I found my dream lamps, I have always fantasized about this  


https://www.alibaba.com/product-detail/ED-shape-150W-Self-ballast-Metal_60417448635.html

Scrolling down my question is, are the lumens per what to good to be true?


Power(W)   Base   Voltage(V)   Color temperature(K)   Lumens(lm)   Average life (H)   RA   Max diameter(mm)   Max length(mm)   Shape
70   E27   220   4200   6100   10000   65   56   141   ED
100   E27   220   4200   9000   10000   65   56   141   ED
150   E27   220   4200   14000   10000   65   56   141   ED
175   E40   220   4000   21000   15000   65   91   215   ED
250   E40   220   4000   25000   15000   65   91   215   ED/BT
400   E40   220   4000   44000   20000   65   122   292   ED/BT
250   E40   220   3700   22000   15000   70   91   215   ED/ BT
400   E40   220   3700   42000   20000   70   122   292   T/ED/BT
250   E40   220   4000   22000   15000   75   47   257   T/ BT
400   E40   220   4000   40000   20000   75   63   292   TO
175   E27   220   4000   14000   10000   65   39   175   T
250   E40   220   4000   21000   10000   65   47   257   T
400   E40   220   4000   38000   12000   65   47   292   T
1000   E40   220   4000   90000   6000   65   78   340   T/BT
175   E40   220   4000   14000   10000   65   91   225   T/BT
250   E40   220   4000   21000   10000   65   91   225   ED/BT
400   E40   220   4000   36000   20000   65   122   285   ED/BT
250   E40   220   6000   19000   8000   80   47   257   T
400   E40   220   6000   34000   8000   80   53   292   TO








For example, the 400 watt versions are listed at 40,000 initial lumens and 20,000 hour life- and even the dimmest 400 lamp is 34,000 lumens at 8000 hours. Huh? This seems offensively bright for a mercury lamp, let alone one that is self ballasted. I call them mercury vapor lamps, because all the ones pictured appear to have a mercury arc tube (by the length and absence of paint at the arc tube end)

Which leads to my next question. Has China ever produced a genuine self ballasted metal halide lamp? I found this picture which shows a lamp at top with what appears to be a halide arc tube and a ballasting linear halogen:

https://sunlecn.en.made-in-china.com/product/xvnmWkXPJuch/China-Self-Ballast-Metal-Halide-Lamps-T-JLZ-T-.html



Honestly, what are these lamps really like?

I think it is impossible to make a self-ballasted MH lamp using a resistive ballast like a halogen lamp.

What would explain the pain on the arc tube though?


https://sunlecn.en.made-in-china.com/product/xvnmWkXPJuch/China-Self-Ballast-Metal-Halide-Lamps-T-JLZ-T-.html

I don't believe the light output claims at all.
Even if it were MH's, the only part that may each 100lm/W is the discharge burner itself and only those of higher wattages (400W overall rating). And even if it were operating as a MV (so with the probe active all the time, serving as a zero cross restart aid), there is barely 40% of the total input power that makes it to the discharge, the rest is in the halogen incandescent ballast.
The fact the arctube on some types is coated by a heat reflective material does not mean they are really MH's, but it just could mean these tubes need the electrode heat to be retained for proper functionality. I guess the reason is, the close proximity of the halogen makes the maximum possible burner loading rather low, so the electrode area (or better to say the seal area) may need some thermal help to prevent mercury condensation there, so to maintain the pressure.

So how I see it:
Rather wide wattage range of a halogen-incandescent ballasted clear MV's.
40 of power in an MV burner with about 40..50lm/W, 60% in an (compare to normal halogens underpowered) halogen filament with about 15..20lm/W, so corresponding total about 25..32lm/W (lower figure for a lower wattage, higher for a higher wattage models).
The expected performance to me would be something around:
150W ->  3700lm
250W ->  7000lm
400W -> 12800lm

The light output claims they made is a nonsense, I won't believe that even a bit. Sad the went that way, because these amps wuld be interesting even without such "balooning"...


The key point against legislation (at least European) is, "mixed light" lamps are not governed by the efficacy standards (responsible for effective banning the incandescents), because these are treated as "special spectrum" (rahter than general lighting) lamps.

So its 40% watts in the arc tube and 60% in the halogen? I thought it was 50/50- or rather what prevents 50/50?


But still, my thoughts as well. The lumens per watt are just to high. I know GE used to offer a metal halide bulb that dropped into a standard MV ballast. Life was 15000 hours (if I remember right) of 24,000 hours. In theory one could take that bulb and use a resistive ballast. I mean if it can handle a 240 volt OCV without the "spike" of a CWA ballast, I guess why not on a resistive.


How would phosphor fair for these clear SBMVs?  

The problem is, the MH fill is very corrosive and so can not tolerate too high electrical fields on the quartz surface (or better to say the quartz with the electrode material can not tolerate it in the presence of molten salts). Therefore the probe start MHs feature a bimetal switch, which shorts out the auxiliary probe towards the main (so there is no voltage across the small distance) once the lamp ignites and warms a bit.
So what you are left with is an arc burner, with just two electrodes 2cm apart. This needs elevated voltage to restrike after each time the feed current crosses zero, forming a reignition voltage overshoot. This overshoot should be supplied by the ballast, otherwise the arc won't restrike, so effectively extinguish.
And with a resistive ballast the time with no current is rather long, what makes the required voltage for the overshoot even higher, so delaying the possible reignition even further, normally to "impossible".

The MV work, because the auxiliary probe there has very small gap, so the discharge (although limited in current by the starting resistor) is present in the burner before the voltage reaches the 110V (the main arc sustaining voltage), the emitted electrons (either from the auxiliary discharge alone, or photons emitted by it) then establish the main arc immediately once the mains voltage reaches the 110V, so the overshoot is very minimal.


The CWA is a lead type ballast, but with a significant inductive component, so around the current zero cross provides nearly full OCV for reignition (even possible boosted by the LC formed with the lamp wiring capacitance), so the MH is able to work there.
But the resistive ballast can not provide it, so it needs either the auxiliary discharge on the starting probe (the selfballasted MV's), or from the separate "arc keeper" circuit maintaining small current even when the mains can not deliver it (how the Halarc worked; to have this function reasonably simple, it required the discharge to operate from a rectifier).

Here you see no shorting contact for the auxiliary probes (and no room for any arc keeper electronic), this is a clear indication these are just plain MV burners. With halide salt fill this design would fail in few 100's hours...

I've seen US style MH lamps without shorting switches. It was either Phillips or GE up until the late 90s. 

Probe-start MH lamps, have a bi-metal switch, that shorts between the starting and the main electrodes when the lamp is hot, to prevent corrosion at that area of the arctube. From the same reasons, Universal burning MH lamps, should never operated base-down.
When the bi-metal shorts these electrodes, the aux discharge can't be form, when the lamp cooling down to hot restrike. If this bi-metal opens when the lamp is sufficiently cooled, the main arc would restrike immediately, without the appearance of the aux discharge associated with MV hot restrike.
This is a similar action in internal ignitor HPS lamps (To prevent from the starter from restriking a hot lamp) and penning start MV retrofit lamps (To prevent sodium loss from the arctube by the capacitive coupled external starting electrode around the arctube), but with HPS lamps, the bi-metal action is the opposite (Disconnecting [So opening], rather than shorting [So closing]).

Maybe you are right- all of them have a switch. But light me crazy, I swear I saw a Halide lamp without a switch.

So what you are left with is an arc burner, with just two electrodes 2cm apart. This needs elevated voltage to restrike after each time the feed current crosses zero, forming a reignition voltage overshoot. This overshoot should be supplied by the ballast, otherwise the arc won't restrike, so effectively extinguish.
And with a resistive ballast the time with no current is rather long, what makes the required voltage for the overshoot even higher, so delaying the possible reignition even further, normally to "impossible".

Regarding not being able to re-strike after the zero crossing on a resisitve ballast, I found a vid to the contrary:


https://www.youtube.com/watch?v=SDP58iqGNfk

As you can all tell, I am very optimistic about the concept of a self ballasted metal halide.  

There is the ignitor, it generates spikes every time the arctube haven't ignited and so reignites it for every half wave. Yes, I haven't mentioned this method (using external voltage pulser for reignition), but indeed it works as well.
But still you would not fit ignitor into a common outer with the arctube and its ballasting filament capsule. Mainly because the electronic won't stand the heat there.
Plus I would doubt about the lifetime of an ignitor used that way.

There is really not much to ballast an MH by just a resistor. There is no way to integrate all required components into a common outer.
And when some parts have to be external, why not to put there decent high efficiency ballast (a choke,...)?
And if we are talking about a rather physically large assembly with rather low power, even the complete high efficiency electronic ballast may fit there (like with the modern selfballasted reflector CMH's). Yes, we are then talking about a selfballasted lamp, yet not about a resistive ballasted one...

What ever these are, their NOT halide lamps, it’s evident in the lack of visable halide salts in the discharge tube?

What ever these are, their NOT halide lamps, it’s evident in the lack of visable halide salts in the discharge tube?

Sometimes the pool sits rather hidden in some corner, so I didn't dare to guess that jut from the marketing photos. But you are right, the lack of visible salts is indeed a clue here...

But I've already bought similar (advertised as) "metal halide lamp" and I'm sure they are MV's. I've expected and hoped them to be MV and I was really happy they really were, because there are lot of clear MHs around, but rather none clear MV's here... ;-)

The lumen and lamp life figures are simply taken from standard metal halide lamps at the same wattages, 36-40K lumens is typical of a newly installed 400w probe start Na-Sc MH lamp. The same goes for all the other wattages listed. As Medved pointed out, the actual output of these SBMV will be 1/4 of that. Lamp life should be 6000-10,000 hours if these are decent quality, but likely 3000 hours or less given these are on alibaba and going for $2 ea.

Also, some unscrupulous chinese dealers will list mercury lamps as metal halide, I am not sure why. My company once imported a case of 250w metal halide lamps (standard M58 lamps) from alibaba or madeinchina, what arrived were mercury lamps, interestingly enough the lamp's packaging as well as the etch on the bulb itself stated "metal halide" "MH250/U".

I'd say you are spot on BlueHalide, the values do indeed match spot on. Would these be more inline with a self ballasted system?


http://www.tcntcn.com/product/55457.html


5000 hours sadly is kind of low for such a complicated lamp.

I'd say you are spot on BlueHalide, the values do indeed match spot on. Would these be more inline with a self ballasted system?


http://www.tcntcn.com/product/55457.html


5000 hours sadly is kind of low for such a complicated lamp.

These are the figures I would expect too.
Maybe the life rating could be a bit longer, but it is a question what ON/OFF frequency they have anticipated for a 10hour/start.
But for a 3hour/start pattern it becomes rather excellent, given the normal principal wear occurs virtually only first few minutes after ignition (standard 16k rated SBMV rating once common with high quality SBMVs mean 1600 starts at 10h/start pattern, with 3h/start pattern the same 1600 starts means 4800hours life)...