Boost line voltage for LED Florescent bulbs

[Windy Player]

I have a project that requires 3ft florescent tubes and I've purchased some 3ft T8 bulbs for the project. These are hard to find apparently, and my bulbs came from China.
http://www.bestlightingbuy.com/3ft-t8-led-tube.html

They say they operate on ac voltage from 85 - 265, but they don't work at my house with my 120.6 Volt power. I have a Kill a watt and I carefully hooked them up with a Variac and they operate perfectly at 126 volts or above, but below that they blink at about a 1 sec rate - apparently saying that the voltage is wrong. The Kill A Watt and the Variac give me confidence in my conclusions.

So, after getting 4 of these bulbs for about $100, I need to boost the input voltage to use them. They apparently come from China and I can't really find others at a reasonable price so returning them is not really an option.

Can someone help me find a transformer to boost the voltage from 120 to about 130 to run these bulbs? If they operate up to 265, I'm not worried about too much voltage, with in reason. But two bulbs pull about 30 watts, so I'll need a transformer that supplies about 40 watts to be conservative. Obviously, purchasing variacs - I saw a small "cheap" one for $121 - is not viable. I'd like something in the$20 ballpark. I need two transformers, each to run 2 of these bulbs for my project.

Thanks a bunch,
Windy

 

[hevans1944]

You could purchase a low-voltage transformer, commonly called a control transformer, such as the one shown at the link. Wire it up as a "boost" transformer with the secondary in series with the "hot" side of the power line and the primary connected between "hot" and neutral. You need to "phase" the secondary winding properly with respect to the "hot" mains feed so the transformer voltage adds to the mains voltage Get it backwards and it will subtract. Secondary current rating should be at least as great as the lamp current.

Transformers used this way are also sold as buck-boost transformers, presumably with better insulation between the secondary and primary windings, since the secondary is "floating" at the mains voltage. Buck-boost transformers (sold as such) tend to be a little pricey IMO. But Google "buck-boost transformer" to see what is available.

 

[Windy Player]

Thanks much. Looks like exactly what I need at a very reasonable price. I'll get a pair and give them a try. Using the variac and Kill A Watt it should be easy to figure out if the phase is correct to boost rather than buck. Yes I see the standard "Buck-Boost" transformers. They seem to have higher ratings, like 500VA, and look like they are meant to conform to house or industrial wiring standards rather than hobby project wiring practices. I understand that anything that pulls 40 VA is not strictly a "toy", but these look like way overkill for my simple project. Again, my testing methodology should keep me safe as I make sure this works.

Thanks again.
wp

 

[Scotophor]

Are you trying to run those LED tubes directly from mains power? They're meant to replace a fluorescent tube, so should be run from a ballast -- the same ballast type as would be used to drive a fluorescent tube of that size. Fluorescent lamps need a certain minimum voltage to begin conducting, a.k.a. "strike", but once the current starts flowing they require a specific current rather than a specific voltage. The voltage across the tube will change depending upon several variables such as temperature and lamp age. The function of a ballast (whether old-style transformer type or modern electronic type) is to provide the proper strike voltage and running current. Since these LED lamps are meant to replace fluorescent tubes, I would assume that they are designed to mimic the current-dependent operation, so the ballasts will not be adversely affected.

If you try to power a real fluorescent lamp directly from mains, it will likely fail catastrophically, possibly even spectacularly, because the lamp itself has no built-in current-limiting capability; it will try to draw as much current as the mains can provide. While I doubt that running one of these LED tubes directly from mains will result in a very dramatic failure, it is a strong possibility that it will simply not work properly or at the very least not last nearly as long as it should.

 

[hevans1944]

Are you trying to run those LED tubes directly from mains power? They're meant to replace a fluorescent tube, so should be run from a ballast -- the same ballast type as would be used to drive a fluorescent tube of that size ... .

Go to the web site @Windy Player cited and scroll down to see what's inside, and how to connect them in place of a fluorescent lamp. It specifically says you must remove the starter (if present) and short the two wires on an inductive ballast. All the magic, including LED constant-current drivers is already included in the LED replacement for fluorescent lamps. It just needs a minimum voltage to get the ball rolling. Apparently the Chinese vendor was optimistic about the capability of these lamps to work as low as 85 volts. Or, more likely, their quality control sucks.

 

[Windy Player]

Re: Scotophor: Like I said. I've tested these with the Variac and a safety 40 watt bulb as a test. They work as advertised - no ballast - , except they require more than 126 volts to work. Apparently they have a very smart circuit in them that limits the current over a wide range of voltage - 126[sic] > 200+ volts. Smarter than a simple resistor at least. I ordered the 24v boosters. So we'll soon see.

BTW, using the Variac and Kill A Watt sensor, I tested the pair at 127 > 140 volts and the current varies from .333a > .28a. So the control circuit is pulling constant power not constant amps. I'll bet someone makes a LED control chip for just this purpose.

Thanks again,
d

 

[Scotophor]

OK, my mistake. Rather an odd way to design a product though, IMO. Especially the voltage requirement, which is right on the borderline of whether it will work reliably (or more likely, not) in the U.S and Canada.

 

[Windy Player]

Just got my transformers and looks like they work fine.

I'm running them for a few hours and monitoring the temperature of the transformer.
Thanks for your help,
wp

 

[CDRIVE]

You could purchase a low-voltage transformer, commonly called a control transformer, such as the one shown at the link. Wire it up as a "boost" transformer with the secondary in series with the "hot" side of the power line and the primary connected between "hot" and neutral. You need to "phase" the secondary winding properly with respect to the "hot" mains feed so the transformer voltage adds to the mains voltage Get it backwards and it will subtract. Secondary current rating should be at least as great as the lamp current.

Since step-down transformers typically provide higher current in the secondary winding than the primary, I would think that the primary winding is going to present a greater limitation to maximum output current than the secondary would. After all, a transformer wired this way is roughly equivalent to a high impedance element (the primary) in series with a low impedance (the secondary) element.

Chris

 

[hevans1944]

@CDRIVE the transformer primary is wired in parallel with the mains feed, not in series with the secondary. The secondary is wired in series with the "hot" side of the mains feed, in series with the load. The current limitation of this "boost" configuration is determined by the secondary current rating. The circuit costs a little, does a lot, lasts a long time, works just fine.

 

[(*steve*)]

Yes, in this application a very small transformer can be used to boost the voltage to a comparatively large load. Also, the voltage across the transformer will be no more than in normal use so there are no special concerns about insulation within the transformer.

On another topic, if the rating 85-265 is out so far on the low side I fear for what happens at the high end

 

[CDRIVE]

@CDRIVE the transformer primary is wired in parallel with the mains feed, not in series with the secondary. The secondary is wired in series with the "hot" side of the mains feed, in series with the load. The current limitation of this "boost" configuration is determined by the secondary current rating. The circuit costs a little, does a lot, lasts a long time, works just fine.

When wired in boost as you describe the secondary winding is in series with the primary winding. I.E. Primary and Secondary current are equal and in phase.

Chris

 

[hevans1944]

When wired in boost as you describe the secondary winding is in series with the primary winding. I.E. Primary and Secondary current are equal and in phase.

Chris

You are mistaken. The primary of the transformer connects between the "hot" line and "neutral". The secondary of the transformer connects between the "hot" line and the load. The other side of the load connects to neutral.

Depending on how the secondary is "phased" with respect to the "hot" line, the secondary voltage will either add to or subtract from the line voltage. The secondary winding is completely isolated from the primary winding and is not connected in series with the primary winding. The load current flows through the secondary winding, not through the primary winding.

 

[CDRIVE]

You are mistaken. The primary of the transformer connects between the "hot" line and "neutral". The secondary of the transformer connects between the "hot" line and the load. The other side of the load connects to neutral.

The secondary winding is completely isolated from the primary winding and is not connected in series with the primary winding. The load current flows through the secondary winding, not through the primary winding.

Yes, you're correct. I am quite mistaken. I have no idea what I was thinking. Possibly a case of not thinking enough or thinking too much!

There is still one statement (marked in bold) that gives me pause. Wiring like this provides a mechanical (secondary) connection to the hot side of the mains. I would think that isolation from the mains is lost here but then I'm only on my third cup of coffee.

Chris

 

[hevans1944]

... There is still one statement (marked in bold) that gives me pause. Wiring like this provides a mechanical (secondary) connection to the hot side of the mains. I would think that isolation from the mains is lost here but then I'm only on my third cup of coffee.

Chris

Yeah, that statement gave me pause too after I wrote it. I think it is technically true because the primary and secondary, considered alone, are galvanically isolated no matter what external connections are made. Connecting one end of the secondary to one end of the primary does not, per se, invalidate the galvanic isolation between primary and secondary, but it certainly changes the external behavior. You can consider the connection to constitute an autotransformer, which of course is not galvanically isolated. As for losing isolation from the mains... there was never any isolation to begin with. The load is either directly connected to the mains or it is connected to one end of the secondary winding. No isolation in either case.

My only concern with using a step-down transformer in this manner is the insulation between the secondary winding and the steel core. For the usual interleaved E-I lamination shell-type construction, the secondary is usually wound on top of the primary on a central bobbin, with adequate insulation to galvanically separate the primary from both the secondary and the steel core. The galvanic isolation between the secondary and the steel core (which is often connected to earth ground or neutral) is another matter. The secondary insulation needs to be good enough to allow full line potential between the secondary and the steel core. In my experience this is usually the case, but most step-down transformers are not purposely designed to operate the secondary at an elevated potential.

In retrospect, the highlighted phrase adds nothing to the discussion and can safely be ignored lest we wander off into murky waters.

BTW, my wife just presented me with my second cup of coffee... a magical combination of coffee, pure chocolate, coconut oil, and a few drops of stevia, all whipped up in her Vitamix blender. Turns out she put too much stevia in it for her taste, but it suits me just fine.