes matter of minutes.

And if that happen to a series choke ballast, it reault into increasing lamp current, so the lamp overheating. And in the last minute the overheating may be so severe, the arctube may explode. Even with the MV. That is the reason, why the fixture design has to be able to contain the arctube explosion with any lamp type.
The only exception is, when the explosion absorbtion barrier is within the lamp itself (the open rated MH's).
But all other high pressure lamps do need to count with the explosion eventuality. In the past, lamp explosion of an outdoor lantern was not considered as a big issue.

What are the odds of a lamp explosion in any lamp, MV/HPS/MH.  I know early MH lamps tended to explode, I've seen plenty of those, but I have never seen an HPS lamp explode.

The impedance of the ballast with shorted turn is lower than intact ballast, and it will keep going down as the shorting expands, so overdriving the lamp more and more. This can definitely led to arc tube explosion in any type of HID lamp

Whether the arc tube will take out the outer bulb of the lamp with it, depends on the force of explosion (so the pressure that took to blow the arc tube, so lamp type, extent of overdriving from the ballast, time it had to warm up, age of lamp.....), level of protection of the outer (internal shrout, T or E shape, hard or soft glass), and sheer luck....

Very frequently the elevated current destroys the cathodes or some of the internal connection, extinguishing the arc. But when assuming the cathode and all the connections hold on, the elevated current causes higher power feed to the lamp, so the arctube overheats.

HPS uses to operate at rather low pressures, so even the arctube does not have to be designed so strong. That means it ruptures at way lower pressure, so there is not too much energy to cause extensive secondary damage. And if the arctube happens to be stronger, the elevated pressure leads to elevated arc voltage, so the lamp extinguish.

The MV's are unsaturated vapor concept, so once the dose evaporates, the gas density can not increase anymore. That means rather stable arc voltage and the pressure not that much dependent on the temperature. So the primary problem is the high temperature weakening the quartz, so the arctube loosing it's strength. Most often that leads to the formation and then popping of a bubble, so the energy is in big extend absorbed by the plasticity of the overheated quartz. But still the shock wave could be strong enough to rupture the outer and so send the pieces of glass and melted quartz around. What worsen the situation is the rather large volume of the arctube vessel, so even when the pressure may not be that extreme like with QMH, there is still quite a lot of energy in the form of the pressurized gas.

The ceramic MH are rather similar to HPS in the operating pressures, however as part of the fill operates in the saturated vapor and par in an unsaturated vapor state, while the arc uses to be way shorter, the arc "persist" to way higher pressures and temperatures than with the HPS, so it may lead the arctube to explode. Plus the arctube volume uses to be higher, so there could be more energy in the form of the pressurized gas. The ceramic exhibit no plasticity, so the complete energy becomes "available" to form a shockwave inside of the outer, so the risk of the outer rupture is significant.

The worst are the quartz MH. These operate at really high pressures during normal operation, plus some components use to work in a saturated vapor regime, so the pressure tends to rise significantly when overheated, usually it is just the pressure, what tear the quartz before it get weakened by the heat. That means really high energy inside of the arctube, leading to really violent explosion. In that respect the fresh lamps are way more violent than their worn out counterparts, because the weakened arctube ruptures at lower pressure, so with less energy in the pressurized gas.

Very frequently the elevated current destroys the cathodes or some of the internal connection, extinguishing the arc. But when assuming the cathode and all the connections hold on, the elevated current causes higher power feed to the lamp, so the arctube overheats.

HPS uses to operate at rather low pressures, so even the arctube does not have to be designed so strong. That means it ruptures at way lower pressure, so there is not too much energy to cause extensive secondary damage. And if the arctube happens to be stronger, the elevated pressure leads to elevated arc voltage, so the lamp extinguish.

The MV's are unsaturated vapor concept, so once the dose evaporates, the gas density can not increase anymore. That means rather stable arc voltage and the pressure not that much dependent on the temperature. So the primary problem is the high temperature weakening the quartz, so the arctube loosing it's strength. Most often that leads to the formation and then popping of a bubble, so the energy is in big extend absorbed by the plasticity of the overheated quartz. But still the shock wave could be strong enough to rupture the outer and so send the pieces of glass and melted quartz around. What worsen the situation is the rather large volume of the arctube vessel, so even when the pressure may not be that extreme like with QMH, there is still quite a lot of energy in the form of the pressurized gas.

The ceramic MH are rather similar to HPS in the operating pressures, however as part of the fill operates in the saturated vapor and par in an unsaturated vapor state, while the arc uses to be way shorter, the arc "persist" to way higher pressures and temperatures than with the HPS, so it may lead the arctube to explode. Plus the arctube volume uses to be higher, so there could be more energy in the form of the pressurized gas. The ceramic exhibit no plasticity, so the complete energy becomes "available" to form a shockwave inside of the outer, so the risk of the outer rupture is significant.

The worst are the quartz MH. These operate at really high pressures during normal operation, plus some components use to work in a saturated vapor regime, so the pressure tends to rise significantly when overheated, usually it is just the pressure, what tear the quartz before it get weakened by the heat. That means really high energy inside of the arctube, leading to really violent explosion. In that respect the fresh lamps are way more violent than their worn out counterparts, because the weakened arctube ruptures at lower pressure, so with less energy in the pressurized gas.

wHAT ABOUT LPS/SOX lamps? could they explode? or have a violent failure?

From the "mechanical" perspective, the LPS operate practically at vacuum, so no explosion could occur.
But they are rather large vessels with a vacuum inside, so may implode. And the implosion could send the glass fragments around as well, but with by far lower energy, so rather easy to stop (practically any cover will do so completely, the surrounding air will slow them down to just a free fall within few centimeters away), so it may just turn into free falling (no high speed "projectiles") shower of glass fragments and all other internal parts.
Most dangerous there are the bits of the sodium, as these will likely selfignite once exposed to the air oxygen, or when falling into some wet things, so cause fire.
The remains of the cathodes may be glowing hot and so dangerous in that way, but on the other hand they are rather small, so most likely cool down way before hitting something...