I am trying to repair a 150W double ended MH floodlight for a friend

The floodlight contains Eltam 150W MH/HPS magnetic ballast and Vossloh Scwabe Z400MK superimposed ignitor

The floodlight was working ok up to last winter, then it began to start erratically : from what the friend describes, the lamp would glow for long time in dim blue, and if left powered for long time it would eventually start and heat up to normal light output. Mow it does not start anymore

I opened it

There is severe water ingress inside. The ballast was partially submerged in water, ignitor partially submerged, and PFC capacitor is melted completely (probably the water got in too, then it started destroying the capacitor) and is basically some plastic mess with a burned cylinder and 2 wires sticking from it

I tested the ballast with multimeter, no continuity at all. I took off the terminal block and found out that it was destroyed by corrosion - it does not conduct because of the corrosion layer that is on all the parts. I soldered wires directly to the coil connections, then put the terminal block back just as isolation / to hold the wires so they dont break

The ballast after this repair measures 5 ohm with a multimeter (end to end of the coil), and if connected directly to 240V it is silently humming and very slowly heating up - just like an intact ballast, so i assume that its ok

I connected it with the ignitor and lamp, and the lamp strikes, but it keeps just glowing dimly with the starting gas color, with ocassional slight flash of mercury and back to clean starting gas

The lamp was never submerged as the water level did not reach it

I'd perhaps blame the ignitor next, but if there is starting gas glow doesnt it mean that the ignitor is ok ? (the ignitor is potted with resin so sealed but perhaps water could soak in through the terminal block etc)

ANy other guess what this may be ? I'll get there next time with a complete set of gear and lamp to try to find out, but any guess ?

Update

Been there today, and took with me a big 150W HPS floodlight + 400W MH lamp with gear

Test 1 : Double ended 150W MH on HPS gear with semiparallel ignitor : Nothing happens, ignitor is clicking

Test 2 : The same but with superimposed ignitor (taken from the 400W MH gear) : Nothing happens, ignitor buzzing

I tried to take the MH out of the socket and 1 of the white ceramic ends broke off very easily, as if it was allready broken in the 1st place and the socket pressure was holding it together. Under it, the seal to the outer lamp is destroyed by corrosion and shows minr signs of arcing (in series with the lamp where the conductor failed). So ok, its the lamp bad



I tested the gear of that floodlight with my lamps :

The 150W MH ballast with the known good MH ignitor and 150W HPS lamp, worked no problem

The Vossloh Scwabe Z400MK ignitor with 400W MH ballast and lamp. But now whats interesting :

1st attempt the lamp striked ok and ran up to full power, though the ignitor kept buzzing for quite a bit longer than my good ignitor

Then i unplugged and plugged back in the gear, to get a hot restrike. For the 1st 10 minutes there is no reaction and the ignitor is silently buzzing. Then the starting gas in the arctube was glowing blue from the ignitor, and it appeared ok. Then began flashes of mercury color (the lamp approaching restrike, and suddenly it went out completely (ignitor not buzzing)

I unplugged and plugged it back in, i tried to unplug and wait few minutes, ignitor is not buzzing and not striking the lamp. After quite a lot of poking around and disconnecting the ignitor from the circuit and connecting it back, still does not work

Then i put in the known good ignitor instead, and it worked

Then i put in the ignitor in question, and it worked too. After few seconds i unplugged and plugged it back, and it restriked ok the 2nd time. Then ok the 3rd time. Then it did not restrike for the 4th time

I unplugged it, waited a minute maybe more and plugged back, and it worked again, i tried 5 times and it worked in all of them

Is this how a "smart" ignitor with EOL lamp detection is supposed to work when forced to hot restrike, or is it faulty ? (is the  Vossloh Scwabe Z400MK a "smart" ignitor ?)

Hi Ash, it seems that you're not alone with such problems with VS ignitors.
In the past we had several premature failures with some of these ignitors.
IMO the cut out timer works not properly!?

As replacement You shoud try Philips or Tridonic ignitors as they are working better.

Other question is how old is the MH lamp? Some very new lamps are sometimes troublesome.
What about the lamp socket?

I am testing the VS ignitor with this lamp (it is moderately used but in good condition) and this ballast. The known good ignitor (which i put in to verify that the problem is indeed with the VS ignitor and not something else) is this

The socket is a standard E40 HID socket made by BJB

As for choice of ignitors, the most common ones here are some locally made ones, which are pretty good. What i like the most with them, they are not "smart", they keep clicking forever untill the lamp starts. Most simple and most reliable. Dont like the "smart" ones, exactly for the bahavior mentioned here. However this is my 1st encounter with this model of ignitor. I dont know if it is a "smart" ignitor, and whether its behavior is due to it being "smart" or being damaged from the water ingress in the fixture

Ash: According to VS PDF catalog, their ignitors are available in either standard version, or with an automatic shut-off.

Is this how a "smart" ignitor with EOL lamp detection is supposed to work when forced to hot restrike, or is it faulty ? (is the  Vossloh Scwabe Z400MK a "smart" ignitor ?)

The "smart" ignitors are sold with the intelligence on three basic levels:
- Timed only: They give up the restrike attempts after some period of unsuccessful firing, usually few minutes. Then they wait few 10's of minutes (to let the lamp cool down) and retry another attempt cycle.
- Failure to strike detect: After few unsuccessful ignition cycles they give up permanently, tilo the power is cycled (to reset them). The power should be OFF for some minute to let all the internal capacitor's discharge below the PowerOnReset level of the controller.
- Advanced lamp failure detections: Monitor the lamp voltage pattern and shut OFF, when it evaluate the lamp as faulty (e.g. the HPS cycling demonstrate itself as rising arc voltage from the ignition too extinction with repeated waveform). But these use different, frequently proprietary detection. Some of these are tailored to work with exact lamp family (HPS, QMH, CMH,...) and so are not 100% compatible with other than rated lamp family (as it behave differently, so the protection won't work properly).


The failure mode you described could be linked to the bad lamp contact as well. With Rx7s I had big trouble with SLI (European Sylvania) lamps, as they used only tiny drop of the cement and it was insufficient to hold the contact pressure (Well, I hope they fixed the problem in the meantime).
In your case I would guess the cement was damaged by the moisture inside the fixture, so it lost it's strength.
If there is bad contact, the HV puilse jump over, so the lamp flash (or glow from these flashes, as many ignitors give 100pulses/second), but the main, low voltage high current for the main arc could not pass.

When the double ended MH lamp disintegrated (in exactly this way - from moisture), it is now obvious why it was glowing but not starting - it would just light on the ignitor and unable to strike a big arc, except maybe for short times where the failed contact in the seal would flash over for a second

But the further testing i did with the ignitor (where it sometimes stopped buzzng and won't strike), was by connecting it with a whole other ballast, wiring, socket and lamp - ones which i brought with me

Ash: You can order an Osram Powerball HCI-TS 150W/NDL, directly from Nisko, or from a wholesale electricity shop.