Running CFLs from inverters.

[Clive Mitchell]

Are there any known pitfalls of running a parallel circuit of several
standard compact fluorescent lamps from one of those cheap 12V to mains
voltage convertors sold for use of small appliances in cars and motor
homes?

I'm guessing the biggest risk of damage comes from either the peak
voltage charging the internal electrolytics in the lamps to a higher
than normal voltage, or even the strange load of rectifiers and
electrolytics causing the converters problems.

I've got a tricky situation where a moving set piece which was
originally going to be powered from the mains is now to be self
contained. The last thing I want is to blow a dozen compact
fluorescents at the flick of a switch.

 

[Victor Roberts]

Are there any known pitfalls of running a parallel circuit of several
standard compact fluorescent lamps from one of those cheap 12V to mains
voltage convertors sold for use of small appliances in cars and motor
homes?

I'm guessing the biggest risk of damage comes from either the peak
voltage charging the internal electrolytics in the lamps to a higher
than normal voltage, or even the strange load of rectifiers and
electrolytics causing the converters problems.

I've got a tricky situation where a moving set piece which was
originally going to be powered from the mains is now to be self
contained. The last thing I want is to blow a dozen compact
fluorescents at the flick of a switch.

Don't most inexpensive inverters have _lower_ peak voltage for the
same RMS voltage because their output is more square than sine wave?
This could still cause problems - just different ones .

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

 

[Nick Hull]

Are there any known pitfalls of running a parallel circuit of several
standard compact fluorescent lamps from one of those cheap 12V to mains
voltage convertors sold for use of small appliances in cars and motor
homes?

Part of my emergency lighting is just that, a cheap inverter running
100% of the time off Golf cart batteries and a bunch of CF lights, some
of which come on automatically after 5 seconds of power failure. Been
using it for years, no problem.

I use an OLD 300 watt inverter without a fan, I also have a newer 200
watt inverter with a noisy fan. I have the fan wired so it is OFF
unless there is a power failure. I don't want the fan noise or battery
use when there is little load on the inverter, and am also concerned the
fan would wear out if run 24/7/365.

 

[Clive Mitchell]

Don't most inexpensive inverters have _lower_ peak voltage for the same
RMS voltage because their output is more square than sine wave? This
could still cause problems - just different ones .

Hmm, you might be right if it's putting a fairly square wave out. That
said, CFL's seem to work well at a slight reduction in voltage, with a
slightly longer start delay and fractionally lower light output. The
life of the lamps is not an issue here since they will only be lit for
about an hour and then the prop will be consigned to the scrap heap.
Yes wasteful, but that's just what happens with single use props. They
have a short but glorious life.

 

[Clive Mitchell]

The rectifiers and caps deal with it no problem. The converter may not
run quite its rated load in lamps before tripping out.

Thanks. I was going to over rate the converter by a factor of two
anyway.

 

[Zak]

Are there any known pitfalls of running a parallel circuit of several
standard compact fluorescent lamps from one of those cheap 12V to mains
voltage convertors sold for use of small appliances in cars and motor
homes?

I've seen a CFL rated for 12 volts. Normal Edison base. Funny


Thomas

 

[Victor Roberts]

I've seen a CFL rated for 12 volts. Normal Edison base. Funny

Used for boats, RVs, and solar powered homes.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

 

[Clive Mitchell]

I've seen a CFL rated for 12 volts. Normal Edison base. Funny

Even funnier if you screw it into a 240V socket.

 

[Clive Mitchell]

Used for boats, RVs, and solar powered homes.

I bet they cost a fortune then.

In the application I have I need the candle shaped lamps which are
probably too exotic to be available in 12V versions.

(Otherwise it would have simplified things a bit.)

 

[JM]

quoting:

Down to quite a lot lower.

I've found 240V 8W lamps to start to start reliably and light on 120VDC.
(will of course vary with brand)

You can modify an inverter to output DC. Either by adding an exteranal
recifier, or by removing the "inverter" part so that it outputs DC. Works
wonderfully for operating CFLs.

 

[Don Klipstein]

The peak voltage is the same, as the waveform is not a square wave, but
a 'modified sine wave'.
This has a peak of the same peak voltage as normal AC, lasting for a few
ms, then drops to zero, stays at zero for a few ms, then goes to peak
voltage in the opposite direction.

I examined one of those "modified sine wave" inverters once. The
waveform was positive or negative 75% of the time and zero 25% of the
time. (37.5% of the tme positive, 12.5% of the time zero, 37.5% of
the time negative, 12.5% of the time zero. )That would make the peak
voltage SQR(4/3) of the RMS, or 1.155 times the RMS. With a true sine
wave, the peak is 1.414 times the RMS.
I believe the 75% duty cycle was chosen because that was close to giving
a relative minimum of some combination of the third and fifth harmonics
without making the higher harmonics too much worse than with a square
wave.
A duty cycle of 2/3 would eliminate the third harmonic but give more of
the fifth and most other harmonics if I figure right. And that would make
the peak voltage 1.225 times the RMS voltage.
If I figure right, a duty cycle of 60% would give a peak/RMS ratio of
1.291 and no fifth harmonic but most other harmonics would be worse.
A duty cycle of 50% (25% positive, 25% zero, 25% negative, 25% zero)
would give the same peak/RMS ratio as a sinewave, but the harmonic content
would be greater than that of positive-zero-negative-zero waveforms with
less off time and the same RMS voltage. Some appliances, such as many
motors and lamps with magnetic ballasts, will be "undervoltaged" because
the RMS voltage of the fundamental frequency component of the output
voltage waveform will be significantly lower than the RMS voltage of the
output including the harmonics.

- Don Klipstein ([email protected])

 

[JohnR66]

quoting:


You can modify an inverter to output DC. Either by adding an exteranal
recifier, or by removing the "inverter" part so that it outputs DC. Works
wonderfully for operating CFLs.

Adding an external rectifier would not make sense. It is already inside the
lamps. Tapping the inverter before the chopper transistors to get the
rectified and filtered DC would be a good solution if one wants to tinker
inside the inverter.

I would add a ceramic resistor of a few ohms. This would reduce the peak
current significantly, thus spredading out the charge time of the capacitors
inside the lamps across the wave peaks (or edge). Keep in mind that many
CFLs already contain a supply resistor for this very purpose. Nothong at all
may need to be done.
John

 

[JohnR66]

I examined one of those "modified sine wave" inverters once. The
waveform was positive or negative 75% of the time and zero 25% of the
time. (37.5% of the tme positive, 12.5% of the time zero, 37.5% of
the time negative, 12.5% of the time zero. )That would make the peak
voltage SQR(4/3) of the RMS, or 1.155 times the RMS. With a true sine
wave, the peak is 1.414 times the RMS.
I believe the 75% duty cycle was chosen because that was close to giving
a relative minimum of some combination of the third and fifth harmonics
without making the higher harmonics too much worse than with a square
wave.
A duty cycle of 2/3 would eliminate the third harmonic but give more of
the fifth and most other harmonics if I figure right. And that would make
the peak voltage 1.225 times the RMS voltage.
If I figure right, a duty cycle of 60% would give a peak/RMS ratio of
1.291 and no fifth harmonic but most other harmonics would be worse.
A duty cycle of 50% (25% positive, 25% zero, 25% negative, 25% zero)
would give the same peak/RMS ratio as a sinewave, but the harmonic content
would be greater than that of positive-zero-negative-zero waveforms with
less off time and the same RMS voltage. Some appliances, such as many
motors and lamps with magnetic ballasts, will be "undervoltaged" because
the RMS voltage of the fundamental frequency component of the output
voltage waveform will be significantly lower than the RMS voltage of the
output including the harmonics.

- Don Klipstein ([email protected])

The Radio Shack 140w unit I have seems seems to lengthen the duty cycle when
the supply voltage drops as a form of regulation. At 9volts in, the lowest
voltage it runs before shut down, it is practically a square waveform.
John

 

[Clive Mitchell]

The Radio Shack 140w unit I have seems seems to lengthen the duty cycle
when the supply voltage drops as a form of regulation. At 9volts in,
the lowest voltage it runs before shut down, it is practically a square
waveform. John

As a foot note to this thread, I ended up using a standard off the shelf
UPS for the job. It powers the 20 candle shaped 5W CFL's (Philex) for
25 minutes when it has been fully charged. (Eight hours or more.)

The UPS (Uninterruptable Power Supply) I'm using is a Trust 800VA unit
which contains a modest 12V 9AH battery and a single chunky transformer
to both charge the battery and convert the 12V back to 230V. (Via a well
stuffed PCB.)

It's interesting to note that the manufacturers of UPS units tend to be
rather ambitious with the backup capacity ratings. They use the
disclaimer that the unit will last "up to 60 minutes depending on load".
In reality the units will only last about five minutes on a full load
but will probably last 60 minutes with a few CFL's connected. In my
case I only need about 15 minutes of illumination for the show, so the
25 minutes are a reasonable safety margin.


The internal design suggests that you could easily take fly leads out to
an external SLA battery pack for greater capacity. I think that I'd
treat the battery leads as being at possible mains potential given the
nature of the circuit and no schematic. This would mean that a single
compact unit would act as a battery charger, step-up convertor and
controller in a single unit. Much more cost effective than the discrete
components.

For this application I'm keeping the units as-is with a secondary unit
for back up in case of unit failure or human failure. (Forgetting to
charge the unit.)