From: John De Armond
Subject: Re: Dry camping and lights...
Date: Thu, 19 Sep 2002 11:42:17 -0400
On Thu, 19 Sep 2002 00:19:39 -0800, Curt Martin <CurtM@GCI.NOSPAM.net> wrote:
>I've read a few things lately on different RV sights about swapping from
>the regular incandescent lights to flourescent. Most places that sell
>the flourescents claim power usage benefits just short of actually
>charging your batteries while using them.
They pretty much do what is claimed. Just compare the power spec to that of
an incandescent bulb.
The only thing I don't like about the fluorescents I've looked at are that
they all seem to be equipped with cool white lamps. This is a bluish white
light and makes the RV seem cool and out of place in the woods. Warm white
lamps are still too cool for my tastes. Compact fluorescents all pretty much
use a very warm phosphor that mimicks the color temperature of incandescent
lamps. I've seen 12v CF fixtures advertised on alternative power sites. I
have not found a conventional fluorescent bulb in this phosphor.
What I've done, given that I have a neon shop, is make some lights using the
same phosphor as the CF. The advantage to neon is that the tube lasts forever
and I can bend it to fit the spaces. I plan to do little miniature cove
lighting when I get a round tuit. Neon (actually cold cathode fluorescent) is
a little more efficient than conventional fluorescents because the electrodes
remain cool. If you're serious about power conservation, you might
investigate local neon shops to see if anyone can make you some tubes. The
tubing I use is a tri-phosphor by Tecnolux and is called "incandescent".
From: John De Armond
Subject: Re: Wierd Electricity
Date: Wed, 17 Mar 2004 23:07:46 -0500
Yup. Those fluorescent fixtures put a LOT of high frequency noise (EMI) on
the 12 volt line. This EMI messes with other electronic gadgets such as your
alarm. When the bulb starts dying the required voltage rises. This is
reflected in even more noise on the line.
I've been chasing that problem, only with new fixtures, in my rig. I recently
converted to fluorescent fixtures and the noise scrambled everything from my
CO detector to my E-meter.
The solution is to install hash filters on the power leads inside the
fixtures. And to make sure the fixtures are well grounded. You can buy the
hash filters at Rat Shack for a couple of bux. They're designed to be
installed on the power leads of car radios to cut alternator and ignition
If all your lights are powered from one 12 volt branch circuit (mine are) then
you might get away with a single hash filter on that circuit at the fuse
On 18 Mar 2004 02:17:23 GMT, firstname.lastname@example.org (Janet Wilder) wrote:
>Our fifth wheel is almost 4 years old and has been in continuous use as we are
>fulltimers. The 12V flourescent bulbs are coming to the end of their lives.
>Tonight the wierdest thing happened. The large TV in the living room developed
>static--white stuff on and off. I checked the cables but everything was snug.
> Then Barry noticed that one of the bulbs in the kitchen was faint. He went to
>change the bulb and the propane sniffer began to beep. The light went orange
>and it beeped every 30 seconds. I got out the book and it said it was a
>"malfunction code" I reset it by holding down the button.
>Just to see if the bulb, which was located about 8 feet from the sniffer, was
>the culpret, I had Barry take out the bulbs in the fixture. Shure enough, the
>sniffer went into "malfunction code"
>The book says something about the sniffer needing so many volts to operate.
>Could removing or a defective bulb in a 12V fixture make the voltage drop
>enough to put the sniffer in to malfunction or do we have a problem. (the
>fixture was "on" when the bulbe were removed)
>Checking our electrical panel shows that the batteries are fully charged and
>the DC is 13.0 V, which is normal.
>any one have a clue as to why taking out a light bulb would affect the propane
>The Road Princess
>Spelling and punctuation is up to my editors. I take no responsibility
From: John De Armond
Subject: Re: Fluorescent lamps
Date: Wed, 15 Mar 2006 17:06:10 -0500
On Tue, 14 Mar 2006 08:52:03 -0700, Dapper Dave <email@example.com>
>For some reason, we go through a lot of those bulbs. I have found no
>difference in longevity among the various brands, although some really
>cheap Chinese bulbs I found at an RV supply store once had the most
>unpleasant "cool white" color I have ever seen. More like "steel gray."
Check the voltage to the fixture with the lamp on. The fluorescent
fixtures I have (that big name brand but I can't recall it)
under-drive the filaments. This results in short life and blackened
ends. The voltage should be AT LEAST 12 volts and preferably more.
From: John De Armond
Subject: Re: Fluorescent lamps
Date: Thu, 16 Mar 2006 02:22:03 -0500
On Wed, 15 Mar 2006 16:09:55 -0700, Dapper Dave <firstname.lastname@example.org>
>>Neon John <email@example.com> wrote:
>>On Tue, 14 Mar 2006 08:52:03 -0700, Dapper Dave <firstname.lastname@example.org>
>>>For some reason, we go through a lot of those bulbs. I have found no
>>>difference in longevity among the various brands, although some really
>>>cheap Chinese bulbs I found at an RV supply store once had the most
>>>unpleasant "cool white" color I have ever seen. More like "steel gray."
>>Check the voltage to the fixture with the lamp on. The fluorescent
>>fixtures I have (that big name brand but I can't recall it)
>>under-drive the filaments. This results in short life and blackened
>>ends. The voltage should be AT LEAST 12 volts and preferably more.
>That's probably it. We dry camp a lot, and I have the generator set to
>start up at 11.8 volts. It runs too often when set to come on sooner.
>Those are Thinlite fixtures, by the way.
>Thanks for explaining the mystery.
Yeah, those are the ones I have. You might want to look into getting
a "voltage stabilizer". This is a gadget that takes in battery
voltage and outputs a constant 13.8 volts (or whatever you set it to).
These have become very popular with the boom-boom stereo types so you
should be able to find one at your local car stereo shop. Or a lot
cheaper on the net.
A DC/DC power supply with 12 in/12 out would do the job too and might
be cheaper if you feed only your lighting circuit. Acopan and
Astrodyne are two brands that I know are good.
Any of Astrodyne's SD series, suffix A (9-18 volt input) will work.
Here is the 50 watt page, first table:
I use SD converters in my EV work to step the high voltage traction
battery voltage (72-250 volts) down to 12 volts for the accessories
and am very satisfied. They withstand pretty much any charging abuse
I can throw at the pack and keep on truckin'. This unit could easily
be mounted on your converter and connected to feed your lighting
circuit. They make the SD line in up to 200 watts.
One automotive-oriented voltage stabilizer that I know of is the
Jacobs FR-1500. Look here and click "accessories".
It's a 100 amp unit and costs about $400. I'm not recommending it, as
I've never tested one and other Jacobs products that I've had in my
lab have not fared well. But it's an example of what to look for.
I have the same problem as you do with my ThinLines because I dry camp
a lot. I've been thinking about just building a simple DC/DC
stabilizer. It's a very simple circuit for the low currents we're
talking about. One approach that I've been thinking about is adding
another winding to the transformer inside my inverter, scaled for 14
volts after rectification.
Unfortunately it's in a long line of things waiting for Round Tuits.
I'll probably just pop for an Astrodyne.
I've spent a little time analyzing the inverter inside the ThinLite
fixture with the notion of adding a turn or two to the filament driver
windings but again, a shortage of Round Tuits has stopped progress....
From: John De Armond
Subject: Re: LED light results -- short
Date: Sun, 02 Apr 2006 23:25:27 -0400
On Sun, 02 Apr 2006 14:38:40 GMT, GaryO < @ . > wrote:
>On Thu, 30 Mar 2006 00:14:39 -0500, Neon John <email@example.com> wrote:
>>I have a nifty little 7 watt 12 volt-operated, bayonet-base CF that
>>was sent to me for review (by an importer that went out of business,
>>unfortunately) It outputs many times more light than 7 watts' worth
>>of LEDs. And it's a much more pleasing light, having a spectrum
>>similar to incandescents.
>This sounds like just what I've been needing!! What are the
>dimensions of it? Do you know if anyone else offers it?
It's a spiral tube lamp. Overall length is about 3" and the spiral
diameter is about 1.5". It won't fit in the RV standard cheezy square
flat plastic fixtures but it would fit in a nicer wall fixture where
the bulb is under a shade or whatever.
I don't know where one could get this exact bulb but there is a flood
of ChiCom 12 volt CFs available now. Google for "12 volt CF lamp" and
enjoy the flood :-)
Here's the first hit:
This lamp looks exactly like mine except that mine has the
automotive-style bayonet base. I'll bet that the bayonet base is an
>I've been considering taking one of the inexpensive 7W Coleman battery
>operated fluorescent lamps and gutting it, to use for RV interior
>lighting. I'd like to fit it inside my existing incandescent (#1141
>bulb) fixtures. (I don't like the large size of the thinlites, and
>their tiny power switches.) Since it is intended to run from D cells,
>it should have pretty decent overall efficiency, I'd just need to
>adapt it to 12V from its stock 6V. The problem that I've encountered
>at the moment is that the bulb is just a tad too long for my fixture.
I've noticed at Wallyworld there are several types of replacement
lamps for those lanterns including a spiral lamp.
The problem with those lanterns is that the driver is crap, to use a
technical term. There are two major issues with electronic drivers -
filament drive and mercury migration. The filament drive must be
adequate or else the electron emission will be too low, positive ions
will collide with the filament and knock off bits of same. The result
is blackening of the tube ends and short life. Sometimes very short
Mercury migration is a phenomenon that happens when there is a little
bit of DC bias on the AC driving signal to the lamp. This can come
from several sources - asymmetrical waveform, inadequately driven
filaments, resulting in one end of the tube acting as a rectifier,
actual DC bias from the driver, etc. What happens is that all the
mercury in the tube gets driven to one end, the end that is the most
negative. That end of the tube stays bright but the rest of the tube
gets progressively dimmer and dimmer.
I have several of those Coleman lanterns (I guess that it's a
sorta-hobby of mine accumulating interesting things that give off
light :-) All but one (out of production for a long time) have one or
both of these problems. Neither seem serious in the original
application simply because the lantern can't be used for long
intervals because of the limited life of the batteries. Hook that
sucker up to a source of power, however, and the problem gradually
One of the things that impressed me about the ThinLites is that
they've mostly avoided the filament drive problem and have completely
avoided the merc migration problem with a very simple ballast circuit.
I don't like the ThinLites for the same reasons but given that at 6'7"
tall, my head scrapes the ceiling as it is, having a CF sticking down
just doesn't cut it :-)
If you can't make a 12 volt CF fit then you could drive an "organ
tube" CF with a 12 volt neon transformer. This guy makes the best
ones available, IMO.
That "midget" unit, the first of the 12 volt units, would do the job.
The 30ma current that standard neon uses is about 1/10th that of CFs
so the brightness would drop accordingly. OTOH, 3KV will run many
tubes in series. Tech 22 is a small company and the guy who owns it
(can't pop his name out of the memory banks at the moment) will
willingly do small custom jobs such as supplying more current. I
think he has 60 and 120 ma units that are sorta non-catalog but also
sorta off-the-shelf items.
I provide this last bit as an option but I'd personally much prefer
the 12 volt CFs. I'd put forth a lot of effort to make one adapt.
One other thought. In my last article I mentioned that Welch Allyn
SolArc HID lamp that I'm using on my scooter for a headlight and as a
reading light in my MH. WA also makes a "bare bulb" version. Here is
the web page:
Buying from WA, just the bulb costs more than my entire light from
TrailTech. The reason I mention this bulb is that it has become quite
popular for diving lights. It has the light output of a 55 watt
quartz halogen with only 13 watts' draw (this includes the 10 watts to
the bulb and the ballast overhead). I can't point you to a URL off
the top of my head but I've seen this bulb offered for sale as a
replacement dive light part in the $50 range.
The arc tube in that lamp is barely the size of a grain of rice. The
whole lamp is probably less than an inch long. The light is
dazzlingly blue-white. There is some UV content which makes text
really pop off the page of a book or magazine. There is very little
heat produced. I doubt that the outer envelope gets anywhere near as
hot as an incandescent bulb.
the ballast is about the size of 4 or 5 quarters stacked up. Tiny
From: John De Armond
Subject: Re: LED light results -- short
Date: Tue, 04 Apr 2006 00:16:58 -0400
>>The problem with those lanterns is that the driver is crap, to use a
>>technical term. There are two major issues with electronic drivers -
>>filament drive and mercury migration. The filament drive must be
>>adequate or else the electron emission will be too low, positive ions
>>will collide with the filament and knock off bits of same. The result
>>is blackening of the tube ends and short life. Sometimes very short
>Do you think the ballasts from the commercial 12V CFs are any better
>in this regard?
I have limited experience with the 12 volt ones but in general, yes I
think they will be. Judging by having taken apart and
reverse-engineered several brands of CFs, it appears that the ChiComs
spend considerable effort on developing a reference design for
electronics like this and then a zillion shops clone them. Almost all
the 120 volt CF ballasts that I've looked at have been identical - and
very clever in the circuit design too. I'd expect the 12 volt CFs to
>>I have several of those Coleman lanterns (I guess that it's a
>>sorta-hobby of mine accumulating interesting things that give off
>>light :-) All but one (out of production for a long time) have one or
>>both of these problems. Neither seem serious in the original
>>application simply because the lantern can't be used for long
>>intervals because of the limited life of the batteries. Hook that
>>sucker up to a source of power, however, and the problem gradually
>I was checking out a lantern at Kmart, a 'Northwest Territories' or
>some such brand. It uses a 7W U-tube, 4 D-cells (batteries not
>included), was only $13, and it included an external power adapter - I
>forget if it was a 12VDC or 120VAC adapter.
For that little bit of money, you could buy one, hook it up at home
and let it run a thousand or so hours and see what happens.
One other thing to keep in mind. The UV spectrum from mercury shifts
a bit with temperature. Nisha (sp?) (one of the worldwide leaders in
rare earth phosphor production) has optimized their phosphor for the
spectrum produced by mercury at about 50deg C, about what a CF tube
runs at after it is warmed up. The tube is not nearly as efficient at
other temperatures. What this means is that if the tube isn't to the
proper temperature, the light output will suffer.
I doubt that this NW Territories unit gets the tube hot enough. None
of my Colemans do. There is a significant light output penalty as the
result. I've put some thought into ways of addressing this problem.
Since I blow glass, one option would be to make a glass shell to go
around the lamp and pull a vacuum in it. Another option might be to
get a smaller tube and let the ballast drive it hard. Or, of course,
you might find the existing light output adequate and just use it
>>One of the things that impressed me about the ThinLites is that
>>they've mostly avoided the filament drive problem and have completely
>>avoided the merc migration problem with a very simple ballast circuit.
>>I don't like the ThinLites for the same reasons but given that at 6'7"
>>tall, my head scrapes the ceiling as it is, having a CF sticking down
>>just doesn't cut it :-)
>Then you wouldn't like my trailer either - 6'4" interior headroom, and
>less under the AC and light fixtures! The thinlites are fine for
>under cabinet use, but I wouldn't want to bang my head on the sharp
>corner of one. I also don't want to fumble around in the dark feeling
>for the tiny switches on them.
>What about the supplies which the kids use for all the accessory
>lighting on their cars? Are any of these decent supplies? Do you
>think they would be capable of driving a standard 7W U-tube?
I've had several of these ground effects transformers in my shop,
usually in the hands of a kid wanting a replacement after it burned
out. Every one I've seen is crap. Bad mercury migration. Poor
thermal management (read: melted housings and such). Not something
I'd want to rely on.
>I take it the 'organ pipe' style is just a straight tube. Even the
>smaller ones of these are a bit long for my existing fixtures.
Lights of America makes an outdoor security light that uses this bulb.
The fixture is sold at the big box stores in the $25 range. If the
bulb would fit, then you could either use a 12 volt CF ballast, a 12
volt neon transformer with less light output or a small inverter to
drive the 120 volt one.
>>I provide this last bit as an option but I'd personally much prefer
>>the 12 volt CFs. I'd put forth a lot of effort to make one adapt.
>Ok, sounds reasonable. So maybe I'll try a CF ballast with the
Nothing to lose, though you should match the tube length and diameter
fairly closely since these ballasts operate fairly close to the edge.
Tube diameter has more effect than length as far as operating voltage
>>One other thought. In my last article I mentioned that Welch Allyn
>>SolArc HID lamp that I'm using on my scooter for a headlight and as a
>>reading light in my MH. WA also makes a "bare bulb" version. Here is
>>the web page:
>>Buying from WA, just the bulb costs more than my entire light from
>Ouch! $126 for a bulb!! It better last forever and be immune to
>vibration! Lots of light though.
Yep, the price hurts. It is immune to vibration and the quoted life
is 8,000 hours if my memory serves. Judging from discussions I've
read, the life is actually much better than 8K. WA tends to be
conservative in their ratings.
I REALLY like my $100 Trail-Tech reading light and consider it money
very well spent. For reading and other detail tasks, the bluish white
light with plenty of long wave UV makes seeing much easier. Edge
detection (picking up things, seeing details, etc) is much easier
under this light.
It is NOT good light for general lighting, unless you like very "cold"
and somewhat harsh light. Incandescents and CFs with
Incandescent-spectrum output are much nicer for sitting around and
taking it easy.
Hey, if all else fails, contact me directly. I can probably make you
a custom lamp tube for your fixture. I have some neon tubing that is
coated with a rare earth phosphor that duplicates the spectrum of
incandescent lamps. If you can't see the source, you'd swear that the
light is coming from a bunch of low wattage light bulbs.
From: John De Armond
Subject: Re: Hey, Neon John . . .
Date: Thu, 22 Jun 2006 11:31:03 -0400
On Thu, 22 Jun 2006 03:57:35 GMT, "Gil Faver" <Rowdy'sboss@ND.com>
>I finally got around to looking at your web site.
>What is a "neon fire"?
"Fire" is the term glass workers use for "burner". Saves a syllable,
I guess :-)
>also, does the mercury in the CF lamps return to the mercury
>release device when you turn off the lamp? If so, how?
No. Much of the merc actually binds up with the phosphor, resulting
in a nice uniform distribution throughout the tube.
That's a generic answer. In some tubes where the mfr has bowed to the
mecuriphobia in society today and has tried to minimize the merc
content, there is a merc emitter attached to one or both electrodes.
This is generally an amalgam of some sort. Looks like spongy metal.
This thing CAN reabsorb some merc when the lamp is off. It is driven
right back out again when the tube is turned on and the discharge
heats the emitter. The effect isn't great but it does cause the tube
to burn dim for a bit.
The scarcity of mercury is yet another contributor to short lamp life.
Merc slowly, driven by the ionized environment, binds to both glass
and phosphor. Once bound, it is out of commission as far as UV
emissions go. When the merc runs out the lamp dims dramatically even
though the 'trodes are still in good shape.
>your BBQ pages make me hungry. Are your sauce recipes still
Yup. Who knows? I might get foolish enough to do another restaurant
someday.... My entire recipe portfolio is for sale, though. I can put
you in touch with my agent if you're interested.
From: John De Armond
Subject: Re: Neon John: Is this how Thin-Lite Fluorescent fixtures work?
Date: Fri, 11 Aug 2006 17:25:38 -0400
Yep, a pretty good description. The bit about the Thinlites creating
EMI is also true. There is NO filtering in the standard fixture. I
added simple little EMI filters, available at Rat Shack or equiv, to
mine so that I again could use the AM radio and the 12 volt amplified
The flicker part isn't absolutely true in all cases. True, the glow
discharge does flicker at twice the applied frequency but that doesn't
always translate into flickering light. Some phosphor have long decay
times (decay tails). That is, they glow for a time after the UV shuts
Flicker in a fluorescent light is highly undesirable in industrial and
even home settings because the stroboscopic effect can make moving
objects appear stationary and lead to injury. Therefore fluorescent
lamp mfrs long ago worked up phosphor formulations with sufficient
decay times to suppress most of the flicker.
Phosphors (conventional halophosphates) in the yellow to red end of
the spectrum tend to have the longest decay tails so the warmer
versions flicker less.
Unfortunately the very efficient rare earth phosphors used in CFs and
even some conventional fluorescent lamps all seem to have very short
decay tails so flicker is again present. The frequency of a solid
state CF lamp is too high for us to perceive flicker but it might
interfere with a poorly designed remote control receiver. The light
would have to be pretty intense to punch through the IR filter in
front of the photo-transistor AND punch through the narrow band filter
that follows the photo-transistor. Moving the light just a little
should solve the problem.
On Fri, 11 Aug 2006 10:56:35 -0700, Dapper Dave <firstname.lastname@example.org>
>I saw this is another forum this morning:
>>I'm not intimately familiar with the thin-lite fixtures, but the basic
>>theory goes: If you have AC, run it though a ballast (step-up
>>transformer) to a few thousand volts and excite the gas in the tube.
>>This in-turn excites the phosphor in the tube and that makes the light.
>>The difference in our coaches is they have DC as a source. They run the
>>DC through a "chopper" circuit. This makes pulsating DC. The pulsating
>>DC is square waves instead of a nice sine-wave. Square-waves are
>>probably the noisiest thing on earth when it comes to electronics but
>>make a suitable substitute for an A/C input to a step-up transformer.
>>Here is the source of your noise.
>>The filters would probably be on the output of the chopper. Cleaner
>>sine wave in...cleaner sine wave out.
>>From then on A/C and D/C starterless lights are the same.
>>As a side note, all fluorescent lights flicker as the A/C changes
>>polarity. This flickering frequently drives the Automatic Gain circuit
>>in IR remote controls crazy and can overpower the weak signal from the
>>remote control. That's why the receiver port has a deep red (near infra
>>red) cover. It's to block out as much ambient light as possible. Some
>>are better than others at blocking.
>>> As I understand it, the Thinlites have some sort of
>>>inverter in them to take 12 VDC to 110AC to run the tubes. I had a
>>>problem with noise from a Thinlite interfering with my climate control
>>>thermostat. When I spoke to Thinlite about it they told me that "Sure,
>>>we know all about it and make a unit that has extra filter to solve that
>>>problem. But Monaco won't pay the extra $2/fixture for the extra
>>>So when they rebuilt the one that was interfering with my thermostat, I
>>>had them put in the extra filters and no more problems.
From: John De Armond
Subject: Re: Thin Lite 12VDC fixture fluorescent lamp life
Date: Wed, 13 Aug 2008 21:46:45 -0400
On Wed, 13 Aug 2008 11:20:51 -0400, Al Malone <email@example.com> wrote:
>Neon John wrote:
>> What kind of 12 volt distribution panel does it have?
>> This sounds like classic symptoms of un-filtered DC from a Magnetek or similar
>> cheap distribution panel. Didn't I just write about that very problem just
>The 12V output to the DC distribution buss / fuse panel, where the light
>fixtures are fed from, is directly connected to the battery bank. The
>Xantrex 2000Watt Inverter / 100 Amp charger (single 12VDC output) is
>directly connected to the 12V battery bank (6 volt golf cart batteries
>in connected in series / parallel).
It STILL sounds like a ripple problem. I have one or two people a month write
me asking about Thinline fixtures eating bulbs. It has ALWAYS been a DC power
quality issue. Most of the time, Magnetek converters. In this case you don't
have one so we have to look farther.
That you and Dapper Dave are having the same problem and both of you have the
same Xantrex converter makes me want to look there.
With a NON-true-RMS responding DVM (most all inexpensive ones are like that),
put the meter on the AC range and measure the voltage at the lamp with the
lamp off. The meter will ONLY respond to the AC component and ignore the DC.
If you have more than a tens of millivolts of ripple then that's your problem.
If the Xantrex is outputting unfiltered DC then you may see several volts of
High DC voltage, within reason, doesn't harm the fixture. That is, less than
about 15 volts. There is a twin bulb Thinline fixture in my rig that has been
on constantly (when I'm not using the rig) for over a year. My DC system
consists of 500 amp-hours of 12 volt batteries supplied by an 80 amp
Intellipower. The periodic equalizing cycle that the Charge Wizard initiates
takes the battery bank up to >14.5 volts.
A few months ago I acquired a lot of over 50 of these fixtures, most new but
some used, in a salvage deal. I've had one running on 14 volts ever since as
a life test. I'm curious to know how long the lamps will last in this type of
high frequency fixture. I suspect practically forever. This fixture has been
on since the first of the summer and the tube ends are still entirely white.
If you don't see anything unusual with your DVM, here's another test that you
can do. Turn off the Xantrex and leave a fluorescent fixture on continuously.
Either put a regular charger on the batteries to keep them charged or
periodically turn the light off and the Xantrex on to recharge the batteries.
The idea is to run the light from as pure a DC source as you can. If the ends
haven't started blackening in a few days then you've identified your culprit -
Double ended blackening is almost always caused by inadequate electrode
temperature, resulting in sputtering of the tungsten filament. The ThinLine
fixture uses a very simple but very cleverly engineered inverter ballast that
operates the little transformer in saturation. That results in a constant
filament drive as long as the voltage is anywhere near nominal. What excess
ripple does is bring the operating voltage down much below nominal 120 times a
second (and probably screws up the inverter's operation in other ways). This
results in inadequate filament heating and sputtering.
Blackening on one end only is caused by there being a DC component on the
tube's operating voltage. This results in mercury migration. That is, the
lamp's mercury fill migrating to the end that is more negative with respect to
the other. Mercury migration has been the bane of solid state neon
transformers. The ThinLine inverter ballast looks to have been designed with
avoiding this problem in mind but if one particular capacitor becomes leaky,
it could cause the problem.
In short, single end blackening is a defect in the inverter. Not enough to
cause the unit to quit working but enough to blacken one end of the tube.
Turning the tube around every so often is a satisfactory solution for most
folks. Try to remember to do it fairly often, before the blackening becomes
If you or Dave want to go to the trouble after you've verified that the DC
feeding the fixture is good, feel free to send me the fixture and I'll check
it out. However, if I were a betting man, I'd be betting on DC power quality
issues when both ends blacken.
From: John De Armond
Subject: Re: Thin Lite 12VDC fixture fluorescent lamp life
Date: Thu, 14 Aug 2008 19:47:42 -0400
On Thu, 14 Aug 2008 13:55:00 -0400, Al Malone <firstname.lastname@example.org> wrote:
>The tubes in the troublesome fixtures are turning black on one end only.
> I have replaced the inverter / ballast in that fixture recently with a
>part that I ordered from Arizona Wind & Sun. The original inverter /
>ballast in that fixture failed completely. The replacement board was a
>different design with what appears to me to be a better heat sink design.
>Since reading your post I found that one tube that I had just replaced
>and not discarded yet had one blackened end. I put it back into the
>same fixture and found that it will work when inserted one way and is
>completely dark when reversed.
>Perhaps I will need to try another inverter / ballast in that fixture.
One other thing that I forgot to mention will cause tube blackening on one end
only. If the filament drive on that end is bad then the cold filament will
sputter. That can be because either the ballast drive for that end failed or
because the filament itself is open. Normal bi-pin ballasts won't light a
lamp with an open filament but an electronic one usually will.
The ThinLines that I have contain an inverter transformer with separate
windings for each filament. I've seen other designs where they play games
with LC resonant circuits to get the filaments both lit from one winding.
Seems to me that the extra components would cost more than the extra winding
on the transformer but that's they way they do it.
BTW, I've read great reports on these 12 volt ballasts
They may be a little large for the smaller ThinLines but they'd be quite handy
for putting conventional fluorescent fixtures under cabinets and the like. The
prices seem fairly reasonable. I'm about to order some to evaluate. If they
perform as advertised then I'm going to put some larger fluorescent fixtures
under my cabinets The ThinLines just aren't bright enough for my aging eyes.
>Also, perhaps I need to put a tube reversing task on my monthly
>maintenance task schedule :-)
That would drive me nutz.
Oh, and you're welcome!