Does anybody have any info on this, I cannot seem to find any online. I have a few Venture 250w 5500K daylight metal halide lamps I use in my garage, they are in typical high bay fixtures with aluminum reflectors. All conatin Advance CWA M80 ballasts (probe start). These lamps are now well burned in and are probably at 300-500 hours of use. Ive noticed recently that the arc in two of them swirls around in the arc tube causing moving shadows on the floor and around the room, its hardly noticable but there is also a slight back and forth color shift with this as well. I removed the lamps doing this and installed different ones by a different brand and so far no swirling. I even went as far as to warm up one of the swirling lamps on a 250w HPS ballast (CWA type) and the arc even swirled on that. So its not the ballast. Does anybody know if the arc swirling is detrimental to lamp life or performance? I like the color of these lamps and dont really mind the swirling, I hope it goes away eventually, but theyve been doing this the last hundred hours or so.

Deleted

If the lamp is close to the ballast, it could be affected by it's magnetic field, so rotatingthe ballast could affect the swirling.
But the primary cause is the lamp itself.
Beside the swirling itself being obtrusive, there is no life problem with that, it would eventually stabilize again...

Medved: In my Nisko fixture, the Radium Relux 9W/840 TC-S, is directly in front of the ballast, yet the ballast don't causes swirling inside the lamp. So swirling on magnetic ballasts, should be caused by the lamp itself.
Generally, is electronic ballasts produces less magnetic fields than magnetic ballast?
Also, swirling in MH lamps is a dangerous situation, unlike in fluorescent lamps, because overheating can increase the risk of arctube explosions.
Generally, swirling in HID lamps, have negative effects on their life, unlike low pressure discharge lamps like fluorescent lamps. This is also the reason why electronic ballasts for HID lamps operating at low frequency square wave current rather than HF sine wave current to prevent reasonances inside the arctube (Which is a similar phenomenon to swirling in appearance).

Deleted

Medved: In my Nisko fixture, the Radium Relux 9W/840 TC-S, is directly in front of the ballast, yet the ballast don't causes swirling inside the lamp. So swirling on magnetic ballasts, should be caused by the lamp itself.

Yes, it is indeed primarily caused by the lamp itself, it could be sometimes only influenced by the magnetic field, so by relocating it it could change the swirling. But that is more the case for high current HID's and not as much for the low current density fluorescents/LPS.

Generally, is electronic ballasts produces less magnetic fields than magnetic ballast?

Generally yes, because for the same VA's it mean way less energy. But I've seen some designs using open magnetic circuits, so there the external field is not as weak. But it drop significantly, as you increase the distance.

Also, swirling in MH lamps is a dangerous situation, unlike in fluorescent lamps, because overheating can increase the risk of arctube explosions.
Generally, swirling in HID lamps, have negative effects on their life, unlike low pressure discharge lamps like fluorescent lamps. This is also the reason why electronic ballasts for HID lamps operating at low frequency square wave current rather than HF sine wave current to prevent reasonances inside the arctube (Which is a similar phenomenon to swirling in appearance).

What damawe the HID's is not the swirling, but a pressure standing wave, causing the arc to burn in some places with way higher power density than the average and/or closer to the arctube wall. But most notably the "bended" arc does not moves, so on places, where it become closer to the wall, it overheat it there.

But the swirling present on low frequency drive mean only the arc is slightly moving as a consequence of the cathode hot spot moving around the electrode tip. And all the time it remain in a safe distance from the wall. So such swirling is not dangerous at all...


@imj: The CWA is more of a constant current source, so it increase the power when the arc voltage rise, what make saturated vapor style lamps (HPS,...) thermally less stable.
But some lamps tend to swirl their arc on the series choke as well - there I expect no difference.
Where the CWA (so no DC current path) could be different from series chokes (DC conductive) is, when the lamp start to rectify (even slightly and temporarily; as a result of a disturbance on the mains,...). On the series choke it cause the 50Hz flicker (with some heavy disturbances it could extinguish), with series capacitor (so the CWA) you do not notice anything happening.

Dont compare the MH to the PL either - a "filament" arc in a HID behaves different than a low pressure discharge in a PL. Besides 400w MH is 3.25A arc and 11w PL is 0.155A arc, and i'd assume the 400w ballast is emitting more field as well

I tried to get moving arc in fluorescents using powrefull permanent magnet (from a hard drive) with allmost 0 results

Medved, this sounds like a good reason to use CWA ballasts - it's a shame we don't have them here as an 80W CWA mercury ballast sounds ideal for my uplighter. The mains round here is terrible, plenty of dips and flickers and the occasional brief period with the lamp partially rectifying as a result.

These voltage dips regularly extinguish the 250W SBMV behind the stairs, at one point it would drop out most evenings, quite disconcerting when it suddenly goes out in the middle of a horror film! I'm sure Roobarb doesn't appreciate his "sun lamp" going off either, he's usually making himself comfy next to it when it's in use!


BG

Dont compare the MH to the PL either - a "filament" arc in a HID behaves different than a low pressure discharge in a PL. Besides 400w MH is 3.25A arc and 11w PL is 0.155A arc, and i'd assume the 400w ballast is emitting more field as well

I tried to get moving arc in fluorescents using powrefull permanent magnet (from a hard drive) with allmost 0 results

It does behave differently: The forces are Curent*Flux, so the 3.25A cause ~20x larger forces on the arc than the 0.155A fluorescent in the same magnetic field, what is quite a difference.
But the main difference is in the arc positional stability: In low pressure discharges, the arc fill the complete tube volume, so it's position is restricted by the tube walls. So it can not move, even when the force would be really high.
On the other hand the HID arc is very narrow and it is basically free to move there, held in place only by the fact, the straight line is the shortest connection between electrodes. But you can see, than even the gas convection is able to bow it very significantly upwards, while the convection forces are not as high.

Generally the free-to-move narrow arc is from nature unstable (if it bend, the magnetic forces tend to bend it further,...), to stabilize it even asked for extra additives into the gas mix (which lower the efficacy) to fatten the arc a bit, so make it more stable. And as these additives lower the lamp performance, only the minimum necessary amount is used to make the arc just moreless stable at low frequency drive (so without acoustic resonances).

But what happen when the arc swirl, is the cathode hot spot (so the arc root) moving around the cathode surface, so the arc move then accordingly. These are ~1..2mm motions, so not significant in terms of arc proximity to the walls (the convection shift it way more), but as it is moving, it cause visible flicker effects.

Thanks for the info guys, well the arc is still swirling in the lamps, no improvement at all since I posted this topic. I got on a ladder and looked at the arc through some thick sunglasses and saw something else interesting, there appeared to be a rather large blob of liquid halide salt moving around the bottom of the arc tube as the arc swirled, the blob would move in the opposite direction of the arc's movement. It appears to be reddish orange in color but difficult to tell with the sunglasses. Both swirling lamps have this. funny thing is, the lamps are still operating at full brightness and the correct color. Kill-a-watt reads 298w, (I took a reading because I thought maybe the lamps were underdriven due to the excess un-vaporized halide salts.) If I move the fixture and hold it at a different angle (horizontally) the arc stops swirling. These lamps are Universal-burn rated.

Are these quartz halides or ceramic ones because every quartz one I've had swirls on high frequency ballasts but not on magnetic ones. Ceramic halides don't seem to have this problem. Its interesting... 

Are these quartz halides or ceramic ones because every quartz one I've had swirls on high frequency ballasts but not on magnetic ones. Ceramic halides don't seem to have this problem. Its interesting... 

Ceramic have different fill and mainly different (way lower) operating pressure, what mean the acoustic waves can not spread inside with so much energy, so the acoustic resonance does not have so high quality factor. Other aspect from the different pressure is the difference in the speed of the sound waves inside the atmosphere, so the resonance frequencies are different.
For the resonance problems to appear there is necessary the ballast operating frequency to match the acoustic resonance of the arctube and build up high amplitude standing waves.

Other difference is the arc itself: In the lower pressure ceramic the arc could become wider, so a bit more stable, while the thermal conductivity of the higher pressure gas prevent the arc to widen as much in QMH's, so it swirl there more violently.

But in any way, unless the ballast is explicitly rated for the exact lamp type (by exact I mean exact, so no "equivalent" from another manufacturer - the "equivalence" is valid only for parameters governed by IEC standards, but these do not govern parameters important for safe operation on HF). Different lamp (different maker,...) could build up the dangerous standing waves and get damaged. And it could happen so not immediately, but after some hours of use,...

They're not ceramic. They're standard quartz probe-start universal burn daylight color metal halides 250w. These lamps appear to use Indium, they go through a strong blue color during warm-up and have the purple-pinkish deposits on the arc tube wall when the lamp is off. Ballasts are magnetic probe start CWA type. Still swirling too...which I have a feeling will never go away

An electronic ballast corrects the swirling problem, I tried it, or operating the lamp horizontally on the magnetic ballast, swirling stops. It only occurs when base-up on the magnetic ballast. It also seems there is far too much halide salts in the arc tube, they form a liquid blob on the bottom of the arc tube when running at full power on a CWA thats drawing almost 300w

It seems your lamp have an overdose manufacturing fault.
The LFAC electronic ballasts run the arc more steadily, because the actual arc power is constant even over the period of the feeding curent. The 100Hz ripple present on magnetic already cause some instabilities, mainly as a result of the periodic extinction/restrikes around the current zero crossings.
In this matter the worst would be the CWA, as it feed rather narrow current pulses (crest factor ~1.7 and more), the series reactor (and/or HX autotransformer) feed the lamp nearly a sinewave, so crest factor ~1.4, while the electronic LFAC feed the rectangular, so CF=1.