O.T. Low volate for UPS

I recently bought an APC UPS system.

With no power to it, the voltage of the outlets is 86 Volts.

Is that normal ?

Battery is fully charged.

Andy wrote:
I recently bought an APC UPS system.

With no power to it, the voltage of the outlets is 86 Volts.

Is that normal ?

Battery is fully charged.

Connect a 60W lightbulb.
Make another voltage measurement with the new load in place.

Paul

On 7/12/2016 12:22 PM, Andy wrote:
I recently bought an APC UPS system.

With no power to it, the voltage of the outlets is 86 Volts.

Is that normal ?

Battery is fully charged.

How are you measuring it?
What waveform does the measurement system expect for accurate
readout?
What waveform does the ups supply?

"Andy" wrote in message
...

I recently bought an APC UPS system.
With no power to it, the voltage of the outlets is 86 Volts.
Is that normal ?

What did APC answer when you asked?
--
Don Phillipson
Carlsbad Springs
(Ottawa, Canada)

On 7/12/2016 3:22 PM, Andy wrote:
I recently bought an APC UPS system.

With no power to it, the voltage of the outlets is 86 Volts.

Is that normal ?

Battery is fully charged.


What model is the APC UPS ?

I just checked two units...

Back-UPS CS 500:
----------------
On AC Power - 118.6 VAC
On DC Power - 101.5 VAC

Back-UPS CS 350:
----------------
On AC Power - 119.5 VAC
On DC Power - 92.5VAC

Measured with a DMM on 200VAC range




--
Dave
Multi-AV Scanning Tool - http://multi-av.thespykiller.co.uk
http://www.pctipp.ch/downloads/dl/35905.asp

David H. Lipman wrote:
On 7/12/2016 3:22 PM, Andy wrote:
I recently bought an APC UPS system.

With no power to it, the voltage of the outlets is 86 Volts.

Is that normal ?

Battery is fully charged.


What model is the APC UPS ?

I just checked two units...

Back-UPS CS 500:
----------------
On AC Power - 118.6 VAC
On DC Power - 101.5 VAC

Back-UPS CS 350:
----------------
On AC Power - 119.5 VAC
On DC Power - 92.5VAC

Measured with a DMM on 200VAC range

Tried a measurement here. Surprised but not surprised.

On AC power - both meters read ~120V
- the power company tells me the mast voltage should
be 113V, but we're back to the 120-121V range, which
is close to being too high.

On internal battery - Kill-O-Watt P4400 - reads 120-122V from inverter
- RS 22-168A reads around 100V from inverter

Test load - 40W light bulb and other stray loads
- two computers sleeping (one 5W, one 7.5W)

The Kill-O-Watt has the same kind of front end
as the digital power meter on the outside of
our houses. Whole house power meters uses sigma-delta
converters running at high speed (125KHz to 1MHz range),
and integrate the area under the curve. The processing
is lots of good digital math.

The RS 22-168A on the other hand, converts whatever
AC waveform is coming in, into a DC voltage, and
an ADC running at 1Hz or so takes a reading (probably
using the dual slope method). So it takes one "sniff"
with a sample-and-hold, then does the dual slope conversion.
I don't really think it is "TrueRMS capable". And that's
why the reading is off. On authentic sine waves, the
RS 22-168A and the Kill-O-Watt, agree.

https://en.wikipedia.org/wiki/True_RMS_converter

Owning a Kill-O-Watt is highly recommended for
power-line functions.

I've also tried measuring current flow with my
clamp-on ammeter (Hall probe based), and that suffers
the same problem. Namely, on distorted waveforms, the
readings are way off. The Kill-o-Watt for example, when
measuring the ATX PSU, gets the right answer (~5W) for
computer sleep. The computer with twice as many sticks
of RAM is the 7.5W one, when it is sleeping. These are
reasonable values.

So I really only have one TrueRMS instrument here.

One other thing to note - in my test (using the Test
button on the front of the UPS), the brilliance of the
light bulb is briefly higher than normal for the
first second. Which I've never noticed before. It
looks like the inverter actually overshoots when it
first starts to run. It's a little too quick to
detect with anything but an oscilloscope.

Paul

On 7/13/2016 8:38 AM, Paul wrote:
David H. Lipman wrote:
On 7/12/2016 3:22 PM, Andy wrote:
I recently bought an APC UPS system.

With no power to it, the voltage of the outlets is 86 Volts.

Is that normal ?

Battery is fully charged.


What model is the APC UPS ?

I just checked two units...

Back-UPS CS 500:
----------------
On AC Power - 118.6 VAC
On DC Power - 101.5 VAC

Back-UPS CS 350:
----------------
On AC Power - 119.5 VAC
On DC Power - 92.5VAC

Measured with a DMM on 200VAC range

Tried a measurement here. Surprised but not surprised.

On AC power - both meters read ~120V
- the power company tells me the mast voltage should
be 113V, but we're back to the 120-121V range, which
is close to being too high.

On internal battery - Kill-O-Watt P4400 - reads 120-122V from inverter
- RS 22-168A reads around 100V from inverter

Test load - 40W light bulb and other stray loads
- two computers sleeping (one 5W, one 7.5W)

The Kill-O-Watt has the same kind of front end
as the digital power meter on the outside of
our houses. Whole house power meters uses sigma-delta
converters running at high speed (125KHz to 1MHz range),
and integrate the area under the curve. The processing
is lots of good digital math.

The RS 22-168A on the other hand, converts whatever
AC waveform is coming in, into a DC voltage, and
an ADC running at 1Hz or so takes a reading (probably
using the dual slope method). So it takes one "sniff"
with a sample-and-hold, then does the dual slope conversion.
I don't really think it is "TrueRMS capable". And that's
why the reading is off. On authentic sine waves, the
RS 22-168A and the Kill-O-Watt, agree.

https://en.wikipedia.org/wiki/True_RMS_converter

Owning a Kill-O-Watt is highly recommended for
power-line functions.

I've also tried measuring current flow with my
clamp-on ammeter (Hall probe based), and that suffers
the same problem. Namely, on distorted waveforms, the
readings are way off. The Kill-o-Watt for example, when
measuring the ATX PSU, gets the right answer (~5W) for
computer sleep. The computer with twice as many sticks
of RAM is the 7.5W one, when it is sleeping. These are
reasonable values.

So I really only have one TrueRMS instrument here.

One other thing to note - in my test (using the Test
button on the front of the UPS), the brilliance of the
light bulb is briefly higher than normal for the
first second. Which I've never noticed before. It
looks like the inverter actually overshoots when it
first starts to run. It's a little too quick to
detect with anything but an oscilloscope.

Paul

There are many factors here. One must realize the difference between
inverted DC and pure AC. The difference lies in the waveform.

What we get from the Power Company should be a pure sinusoidal waveform.
What we get from a Power Inverter is the attempt at creating pure
sinusoidal waveform. Usually a square wave generation modified into a
curve.

We have to also understand the phaser relationship of real and apparent
power.

Then we have Loaded and unloaded power sources. As the load impedance
matches the source impedance the voltage on the load will decrease up to
50%. The mark of a good power supply is that as the load impedance
decreases the voltage on the load stays steady and does not decrease.
I learned something interesting during SuperStorm sandy.

I have APC UPS and I also have a Marine Battery and a 12VDC 350W power
inverter. I also have a 150W power inverter built into my Dodge Grand
Caravan.

Here is where is got "interesting". I prefer 13~13W Compact Flourescent
(CF) bulbs. I thought I would use the Marine Battery and Power light in
my residence. It worked fine on one CF bulb but once I added a second
CF bulb, their ability to provide light was greatly inhibited.

Their reliance on a pure sinusoidal AC source had a codependency on
their productivity. Once a second or 3rd bulb was added they created AC
frequency patterns that inhibited their ability to produce light.


--
Dave
Multi-AV Scanning Tool - http://multi-av.thespykiller.co.uk
http://www.pctipp.ch/downloads/dl/35905.asp

On 7/13/2016 10:01 AM, David H. Lipman wrote:


I have APC UPS and I also have a Marine Battery and a 12VDC 350W power
inverter. I also have a 150W power inverter built into my Dodge Grand
Caravan.

Here is where is got "interesting". I prefer 13~13W Compact Flourescent
(CF) bulbs. I thought I would use the Marine Battery and Power light in
my residence. It worked fine on one CF bulb but once I added a second
CF bulb, their ability to provide light was greatly inhibited.

Their reliance on a pure sinusoidal AC source had a codependency on
their productivity. Once a second or 3rd bulb was added they created AC
frequency patterns that inhibited their ability to produce light.

Not clear what you meant by that last sentence. Can you be more specific
about what you mean by "AC frequency patterns?"
If the inverter voltage waveform is independent of load,
I can't imagine adding more lamps making any difference until
the supply voltage changes.

Did you look at the voltage waveform from the inverter?

I'd suggest that you might have cause and effect reversed.
The 100W CFL that I just measured has a power factor of .6
as measured by a kill-a-watt.
But that's not the whole story. Power factor assumes sinewave
current that's out of phase with sinewave voltage. Not sure what
the KAW is measuring or how it decides what math to use to generate
that PF number for non-sinusoidal inputs.

When you drive a CFL with a fast-rise input, you cram much of the energy
into that edge instead of spreading it over a much longer portion
of a real sine wave. That requires a much higher peak current.
Perhaps multiple lamps exceed the peak current
that the inverter can supply causing a voltage drop and consequent
lower light output.

On 7/13/2016 2:09 PM, mike wrote:
On 7/13/2016 10:01 AM, David H. Lipman wrote:


I have APC UPS and I also have a Marine Battery and a 12VDC 350W power
inverter. I also have a 150W power inverter built into my Dodge Grand
Caravan.

Here is where is got "interesting". I prefer 13~13W Compact Flourescent
(CF) bulbs. I thought I would use the Marine Battery and Power light in
my residence. It worked fine on one CF bulb but once I added a second
CF bulb, their ability to provide light was greatly inhibited.

Their reliance on a pure sinusoidal AC source had a codependency on
their productivity. Once a second or 3rd bulb was added they created AC
frequency patterns that inhibited their ability to produce light.

Not clear what you meant by that last sentence. Can you be more specific
about what you mean by "AC frequency patterns?"
If the inverter voltage waveform is independent of load,
I can't imagine adding more lamps making any difference until
the supply voltage changes.

Did you look at the voltage waveform from the inverter?

I'd suggest that you might have cause and effect reversed.
The 100W CFL that I just measured has a power factor of .6
as measured by a kill-a-watt.
But that's not the whole story. Power factor assumes sinewave
current that's out of phase with sinewave voltage. Not sure what
the KAW is measuring or how it decides what math to use to generate
that PF number for non-sinusoidal inputs.

When you drive a CFL with a fast-rise input, you cram much of the energy
into that edge instead of spreading it over a much longer portion
of a real sine wave. That requires a much higher peak current.
Perhaps multiple lamps exceed the peak current
that the inverter can supply causing a voltage drop and consequent
lower light output.


Maybe the term "interference" patterns would be better. Like when you
use a strobe light on a rotating object and wither the light appears to
go in one direction, the other direction or appears to stand still.
https://en.wikipedia.org/wiki/Afshar_experiment

When two dissimilar frequencies "beat". You will have super nodes,
super anti nodes, the frequencies subtract and the frequencies add.

https://en.wikipedia.org/wiki/Interf...ve_propagation)

When I was younger I built a Zenith/HeathKit Oscilliscope. I no longer
have it or another other scope to have examined the wave forms.

--
Dave
Multi-AV Scanning Tool - http://multi-av.thespykiller.co.uk
http://www.pctipp.ch/downloads/dl/35905.asp

mike wrote:
On 7/13/2016 10:01 AM, David H. Lipman wrote:


I have APC UPS and I also have a Marine Battery and a 12VDC 350W power
inverter. I also have a 150W power inverter built into my Dodge Grand
Caravan.

Here is where is got "interesting". I prefer 13~13W Compact Flourescent
(CF) bulbs. I thought I would use the Marine Battery and Power light in
my residence. It worked fine on one CF bulb but once I added a second
CF bulb, their ability to provide light was greatly inhibited.

Their reliance on a pure sinusoidal AC source had a codependency on
their productivity. Once a second or 3rd bulb was added they created AC
frequency patterns that inhibited their ability to produce light.

Not clear what you meant by that last sentence. Can you be more specific
about what you mean by "AC frequency patterns?"
If the inverter voltage waveform is independent of load,
I can't imagine adding more lamps making any difference until
the supply voltage changes.

Did you look at the voltage waveform from the inverter?

I'd suggest that you might have cause and effect reversed.
The 100W CFL that I just measured has a power factor of .6
as measured by a kill-a-watt.
But that's not the whole story. Power factor assumes sinewave
current that's out of phase with sinewave voltage. Not sure what
the KAW is measuring or how it decides what math to use to generate
that PF number for non-sinusoidal inputs.

When you drive a CFL with a fast-rise input, you cram much of the energy
into that edge instead of spreading it over a much longer portion
of a real sine wave. That requires a much higher peak current.
Perhaps multiple lamps exceed the peak current
that the inverter can supply causing a voltage drop and consequent
lower light output.


I'm not following this, but I thought a CFL has its
own inverter in the base ?

http://www.homemade-circuits.com/201...tubelight.html

Paul

On 7/13/2016 5:30 PM, Paul wrote:
mike wrote:
On 7/13/2016 10:01 AM, David H. Lipman wrote:


I have APC UPS and I also have a Marine Battery and a 12VDC 350W power
inverter. I also have a 150W power inverter built into my Dodge Grand
Caravan.

Here is where is got "interesting". I prefer 13~13W Compact Flourescent
(CF) bulbs. I thought I would use the Marine Battery and Power light in
my residence. It worked fine on one CF bulb but once I added a second
CF bulb, their ability to provide light was greatly inhibited.

Their reliance on a pure sinusoidal AC source had a codependency on
their productivity. Once a second or 3rd bulb was added they created AC
frequency patterns that inhibited their ability to produce light.

Not clear what you meant by that last sentence. Can you be more specific
about what you mean by "AC frequency patterns?"
If the inverter voltage waveform is independent of load,
I can't imagine adding more lamps making any difference until
the supply voltage changes.

Did you look at the voltage waveform from the inverter?

I'd suggest that you might have cause and effect reversed.
The 100W CFL that I just measured has a power factor of .6
as measured by a kill-a-watt.
But that's not the whole story. Power factor assumes sinewave
current that's out of phase with sinewave voltage. Not sure what
the KAW is measuring or how it decides what math to use to generate
that PF number for non-sinusoidal inputs.

When you drive a CFL with a fast-rise input, you cram much of the energy
into that edge instead of spreading it over a much longer portion
of a real sine wave. That requires a much higher peak current.
Perhaps multiple lamps exceed the peak current
that the inverter can supply causing a voltage drop and consequent
lower light output.


I'm not following this, but I thought a CFL has its
own inverter in the base ?

http://www.homemade-circuits.com/201...tubelight.html


Paul


Yes, for its electronics. But then it creates an high frequency at a
high voltage to energize phosphorous and "fluoresce".

http://www.edisontechcenter.org/Fluorescent.html
"Electronic Ballasts: The electronic ballasts use semiconductors to
limit power to a fluorescent lamp. First the ballast rectifies the AC
power, then it chops it to make a high frequency for improved
efficiency. The ballast can more precisely control power than a magnetic
ballast but does have a number of problems."

--
Dave
Multi-AV Scanning Tool - http://multi-av.thespykiller.co.uk
http://www.pctipp.ch/downloads/dl/35905.asp

On 7/13/2016 12:57 PM, David H. Lipman wrote:
On 7/13/2016 2:09 PM, mike wrote:
On 7/13/2016 10:01 AM, David H. Lipman wrote:


I have APC UPS and I also have a Marine Battery and a 12VDC 350W power
inverter. I also have a 150W power inverter built into my Dodge Grand
Caravan.

Here is where is got "interesting". I prefer 13~13W Compact Flourescent
(CF) bulbs. I thought I would use the Marine Battery and Power light in
my residence. It worked fine on one CF bulb but once I added a second
CF bulb, their ability to provide light was greatly inhibited.

Their reliance on a pure sinusoidal AC source had a codependency on
their productivity. Once a second or 3rd bulb was added they created AC
frequency patterns that inhibited their ability to produce light.

Not clear what you meant by that last sentence. Can you be more specific
about what you mean by "AC frequency patterns?"
If the inverter voltage waveform is independent of load,
I can't imagine adding more lamps making any difference until
the supply voltage changes.

Did you look at the voltage waveform from the inverter?

I'd suggest that you might have cause and effect reversed.
The 100W CFL that I just measured has a power factor of .6
as measured by a kill-a-watt.
But that's not the whole story. Power factor assumes sinewave
current that's out of phase with sinewave voltage. Not sure what
the KAW is measuring or how it decides what math to use to generate
that PF number for non-sinusoidal inputs.

When you drive a CFL with a fast-rise input, you cram much of the energy
into that edge instead of spreading it over a much longer portion
of a real sine wave. That requires a much higher peak current.
Perhaps multiple lamps exceed the peak current
that the inverter can supply causing a voltage drop and consequent
lower light output.


Maybe the term "interference" patterns would be better. Like when you
use a strobe light on a rotating object and wither the light appears to
go in one direction, the other direction or appears to stand still.
https://en.wikipedia.org/wiki/Afshar_experiment

When two dissimilar frequencies "beat". You will have super nodes,
super anti nodes, the frequencies subtract and the frequencies add.

https://en.wikipedia.org/wiki/Interf...ve_propagation)

When I was younger I built a Zenith/HeathKit Oscilliscope. I no longer
have it or another other scope to have examined the wave forms.

I understand the concept, but don't think it's relevant.
The typical cheapo modified sinewave, aka square wave with some dead
time, uses an inverter to generate a DC voltage somewhere around the
peak of the desired output voltage. It uses a bridge of FETS or
similar to generate the modified sine wave from that DC voltage.
I'm assuming there are various ways to protect the bridge if the
load is too much.

I bought a B&D inverter at the thrift store yesterday.
Quick look suggests that they use a fixed duty factor
that's independent of load. The only thing that appears
to be regulated is the voltage ahead of the FET bridge that
turns it into a bipolar pulse. The datasheet claims it
has soft start technology, suggesting that it somehow
monitors peak current. Not clear how they use that tho...

If your CFL has a diode bridge followed by a capacitor, the input
current with sinewave drive happens over a relatively small portion
of the input sinewave. Power factor is low. When you drive it with
a fast rise pulse from the inverter, the peak current is much higher.
I = C * dV/dT. For a sinewave, the dV/dT is relatively low near the
peak where the diodes conduct. For square wave input, the dV/dT is
WAY higher, so the current is much higher.

I'm suggesting that the inverter can handle the peak current for
one lamp, but not two.

Should be very easy to verify by looking at the waveform.
I expect there's lots of noise on the waveform, but it's unlikely
that is the cause of a major brightness decrease.

David H. Lipman wrote:
On 7/13/2016 5:30 PM, Paul wrote:
mike wrote:
On 7/13/2016 10:01 AM, David H. Lipman wrote:


I have APC UPS and I also have a Marine Battery and a 12VDC 350W power
inverter. I also have a 150W power inverter built into my Dodge Grand
Caravan.

Here is where is got "interesting". I prefer 13~13W Compact
Flourescent
(CF) bulbs. I thought I would use the Marine Battery and Power
light in
my residence. It worked fine on one CF bulb but once I added a second
CF bulb, their ability to provide light was greatly inhibited.

Their reliance on a pure sinusoidal AC source had a codependency on
their productivity. Once a second or 3rd bulb was added they
created AC
frequency patterns that inhibited their ability to produce light.

Not clear what you meant by that last sentence. Can you be more
specific
about what you mean by "AC frequency patterns?"
If the inverter voltage waveform is independent of load,
I can't imagine adding more lamps making any difference until
the supply voltage changes.

Did you look at the voltage waveform from the inverter?

I'd suggest that you might have cause and effect reversed.
The 100W CFL that I just measured has a power factor of .6
as measured by a kill-a-watt.
But that's not the whole story. Power factor assumes sinewave
current that's out of phase with sinewave voltage. Not sure what
the KAW is measuring or how it decides what math to use to generate
that PF number for non-sinusoidal inputs.

When you drive a CFL with a fast-rise input, you cram much of the energy
into that edge instead of spreading it over a much longer portion
of a real sine wave. That requires a much higher peak current.
Perhaps multiple lamps exceed the peak current
that the inverter can supply causing a voltage drop and consequent
lower light output.


I'm not following this, but I thought a CFL has its
own inverter in the base ?

http://www.homemade-circuits.com/201...tubelight.html



Paul


Yes, for its electronics. But then it creates an high frequency at a
high voltage to energize phosphorous and "fluoresce".

http://www.edisontechcenter.org/Fluorescent.html
"Electronic Ballasts: The electronic ballasts use semiconductors to
limit power to a fluorescent lamp. First the ballast rectifies the AC
power, then it chops it to make a high frequency for improved
efficiency. The ballast can more precisely control power than a magnetic
ballast but does have a number of problems."

It chops it at high frequency, so it is above human hearing.
For the same reason the backlight in a laptop runs at
25KHz.

The article here, says it runs 40-50KHz.

http://www.cliftonlaboratories.com/compact_fl.htm

We learned a lot of lessons from television and how 15KHz
(in the flyback) isn't high enough.

As for "The ballast can more precisely control power",
the circuit in that Sylvania bulb doesn't look all that
sophisticated. I think I only see one transistor in there.

And the idea of using a 1/4Watt resistor as a fuse,
now, how clever is that ? What a nice smell.

Paul

On 7/13/2016 6:52 PM, mike wrote:
On 7/13/2016 12:57 PM, David H. Lipman wrote:
On 7/13/2016 2:09 PM, mike wrote:
On 7/13/2016 10:01 AM, David H. Lipman wrote:


I have APC UPS and I also have a Marine Battery and a 12VDC 350W power
inverter. I also have a 150W power inverter built into my Dodge Grand
Caravan.

Here is where is got "interesting". I prefer 13~13W Compact
Flourescent
(CF) bulbs. I thought I would use the Marine Battery and Power
light in
my residence. It worked fine on one CF bulb but once I added a second
CF bulb, their ability to provide light was greatly inhibited.

Their reliance on a pure sinusoidal AC source had a codependency on
their productivity. Once a second or 3rd bulb was added they
created AC
frequency patterns that inhibited their ability to produce light.

Not clear what you meant by that last sentence. Can you be more
specific
about what you mean by "AC frequency patterns?"
If the inverter voltage waveform is independent of load,
I can't imagine adding more lamps making any difference until
the supply voltage changes.

Did you look at the voltage waveform from the inverter?

I'd suggest that you might have cause and effect reversed.
The 100W CFL that I just measured has a power factor of .6
as measured by a kill-a-watt.
But that's not the whole story. Power factor assumes sinewave
current that's out of phase with sinewave voltage. Not sure what
the KAW is measuring or how it decides what math to use to generate
that PF number for non-sinusoidal inputs.

When you drive a CFL with a fast-rise input, you cram much of the energy
into that edge instead of spreading it over a much longer portion
of a real sine wave. That requires a much higher peak current.
Perhaps multiple lamps exceed the peak current
that the inverter can supply causing a voltage drop and consequent
lower light output.


Maybe the term "interference" patterns would be better. Like when you
use a strobe light on a rotating object and wither the light appears to
go in one direction, the other direction or appears to stand still.
https://en.wikipedia.org/wiki/Afshar_experiment

When two dissimilar frequencies "beat". You will have super nodes,
super anti nodes, the frequencies subtract and the frequencies add.

https://en.wikipedia.org/wiki/Interf...ve_propagation)

When I was younger I built a Zenith/HeathKit Oscilliscope. I no longer
have it or another other scope to have examined the wave forms.

I understand the concept, but don't think it's relevant.
The typical cheapo modified sinewave, aka square wave with some dead
time, uses an inverter to generate a DC voltage somewhere around the
peak of the desired output voltage. It uses a bridge of FETS or
similar to generate the modified sine wave from that DC voltage.
I'm assuming there are various ways to protect the bridge if the
load is too much.

I bought a B&D inverter at the thrift store yesterday.
Quick look suggests that they use a fixed duty factor
that's independent of load. The only thing that appears
to be regulated is the voltage ahead of the FET bridge that
turns it into a bipolar pulse. The datasheet claims it
has soft start technology, suggesting that it somehow
monitors peak current. Not clear how they use that tho...

If your CFL has a diode bridge followed by a capacitor, the input
current with sinewave drive happens over a relatively small portion
of the input sinewave. Power factor is low. When you drive it with
a fast rise pulse from the inverter, the peak current is much higher.
I = C * dV/dT. For a sinewave, the dV/dT is relatively low near the
peak where the diodes conduct. For square wave input, the dV/dT is
WAY higher, so the current is much higher.

I'm suggesting that the inverter can handle the peak current for
one lamp, but not two.

Should be very easy to verify by looking at the waveform.
I expect there's lots of noise on the waveform, but it's unlikely
that is the cause of a major brightness decrease.


I don't think the CF uses just a simple bridge rectifier with
electrolytic capacitor DC converter.

The thing is a CF bulb has little load. A 13~14W CF on a 350W inverted
12VDC supply is not going to have a load. Therefore going from 13~14W
CF to 26~28W from two CF bulbs should not go from a condition of light
production to not producing light. Therefore there has to be some other
condition involved. Especially considering I took a 36W LED Rope Light
( Orange LEDs for Halloween ) and also put it on the Inverter and it
worked if I had 1 CF or 2 CF bulbs on the circuit. It was only the two
CF bulbs that interacted with each other.

BTW: It is a Tripp-Lite PowerVerter Ultra-Compact Inverter, 375 Watts
http://www.homedepot.com/p/Tripp-Lit...gcl src=aw.ds


--
Dave
Multi-AV Scanning Tool - http://multi-av.thespykiller.co.uk
http://www.pctipp.ch/downloads/dl/35905.asp

ADDENDUM:
I forgot that after SuperStorm Sandy I had put up a photo on my web site.

http://multi-av.thespykiller.co.uk/v...ter(LoRes).jpg



--
Dave
Multi-AV Scanning Tool - http://multi-av.thespykiller.co.uk
http://www.pctipp.ch/downloads/dl/35905.asp