Brand new MH bulbs flicker but won't fire

[LSWM]

I've got 3 400 watt MH ballasts gifted from a friend. He was using it them as aquarium lights, but has LED's now, so the current bulbs are like 20,000K spectrum.

The original bulbs fire, as does another old bulb I have laying around, however I bought some 7,500k bulbs, and they flicker once or twice when I turn the ballast on, and again when I shut it off, but they never fully ignite and fire up.

Anything I'm missing here?

 

[LSWM]

I had changed the bulbs during the dark cycle and had sat and waited for 2-3 mins and flipped them on and off 3 times. I just went over there to change the bulbs for the day cycle, and they had turned on by themselves after the timer flipped them on.

Guess false alarm? I'm hoping I don't have any issues with them randomly not firing.

 

[Jhhnn]

There are two kinds of MH ballasts & two kinds of lamps. Probe start lamps will fire in either probe or pulse ballasts, pulse start lamps won't fire reliably if at all on probe start ballasts.

Probe start lamps have an electrode that pulse start lamps lack, illustrated here-

http://www.lrc.rpi.edu/programs/nlpip/lightinganswers/mwmhl/differenceprobepulse.asp

Probe start ballasts have only a capacitor & transformer, pulse start ballasts also have an ignitor as with HPS systems. I think that most MH remote ballasts marketed for growing are probe start, don't really know. Both of my salvaged industrial 400w MH ballasts are pulse start, otoh.

A weak capacitor can also lead to erratic lamp starting, probably a weak ignitor in pulse start systems, as well. It's easy to test the capacitor using a kill-a-watt meter. Fully warmed up, the system should have a power factor >.90, ideally .97-.98.

Always check the obvious first, in any case, like the wiring connections.

 

[LSWM]

All of my lamps are probe start. I definitely checked the wiring first thing.

Would a weak capacitor cause a delayed fire? Like 10+ mins?

It's beginning to look like I need to buy a bunch of new caps or transformers. My multimeter broke recently so I'll need to grab a new one before testing the ignitor correct?

Thanks for the help Jhhhnn.

 

[Jhhnn]

All of my lamps are probe start. I definitely checked the wiring first thing.

Would a weak capacitor cause a delayed fire? Like 10+ mins?

It's beginning to look like I need to buy a bunch of new caps or transformers. My multimeter broke recently so I'll need to grab a new one before testing the ignitor correct?

Thanks for the help Jhhhnn.

I don't know if an ignitor can be checked with a multimeter. They kick out a very high voltage pulse that may not agree with the meter or the leads, dunno. Never tried it. They're generally tested by substitution.

Might want to check everything independent of the timer, too.

 

[Miraculous Meds]

Sounds like capacitor maybe. They are cheap and easy to change. Just discharge it before u pull the old one out so u don't shock urself. Are u running the vert or horiz.

 

[Drek]

A weak capacitor can also lead to erratic lamp starting, probably a weak ignitor in pulse start systems, as well. It's easy to test the capacitor using a kill-a-watt meter. Fully warmed up, the system should have a power factor >.90, ideally .97-.98.

Testing a capacitor by means of power factor isn't all that conclusive. All parts of the circuit contribute to power factor; resistive power(true) and the calculations of inductive and capacitive reactance(apparent power).

It's always best to test a component, by measuring the component. Visual inspection of the capacitor(leaking/bulging) and direct testing(by means of a multimeter set to capacitance) it's value in comparison to the indicated value and tolerance.

Ignitors, at least in my field, are tested by means of testing it's resistance(ohms). I'm not certain as to testing commercial lighting ignitors, however I'm assuming the same approach is probably valid, as the ignitor(just from guessing) probably involves some sort of step up coil(as in direct spark ignition) which again can be tested by verifying it's resistance(open coil, good coil, bad resistance value(indicating impending failure).

 

[LSWM]

Testing a capacitor by means of power factor isn't all that conclusive. All parts of the circuit contribute to power factor; resistive power(true) and the calculations of inductive and capacitive reactance(apparent power).

It's always best to test a component, by measuring the component. Visual inspection of the capacitor(leaking/bulging) and direct testing(by means of a multimeter set to capacitance) it's value in comparison to the indicated value and tolerance.

Ignitors, at least in my field, are tested by means of testing it's resistance(ohms). I'm not certain as to testing commercial lighting ignitors, however I'm assuming the same approach is probably valid, as the ignitor(just from guessing) probably involves some sort of step up coil(as in direct spark ignition) which again can be tested by verifying it's resistance(open coil, good coil, bad resistance value(indicating impending failure).

Your average cheap multimeter doesn't check capacitance does it? That was how I intended to use the multimeter to test the ignitor was with it set to Ohms. Works for speakers too!

 

[Drek]

Many multimeter's these days can test capacitance. It's good to buy one that has this feature. Symbol is: http://www.clipartbest.com/clipart-yTkqAnbTE

Ebay(decent and cheap): http://www.ebay.com/itm/New-VC99-3-...055?pt=LH_DefaultDomain_0&hash=item2a1d16b5e7

Cheers!

 

[Jhhnn]

One thing I'm sure of, Drek- if the power factor is correct, & the system has a compatible lamp, then the transformer & capacitor are working as designed. It can be no other way. Many lighting capacitors have internal discharge resistors, as well, making them difficult to test other than operationally. Internal transformer values don't change, other than to open or short, so if it works at all then the transformer isn't a problem. If it's shorted, you'll probably smell it, and if it's open then that can be determined with an ohm meter.

 

[Drek]

Hey JH. I know one thing for sure, after going to school for 6 years, Power factor is not capacitance. It's a determination of how efficient the load/circuit is operating. It's determined by taking everything into account. By altering the capacitance, you alter the phase relationships that affect how efficient the load is operating. Any of the elements involved in the circuit can affect the power factor, according to the math. I sat there, in class, and did all the calculations. The lamp for instance, can affect the power factor, even if the capacitor is fine...etc...and these are just simple circuits. Deductive reasoning isn't all that objective. ie: If this is good, than this has to be good...

That's why mechanics test capacitance and components with meters; that's how it's done in the field...and not the marijuana field.


- So I guess my point was, you either measure the component directly...or you guess.

 

[intotheunknown]

There are two types of ballasts , each require a certain bulb.
For example a digital ballast will require an ANSI type bulb.
Make sure youve got that right first.

 

[LSWM]

So after looking around some more, one of the ballasts has a electric diagram on it for wiring. The ballasts are electronic and it says "Blue Line Electronic Metal Halide Ballast 400w" on the top.

Input Voltage: 90-260v 50-60 HZ
Power Factor >0.95
THD: <10%

And then there is a sticker on it in Chinese (Guessing Chinese not some other asian language). It also says that the ballast has a 5 minute delay on power failure, so I may have been trying to flip it off and on before I had waited the full delay, thus delaying again?

The bulbs I bought are UltraSun 7,500k 400w Probe Start MH. Says only use in M59 ballast.

I have since had no problems with the bulbs firing by way of the timer, haven't tried to fire them manually. I'm about to wire this last one up to a mogul socket and see if I can't use it in a stacked donut setup.

 

[Jhhnn]

Hey JH. I know one thing for sure, after going to school for 6 years, Power factor is not capacitance. It's a determination of how efficient the load/circuit is operating. It's determined by taking everything into account. By altering the capacitance, you alter the phase relationships that affect how efficient the load is operating. Any of the elements involved in the circuit can affect the power factor, according to the math. I sat there, in class, and did all the calculations. The lamp for instance, can affect the power factor, even if the capacitor is fine...etc...and these are just simple circuits. Deductive reasoning isn't all that objective. ie: If this is good, than this has to be good...

That's why mechanics test capacitance and components with meters; that's how it's done in the field...and not the marijuana field.


- So I guess my point was, you either measure the component directly...or you guess.

Or you test system performance first. Like I said, the PF can't be right if the system is wrong. If the system operates at a poor PF with a known good bulb, the capacitor value has changed. The transformer values can't change other than dramatically- good, open, or short.

I recently tested a batch of CMH lamps, both 330w & 860w because they were beat up in shipment. I also tested some other lamps at the same time to see if the PF changed at all. 400w MH &330w CMH yielded the same PF, .98 on that particular ballast. 860w CMH, 1000w HPS, dual arc, MH & MH conversion lamps also tested the same, .97 on a given 1000w switchable ballast that received a new capacitor in order to work at all.

How do you test a capacitor with an internal discharge resistor? There's only one way- operationally.

I worked on exotic electrical equipment in light rail vehicles for decades- it was my specialty. DC from 5v to 750v. Chopped high voltage DC. 3 phase AC. Variable frequency 3 phase AC. Exotic relay logic. PLC's. Special purpose industrial computers. High current inverters & converters. 10 ton HVAC systems. I passed the test to wire my own house & the work passed inspection. I know how to troubleshoot electrical problems. Start with the easy stuff, and testing the PF is as easy as it gets.

 

[Jhhnn]

So after looking around some more, one of the ballasts has a electric diagram on it for wiring. The ballasts are electronic and it says "Blue Line Electronic Metal Halide Ballast 400w" on the top.

Input Voltage: 90-260v 50-60 HZ
Power Factor >0.95
THD: <10%

And then there is a sticker on it in Chinese (Guessing Chinese not some other asian language). It also says that the ballast has a 5 minute delay on power failure, so I may have been trying to flip it off and on before I had waited the full delay, thus delaying again?

The bulbs I bought are UltraSun 7,500k 400w Probe Start MH. Says only use in M59 ballast.

I have since had no problems with the bulbs firing by way of the timer, haven't tried to fire them manually. I'm about to wire this last one up to a mogul socket and see if I can't use it in a stacked donut setup.

Given that the ballasts are electronic, they're basically non-repairable outside of an electronics shop. Even then, it's not likely to be cost effective to even try except by techs very experienced with that particularly bit of kit. As you've discovered, it's very important to understand what it's supposed to do, simply because digital stuff can be programmed to do things you might not expect, like having a hot strike/ re-strike delay.

 

[rives]

The power factor method that you are describing assumes that the power system that you are using has a high power factor to start with. That would normally be true in a residential setting, but isn't a given. A base line measurement of the system power factor and then a measurement of the ballast's impact on it would be more helpful.

 

[Jhhnn]

The power factor method that you are describing assumes that the power system that you are using has a high power factor to start with. That would normally be true in a residential setting, but isn't a given. A base line measurement of the system power factor and then a measurement of the ballast's impact on it would be more helpful.

Technically correct. Plug in an incandescent lamp. The PF should be 1. Chances of finding anything else in a residential setting are very slight.

 

[rives]

Very slight?

It depends on what type of loads are normal for your house, your neighborhood, and your proximity to industry. House fans have a pf of .5-.8, cfl's around .6, a heat pump .8, a clothes washer .5 - .6, computer monitors & televisions .6, well pump .8, etc, etc.

 

[Jhhnn]

Very slight?

It depends on what type of loads are normal for your house, your neighborhood, and your proximity to industry. House fans have a pf of .5-.8, cfl's around .6, a heat pump .8, a clothes washer .5 - .6, computer monitors & televisions .6, well pump .8, etc, etc.

Meh. I just tested it again. There's a 1/4 hp whole house fan running & a 1/4 hp swamp cooler. Opened the fridge & the freezer to make sure they're running, Turned the forced air furnace to vent, & turned on my 315w radial arm saw which has a terrible PF of .23. Plugged into the same 20A receptacle as the running saw, my heat lamp has a PF of .99. Plugged into the same circuit across the room, it's 1, as it should be. I could have added more inductive load power tools & a couple of vacuum cleaners, but I doubt it would have mattered much. It's a warm day, so my neighbors' air conditioners are probably running, too. The electrical distribution net in my neighborhood is ancient, although there is a substation nearby.

Go figure.

Yeh, sure, test a resistive load to establish a baseline for your home, but it likely won't make any difference considering what you're looking for when testing a HID fixture, which is a PF less than .90.

 

[rives]

turned on my 315w radial arm saw which has a terrible PF of .23. Plugged into the same 20A receptacle as the running saw, my heat lamp has a PF of .99.

That's strange. I'm traveling so I don't have mine available to play with, but I wonder how they are correcting the power disruption from the saw to arrive at what the heat lamp is doing?