HID lamps in series, is it safe? (Solved)

[cs.dk]

I know, maybe a dumb question. But i can't figure out what can happen.
On my 48V battery forklift, i have 2 worklights mounted (24V H1 bulbs) if one bulb blow the circuit is open and nothing more happens.

I have a couple of HID worklights on my shelf, and decided to use them, instead of tossing them out. This is the model (They are 24V versions); http://www.eagle-lights.com/product_show.aspx?id=75

What happens if ie. one bulb blows? Then the startercircuit is not loaded like the other.

So question is, is it safe to couple them in series, or should i tap of 12 cells with a proper fuse, and a switch? Solution one is the easist for me, as the wires are already there.

[botcrusher]

So long as the bulbs don't have a Christmas bulb failiure mode (failing shorted) i don't see a problem.

[MarkF]

Just to clarify.
This is NOT just putting two bulbs in series (like Christmas lights). 
This is like putting two fluorescent lights in series.  You're placing the two ballasts in series.  I don't know how the ballast will react without its load.
I would guess that the ballast would handle the bulb burning out and as long as you have the units on a fuse, you'd be okay.

[Ian.M]

I would expect nothing good.  Differences in current draw during startup and time to start will invariably cause imbalance.  As soon as the imbalance is enough to brown out the controller, that lamp's arc will quench.  The voltage will then spike dramatically until the other lamp browns out.

[botcrusher]

Oh dur... Right HID uses a ballast.
Well then, my response changes to Idk, YMMV

[cs.dk]

Thanks for the replys. I did have a fear something could go wrong with them. I don't know whats inside a HID ballast though.

I'll tap of 24V with a fuse and some new wiring.

[Ian.M]

Is your battery bank big enough that the 2.8A or so they need wont imbalance it too badly?  I would assume so, just because its a forklift, but I would be cautious if the expected AH draw between charges for the lights is more than 1% of the battery capacity.

[richard.cs]

If you put them in series *and* connect the centre point to the 24V tap you will need a double pole switch or a relay but they will both reliably see 24V without unbalancing the battery much.

[Ian.M]

There's issues with that if the fuse to the tap ever blows - you then end up with the lamps in series.  On a smaller battery bank, that or a DC-DC converter would be essential to keep it balanced, but as it a large traction bank, as long as the average lamp load is significantly smaller than the self-discharge, the imbalance from the simpler tap to ground load shouldn't be a problem.

[Kjelt]

A HID ballast will generate a high voltage ignition at the start time (between 1500V and for cars even 20000V) and then after the lamp is ignited it will continue to run the current and keep track of the voltage in order to keep a constant watt output (current driven, voltage is going up automatically with the lamp).
So a ballast can not do this over two hid lamps in series, they will never ignite and startup. Two the same lamps in parallel would work but then only one (that will ignite) will give light in the end.
If one breaks down or is still hot and you restart the ballast the other lamp will ignite. But that is not what you want, if you want both lamps to work you need two ballasts.

[richard.cs]

There's issues with that it if the fuse to the tap ever blows - you then end up with the lamps in series.

One way to mitigate that would be to use a double pole MCB instead of a fuse so that the 24V tap and the 48V break together.

Agreed that he might get away without it on a large traction pack but if the lights are used a lot then it is not an entirely insignifcant amount of charge.

[Ian.M]

I cant help thinking that a rail splitter to limit the voltage across the lamps is the way to go.  e.g. a pair of complimentary power Darlington emitter followers on a chunky heatsink with a resettable low temperature thermal trip attached to interrupt the 48V feed, with the bases fed from a potential divider with three series diodes to set the base bias so the quiescent current is fairly negligible.

Its only got to cope with the running inbalance between the lamps and short term, one lamp's current foir startup inbalance.  If a lamp fails, the dissipation shoots up, but the thermal trip will shut it off before it gets hot enough to damage anything.

[cs.dk]

Sorry for bringing this old topic up again.
My sparetime has gone with other projects, but it's time to look at this again.

Would you trust a DC/DC module to power the lamps, something like this; http://www.ebay.co.uk/itm/DROK-Waterproof-DC-DC-Buck-Converter-48V-to-24V-15A-Step-down-Voltage-Regulato-/252433178106?hash=item3ac6309dfa:g:AFcAAOSwNuxXaV86

The efficiency is pretty high; "high conversion rate(above 902%), low heat and stable." 

It was my hope that one module could start both lamps. A lab PSU current limited at 5A can't get one running. Might try with 7,5A, if it goes the module should be able to start them both? Once the arc is striked the current consumption falls a lot.

[cs.dk]

Sorry again for bumping.

But i have a question - Can I damage the SMPS if i switch it on/off at the inputside? I don't want to have power on the inverter, when the vehicle is parked. It's chinese, and could possible catch fire, short out, or something else i don't want to happen.

Updated link: http://www.ebay.co.uk/itm/DROK-Waterproof-DC-DC-Buck-Converter-48V-to-24V-10A-Step-down-Voltage-Regulat/182359689287?_trksid=p2047675.c100005.m1851&_trkparms=aid%3D222007%26algo%3DSIC.MBE%26ao%3D2%26asc%3D38661%26meid%3Dc06a8ba18ba746968dd6cde42923dbe2%26pid%3D100005%26rk%3D4%26rkt%3D6%26mehot%3Dlo%26sd%3D272340025388

[cs.dk]

Just to close this thread.
The cheap DROK DC-DC converter said boom; https://www.eevblog.com/forum/chat/watch-out-for-drok-products!/msg1088019/#msg1088019

Got one of these 24V 14,6A Mean Well PSU's ; http://www.tme.eu/en/details/sd-350c-24/dcdc-converters/mean-well/#

The thing fail to start just one of the xenon lamps, it's shutting down instead That seems strange.

Took my Brymen BM089 clamp meter, and in Crest-mode is shows 78A in inrush current per lamp! Don't know for how long time though, but long enough to shut the mean well psu down. Gotta drop this project, and sell the lamps to someone with a 24V vehicle.

We got some 15W LED worklights at my work, two of those will be fitted now.