Maker Pro
Maker Pro

Limitation to dimming power compact tubes ?

G

George Pontis

Jan 1, 1970
0
Hi all,

Another question about dimming fluorscent tubes, in particular power compacts.

I recently learned that IceCap (www.icecapinc.com) makes a model 660 ballast that
can fire and run a wide variety of fluorescent tubes. They list configurations
with 2, 3, and 4 tubes including Std, HO, VHO, T8 or T12. Up through 16 feet of
tube and 440 watts total. They also list the ballast as being rated for PC (power
compacts) up through 96 watts per tube, which is typically a 34.1" twin Japanese
tube (Panasonic) with 4 pin square base.

It turns out that this ballast also has dimming capability. A common application
is in aquarium lighting with VHO tubes and a dimmer control that mimics a solar
cycle. The dimming capability is unique in a ballast of this power.

For some reason IceCap says that PC tubes are not dimmable. I have measured them
with an ohmmeter and found that they have filaments of similar resistance to
conventional 48" T8, so it seems like they have at least that similarity. Any
thoughts as to what could make them not suited for a dimming application ?

FWIW, I asked a tech support person at IceCap. He reiterated that they are not
dimmable did not know the reason.
 
V

Victor Roberts

Jan 1, 1970
0
Hi all,

Another question about dimming fluorscent tubes, in particular power compacts.

I recently learned that IceCap (www.icecapinc.com) makes a model 660 ballast that
can fire and run a wide variety of fluorescent tubes. They list configurations
with 2, 3, and 4 tubes including Std, HO, VHO, T8 or T12. Up through 16 feet of
tube and 440 watts total. They also list the ballast as being rated for PC (power
compacts) up through 96 watts per tube, which is typically a 34.1" twin Japanese
tube (Panasonic) with 4 pin square base.

It turns out that this ballast also has dimming capability. A common application
is in aquarium lighting with VHO tubes and a dimmer control that mimics a solar
cycle. The dimming capability is unique in a ballast of this power.

For some reason IceCap says that PC tubes are not dimmable. I have measured them
with an ohmmeter and found that they have filaments of similar resistance to
conventional 48" T8, so it seems like they have at least that similarity. Any
thoughts as to what could make them not suited for a dimming application ?

FWIW, I asked a tech support person at IceCap. He reiterated that they are not
dimmable did not know the reason.

All fluorescent lamps, except perhaps pre-heat lamps with built-in
glow starters, are dimmable. Instant start fluorescent lamps have a
rather small dimming range (unless you are willing to accept very
short life) because there is no way to provide external power to their
electrodes.

Advance makes two 100% to 3% dimming ballasts for the FT55W/2G11 power
compact lamp, one for 120 volt operating and the other for 277 volt
operation, so we know these lamps can be dimmed.

IceCap may be telling you that power compacts cannot be dimmed only
because they do not have a dimming ballast for these lamps.

I have found some of the information provided by IceCap to be suspect.
It is rather common to find ballasts that can operate fluorescent
lamps of different types and different power ratings as long as all
these different lamps have the same or approximately the same rated
current. In a fluorescent system the ballast determines the current
and the lamp determines the voltage. Ballasts are current sources or
pseudo current sources and many modern electronic ballasts are
designed so that their output voltage will "follow" the
lamp-determined lamp voltage, so lamps with different power ratings
can be properly operated from the same ballast model - AS LONG AS THEY
HAVE THE SAME OR SIMILAR RATED CURRENTS.

The last time I checked, IceCap was claiming that certain models of
their ballasts can operate lamps that have very different rated
currents. There is no known way for a ballast to determine the rated
current of a fluorescent lamp. Since the ballast is a pseudo current
source, at least some of these lamps are being operated above or below
their rated currents by the IceCap ballast. That may be OK in some
applications, for example, where people want more light from a
particular lamp at the expense of lamp life, but I think this fact
should be disclosed instead of implying that a single ballast model
can properly operate standard, VO and VHO lamps.

--
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.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.
 
G

George Pontis

Jan 1, 1970
0
...

I have found some of the information provided by IceCap to be suspect.
It is rather common to find ballasts that can operate fluorescent
lamps of different types and different power ratings as long as all
these different lamps have the same or approximately the same rated
current. In a fluorescent system the ballast determines the current
and the lamp determines the voltage. Ballasts are current sources or
pseudo current sources and many modern electronic ballasts are
designed so that their output voltage will "follow" the
lamp-determined lamp voltage, so lamps with different power ratings
can be properly operated from the same ballast model - AS LONG AS THEY
HAVE THE SAME OR SIMILAR RATED CURRENTS.

Thanks again Victor, your explanation makes sense and confirms what I was
beginning to suspect. In spite of IceCap's questionable statements, their ballast
remains interesting due to the unique combination of being dimmable and high
power. It is also much less expensive than buying multiple dimming ballasts from
another vendor to fire the equivalent lamp load. My target application with the
34.1" 96W PC lamps seems similar to IceCap's mainstream VHO target using 36" 95W
VHO lamps. I will get one and see how it performs.
 
V

Victor Roberts

Jan 1, 1970
0
Thanks again Victor, your explanation makes sense and confirms what I was
beginning to suspect. In spite of IceCap's questionable statements, their ballast
remains interesting due to the unique combination of being dimmable and high
power. It is also much less expensive than buying multiple dimming ballasts from
another vendor to fire the equivalent lamp load. My target application with the
34.1" 96W PC lamps seems similar to IceCap's mainstream VHO target using 36" 95W
VHO lamps. I will get one and see how it performs.

These are very different lamps.

I don't have data on the 34" 96W power compact, but it seems to be a
longer version of the more common 55W lamp.

The Philips PL-L 55W/4P HF has diameter of 18 mm (between T5 and T6)
an overall length of 542 mm and is rated for 550 ma at 101 volts at
high frequency. Your 34.1" lamp has an overall length of 866 mm,
866/542 is 1.6 and 1.6 x 55 = 88 watts. The power and length do not
scale linearly since the overall length includes fixed length sections
at both ends that do not contribute to the operating voltage - so the
power should scale faster than the length. It is therefore reasonable
to conclude that the rated high frequency current of the 34" 96W PC
lamp is also about 550 ma which would make the operating voltage about
175 volts.

I also don't have data on the 95 watt VHO, but all VHO lamps I am
familiar with have a diameter of T12 and a rated operating current of
1.5 amps. An example is the 116-watt, 48-inch T12 VHO that has a rated
current of 1.5 amps and a rated voltage of 84 volts. Scaling this to
36 inches would reduce the power to 87 watts, which is off in the
wrong direction based on the "dead end" theory I used to explain the
scaling error in the PL-L case. But, if we assume that your lamp is a
T12 VHO rated at 1.5 amps it would have a high frequency operating
voltage close to 63 volts.

The PC lamp has an arc length that is twice as long as the VHO lamp
and a diameter that is about half as big. The double length of the PC
lamp alone would make its operating voltage double that of the linear
VHO lamp. When you add the smaller diameter, I'm surprised the
operating voltage is only about triple.

If this analysis is correct, then operating the PC lamp on a ballast
designed for VHO lamps that is generating 1.5 amps would mean that the
PC is being operated at almost three times its rated current. (Or
course, the output current of the IceCap ballast might drop when the
load voltage is as high as it is in the this PC lamp, and that would
reduce the overcurrent factor.)

When you dim fluorescent lamps their operating voltage rises. If the
voltage compliance of the IceCap ballast is barely sufficient to run
the PC lamp at rated current, that would explain why the ballasts
cannot run the lamps when dimmed.

--
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.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.
 
G

George Pontis

Jan 1, 1970
0
...

These are very different lamps.

I don't have data on the 34" 96W power compact, but it seems to be a
longer version of the more common 55W lamp.

The Philips PL-L 55W/4P HF has diameter of 18 mm (between T5 and T6)
an overall length of 542 mm and is rated for 550 ma at 101 volts at
high frequency. Your 34.1" lamp has an overall length of 866 mm,
866/542 is 1.6 and 1.6 x 55 = 88 watts. The power and length do not
scale linearly since the overall length includes fixed length sections
at both ends that do not contribute to the operating voltage - so the
power should scale faster than the length. It is therefore reasonable
to conclude that the rated high frequency current of the 34" 96W PC
lamp is also about 550 ma which would make the operating voltage about
175 volts.

I also don't have data on the 95 watt VHO, but all VHO lamps I am
familiar with have a diameter of T12 and a rated operating current of
1.5 amps. An example is the 116-watt, 48-inch T12 VHO that has a rated
current of 1.5 amps and a rated voltage of 84 volts. Scaling this to
36 inches would reduce the power to 87 watts, which is off in the
wrong direction based on the "dead end" theory I used to explain the
scaling error in the PL-L case. But, if we assume that your lamp is a
T12 VHO rated at 1.5 amps it would have a high frequency operating
voltage close to 63 volts.

The PC lamp has an arc length that is twice as long as the VHO lamp
and a diameter that is about half as big. The double length of the PC
lamp alone would make its operating voltage double that of the linear
VHO lamp. When you add the smaller diameter, I'm surprised the
operating voltage is only about triple.

If this analysis is correct, then operating the PC lamp on a ballast
designed for VHO lamps that is generating 1.5 amps would mean that the
PC is being operated at almost three times its rated current. (Or
course, the output current of the IceCap ballast might drop when the
load voltage is as high as it is in the this PC lamp, and that would
reduce the overcurrent factor.)

When you dim fluorescent lamps their operating voltage rises. If the
voltage compliance of the IceCap ballast is barely sufficient to run
the PC lamp at rated current, that would explain why the ballasts
cannot run the lamps when dimmed.

I fired up two of the 96W twin power compacts and took some measurements. With
200W into the ballast and a two tube load, it is reasonable to assume that each
tube is operating very close to 96W. I measured a lamp current of .78A, which
should correspond to 123V across the lamp. The diameter of these bulbs is larger
than the 55W, about 22mm, so that correlates with the higher tube current and
reduces the mismatch somewhat. Not that I would want to run these tubes at double
the normal current.

I found a chart showing dimmer capability with the 660 ballast here:

http://www.bluelineaquatics.com/products/dimmer/1000.html.

It indicates that one can use the ballast with dimmer on VHO lamps up to 2 by 72".
If the 72" VHO is a longer version of the 116W 48" T12(a guess) then the voltage
should scale up to the 120V range. So I think that there is still a chance of
getting some utility out of this combination with a dimmer. I ordered one and will
test it with a few lamps that I have available. The worst thing that can happen is
that I auction it off for half price and chalk it up to another "learning
experience".

By the way, whenever I post something about "watts" there is always a question as
to how I made the measurement. And since I am new to measuring discharge lamps,
let me say something about how I did it. For AC input power I use a Kill-A-Watt
power meter ( http://www.p3international.com/products/special/P4400/P4400-CE.html
). For tube current, I used a Tektronix P6302 current probe with AM503 amplifier.
The output was connected to an HP 34401A meter having true RMS response. The
frequency response of the current measurement was limited by the meter but is
given as being at least 300KHz, typically rolling off at 1MHz. The ballast was a
Fulham Workhorse 7, which produced reasonably sinusoidal output currents, though
the envelope followed the AC line above about 50% amplitude.
 
V

Victor Roberts

Jan 1, 1970
0
I fired up two of the 96W twin power compacts and took some measurements. With
200W into the ballast and a two tube load, it is reasonable to assume that each
tube is operating very close to 96W.

A reasonable assumption.
I measured a lamp current of .78A, which
should correspond to 123V across the lamp. The diameter of these bulbs is larger
than the 55W, about 22mm, so that correlates with the higher tube current and
reduces the mismatch somewhat. Not that I would want to run these tubes at double
the normal current.

I didn't understand this data until I read the end of the note where I
see you are using a Fulham ballast, not the IceCap. When properly
configured the Fulham ballast should operate the lamps at or near the
proper power. I will try to find some data for the 96 watt power
compact.
I found a chart showing dimmer capability with the 660 ballast here:

http://www.bluelineaquatics.com/products/dimmer/1000.html.

It indicates that one can use the ballast with dimmer on VHO lamps up to 2 by 72".
If the 72" VHO is a longer version of the 116W 48" T12(a guess) then the voltage
should scale up to the 120V range.

I agree with your comment "a guess". Without an indication of lamp
diameter, the chart of dimming capability vs. length unless we assume
that all VHO lamps are T12 diameter, which I believe is correct. The
HO designation is used with T12, T8 and T5 lamps, but VHO seems to
limited to T12 lamps.
So I think that there is still a chance of
getting some utility out of this combination with a dimmer. I ordered one and will
test it with a few lamps that I have available. The worst thing that can happen is
that I auction it off for half price and chalk it up to another "learning
experience".

By the way, whenever I post something about "watts" there is always a question as
to how I made the measurement. And since I am new to measuring discharge lamps,
let me say something about how I did it. For AC input power I use a Kill-A-Watt
power meter ( http://www.p3international.com/products/special/P4400/P4400-CE.html
).

I have no experience with the Kill-A-Watt, but it seems like it would
be suitable for 60 Hz power measurements.
For tube current, I used a Tektronix P6302 current probe with AM503 amplifier.

I use the same current probe. I assume that when you measure the
discharge current you clip the current probe simultaneously around
BOTH leads connected to one end of the lamp. That gives you a measure
of the discharge current and ignores the electrode heating current.
The discharge current is distributed between the two leads in an
unknown manner, while the electrode heating current is the same in
each lead but with opposite polarity. Therefore, when you measure the
current in both lamp leads at the same time, the electrode heating
currents null to zero while the two discharge currents add to the
proper result.
The output was connected to an HP 34401A meter having true RMS response. The
frequency response of the current measurement was limited by the meter but is
given as being at least 300KHz, typically rolling off at 1MHz.

That is sufficient frequency response for the Fulham ballast, which I
believe operates between 75 kHz and 100 kHz.
The ballast was a
Fulham Workhorse 7, which produced reasonably sinusoidal output currents, though
the envelope followed the AC line above about 50% amplitude.

Please post the lamp current when you get the IceCap ballast.

--
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.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.
 
V

Victor Roberts

Jan 1, 1970
0
I use the same current probe. I assume that when you measure the
discharge current you clip the current probe simultaneously around
BOTH leads connected to one end of the lamp. That gives you a measure
of the discharge current and ignores the electrode heating current.
The discharge current is distributed between the two leads in an
unknown manner, while the electrode heating current is the same in
each lead but with opposite polarity. Therefore, when you measure the
current in both lamp leads at the same time, the electrode heating
currents null to zero while the two discharge currents add to the
proper result.

I just realized that this discussion above does not apply to the
Fulham Workhorse 7 since it is an instant start ballast and therefore
does not provide electrode heating current. It will, however, apply to
the IceCap ballast, which I believe does provide electrode heating.

The Fulham Web site provides a graphic indication that the 96-watt PC
lamp operates at much lower current than the 95 watt VHO lamp. Go to
the ballast selector at www.fulham.com and look up the ballast for a
96-watt T12 VHO lamp. The default selection is Workhorse 7 connected
as shown in Diagram 12. You will see THREE red leads from the
Workhorse 7 connected to the lamp.

Now enter TWIN and then 1x96. The default ballast is the Workhorse 5.
Use the drop down box to select Workhorse 7 - the same ballast
discussed in the previous paragraph. The connection diagram for
96-watt twin-tube lamp is Diagram 12, which shows only ONE of the red
leads being used to operate the lamp. From this we can estimate that
the operating current of the 96-watt T12 VHO lamp is approximately 3X
the operating current of the 96-watt power compact lamp.

It will be interesting so see what current is provided to the 96-watt
PC lamp by the IceCap ballast.

--
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.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.
 
G

George Pontis

Jan 1, 1970
0
I just realized that this discussion above does not apply to the
Fulham Workhorse 7 since it is an instant start ballast and therefore
does not provide electrode heating current. It will, however, apply to
the IceCap ballast, which I believe does provide electrode heating.

Heating current understood. I would clip around both wires for any ballast
measurement, even if it is a bit hard to get them in that current probe jaw.
The Fulham Web site provides a graphic indication that the 96-watt PC
lamp operates at much lower current than the 95 watt VHO lamp. Go to
the ballast selector at www.fulham.com and look up the ballast for a
96-watt T12 VHO lamp. The default selection is Workhorse 7 connected
as shown in Diagram 12. You will see THREE red leads from the
Workhorse 7 connected to the lamp.

Now enter TWIN and then 1x96. The default ballast is the Workhorse 5.
Use the drop down box to select Workhorse 7 - the same ballast
discussed in the previous paragraph. The connection diagram for
96-watt twin-tube lamp is Diagram 12, which shows only ONE of the red
leads being used to operate the lamp. From this we can estimate that
the operating current of the 96-watt T12 VHO lamp is approximately 3X
the operating current of the 96-watt power compact lamp.

We are probably interpreting these diagrams too far beyond their original intent.
Also, I suspect that each red wire is not quite the perfect current source,
independent of the other red wires. In my case I fired two 96W twin lamps with a
Workhorse 7, and it called for two red wires to each lamp. Actual measurement of
power input confirmed that the lamps were operating very close to 96W each. So
according to that datum the difference from the 96W T12 VHO would only be 1.5.

The good thing about computer-controlled dimming is that one can afford to lose a
little dynamic range by never going to full-brightness if that is required to
maintain a safe lamp current. (Remember the story about guitar amp that went to 12
?)

It will be interesting so see what current is provided to the 96-watt
PC lamp by the IceCap ballast.

I will just have to take delivery of the IceCap 660 and repeat these measurements
with various tubes. The delivery is scheduled for Monday. If this thing doesn't
perform I would not be beyond making one for myself. More likely modifying an
existing dimmable ballast.

Geo.
 
V

Victor Roberts

Jan 1, 1970
0
Heating current understood. I would clip around both wires for any ballast
measurement, even if it is a bit hard to get them in that current probe jaw.


We are probably interpreting these diagrams too far beyond their original intent.
Also, I suspect that each red wire is not quite the perfect current source,
independent of the other red wires.

I agree the wires are not completely independent current sources. I
tested another Fulham ballast a while back that was designed for one
or two lamps of the same type. When operating one lamp from one red
lead the lamp current was higher than the current in each of two lamps
of the same type when each was operated from their own single red
lead.
In my case I fired two 96W twin lamps with a
Workhorse 7, and it called for two red wires to each lamp. Actual measurement of
power input confirmed that the lamps were operating very close to 96W each. So
according to that datum the difference from the 96W T12 VHO would only be 1.5.

But your own measurements show 0.78 amps to each of the 96W twin tube
lamps, and we know that VHO lamps are rated for 1.5 amps. So, if the
IceCap ballast is providing full operating current, the IceCap
ballasts will operate the 96W twin tube lamps at 1.92 times the
current provided by the Fulham in the configuration you tested.

--
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.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.
 
G

George Pontis

Jan 1, 1970
0
But your own measurements show 0.78 amps to each of the 96W twin tube
lamps, and we know that VHO lamps are rated for 1.5 amps. So, if the
IceCap ballast is providing full operating current, the IceCap
ballasts will operate the 96W twin tube lamps at 1.92 times the
current provided by the Fulham in the configuration you tested.

Finally some measured data on the IceCap 660-009 ballast. The tests were done with
two lamps. I did not have a dimmer connected. It was noted from the current
waveform that the lamp current was close to 20KHz and had little 60 Hz ripple.

In the first case, I used two 65W, 21" power compact lamps. Total input power to
the ballast was approximately 160W. Lamp currents were 873/835mA RMS. The lamps
were inexpensive pieces sold under the brand "Odyssea". After a few minutes one of
the lamps broke (!) at the glass junction of the two tubes.

The next case was using two 34.1" lamps. The only marking on these lamps was "96W
6700K". Lamp currents were within a few percent of the previous case. Ballast
input power was about 240W.

It would appear that the ballast is a bit weak when it comes to firing T12 VHO
tubes, but not a bad match for the 96W power compacts. I may have a dimmer for
this ballast in a day or two and can see what happens with this load.

George
 
V

Victor Roberts

Jan 1, 1970
0
Finally some measured data on the IceCap 660-009 ballast. The tests were done with
two lamps. I did not have a dimmer connected. It was noted from the current
waveform that the lamp current was close to 20KHz and had little 60 Hz ripple.

In the first case, I used two 65W, 21" power compact lamps. Total input power to
the ballast was approximately 160W. Lamp currents were 873/835mA RMS. The lamps
were inexpensive pieces sold under the brand "Odyssea". After a few minutes one of
the lamps broke (!) at the glass junction of the two tubes.

The next case was using two 34.1" lamps. The only marking on these lamps was "96W
6700K". Lamp currents were within a few percent of the previous case. Ballast
input power was about 240W.

It would appear that the ballast is a bit weak when it comes to firing T12 VHO
tubes, but not a bad match for the 96W power compacts. I may have a dimmer for
this ballast in a day or two and can see what happens with this load.

George

Thanks for the data. Based on your data I would say this ballast would
operate VHO lamps significantly below their rated power, considering
that the rated current of VHO lamps is 1.5 amps and this ballast is
producing only about 850 ma. That's only 57% of the rated current. The
lamp power would be somewhat higher than 57% of rated because the lamp
voltage rises as the current is reduced.

--
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.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.
 
Top