I've recently acquired a couple of Osram HQI-E 70W metal halide lamps (rated for open fixtures) and seem to be having a bit of trouble getting them to strike, the last few times I've used them they seem to be OK but I wonder if this is an isue with new bulbs or just the way I have it wired? Perhaps the lead is still too long between the ignitor and lampholder? I couldn't get it to work at all with the ignitor wired alongside the ballast.

The photo here shows the modification to the lamp cord, this is to incorporate a pluggable ignitor or bypass link so the lantern can be used with mercury, high pressure sodium (including my lamps with internal ignitors) or metal halide lamps.

Here is a photo showing the lamp at full power.

When the lamps strike they exhibit the usual run-up displays with pink flashes etc.; the ignitor buzzes loudly and tapers off as the arc forms.


BG


>EDIT< Links changed, hope these work!

Links are broken.

You (likely accidentally) set access restrictions for the photos, so they could not be reached...

The superimposed ignitors perform best, when they are placed close to the lamp and further away from the ballast.
Some superimposed ignitors are lacking the capacitor on their input (between "ballast" input and "Neutral"), those have generally more troubles to ignite the lamps, when there is longer wiring between the ignitor and lamp, but when the ignitor is close by to the ballast.
The reason is, than the ignitor generate the high voltage between the "Ballast" input and "Lamp" output, so the current necessary to charge the wiring between the ignitor and lamp have to go through the ballast coil, while the generated pulse is very narrow, so all that is quite high frequency event.. And as you know, the inductor does not like to conduct the high frequency current as much, so as a result the ballast inductance tend to limit the ignitor's output current. And when this is combined with higher capacitance of the wiring to the lamp, the reached peak voltage become low.

This problem is solved inside many ignitors by connected ~10..100nF capacitor just between the "Ballast" input and "Neutral" common, so providing the bypass path for the (rather high frequency) current pulse. But some ignitors are lacking this capacitor, as they count for the fact, than the parasitic ballast capacitance and eddy current losses are rather high, so provide sufficient current path for the short distance from the ignitor to the lamp.
When you use a ballast with lower eddy current losses, it's impedance for the narrow pulse increase. And the longer wiring to the lamp mean higher load capacitance...

So when moving the ignitor to the lamp side of the cord, you reduce the capacitance the ignitor have to charge to the high voltage, while the longer wiring between the ballast and ignitor provide the required bypass, both effect ensure there is more powerful pulse left to ignite the lamp.

And another experience: "HPS/MH 35..70W" ignitors usually generate only ~2..3kV, what is sufficient for HPS, but frequently not enough for many MH (my experience).
It is better to use "MH 150..400W" ignitors, what generate about 4..6kV pulses, so the starting is way more reliable. Don't worry about the ignitor being rated for higher wattage than the lamp, the ignitor rating is related to the current it could handle (it have to flow through the pulse transformer secondary) without overheating, so the 70W would mean they would run way cooler, so they should last longer. So the "forbidden" combination is to use an ignitor with lamp of a greater power than is the ignitor's rating.

Thanks .. I will have to look at the links as they work at my end. Depending on how I get to the photos I get different URLs

I'll certainly try your suggestion Medved! Cheers


BG

The pink flashes in your Osram HQI-E 70W, when it reaches full brightness (More common in Na-Sc lamps), rising me the question if your "Osram HQI-E 70W", is an original lamp (From the european manufacturer), or a clone (Made in USA, have a plain MV arctube with heat reflective coatings and have sodium and scandium filling).

I will have to look at the links as they work at my end.

Still does not work...
Log out from the LG and try to look at them as an "anonymous guest"...
I have an suspicion, than it is related to the problem described in http://www.lighting-gallery.net/index.php?topic=1251.51 - the browser auto-filling the password...

To prevent the browser auto-filling of the album password box, I don't save the password inside my browser, and remeber it by heart.

The albom was indeed set to "private" .. that explains things, thanks Medved. Hopefully it should be OK now


BG

Update:

I've discovered that tapping the bulb causes it to strike! A mate of mine has done the same with the same result, so it's not just me. Strange! Maybe the shock/vibration knocks some of the salts off the electrodes or arc tube wall, or something?


BG

I don't think the halides on the electrodes could cause the lamp to not start. I would rather guess for the dust (forming from the halides) fall through the arctube and deform the electrical field so, it have more high field strength areas, so easing the breakdown.
So that would really suggest for too low ignition spikes...

I've been using the coated MH lamp more recently as I need extra light in the winter, I've found that the lamp strikes straght away now (at least for the past 3-4 starts) and the pink flashing is MUCH stronger during run-up than when the lamp was first used, probably has 100+ hours on it now. I'll have to do the same with the clear lamp!


BG