PDA

View Full Version : Additions for the ballast overdrive post



Greg Moore
January 8th, 2002, 12:01 PM
Since the original post by StirCrazy is getting long I am starting this in a new post.

This is to add additional information to StirCrazy's original post about overdriving electronic ballasts and includes additional info and tests done by myself.

First off, the ballasts I am using are Sylvania model QuickTronic 4x32/120IS Simply put, electronics made to run up to 4 32w T8 bulbs.
The bulbs used for this test are a ZooMed 10000K, a ZooMed actinic (both T8's) and a Philips 03 (40w standard actinic). I wish now I had a VHO to test as I expect it would produce the results we want.

This thread is longish, but if you want to know what can be done for a much lower cost then conventionals, read on!!

Ballast cost, $2.25US each. Yes, thats right $2.25us, I bought them in bulk from eBay where they are very often found. Currently there are the same ballasts in 240v for about the same price.
Bulb cost, the ZooMed's are on par with other reef bulbs and come in all the reef colour ranges. A T8 by nature is more efficient then T12 (NO's) giving 95% of the par at %80 of the wattage so opperation is cheaper and at 1" dia, more can be fit in a hood.

Power draw results, these are interesting and suggest a VHO bulb would run at VHO output in this configuration. I base that on the fact that a VHO has slightly more mercury in it and will draw more amperage (if available) then NO bulbs. In this test, more amperage is available then the NO's are drawing.

Draw no bulbs 22.92watts
Draw 1 T8, wired for 1 = 50.4 (27.48 net per bulb)
Draw 2 T8's, wired for 2 = 70.8 (24.3 net per bulb)
Draw 1 T8, wired 2X = 67.2 (44.28 net)
Draw 1 T8 wired 3X = 75.6 (56.8 net)
Draw 1 T8 wired 4x = 83.4 (60.48 net)
Draw 1 T12 40W wired 4X = 72w ((49.1 net)

Also of note, the T8 actinic drew less then the T8 10K did (7 watts less in 4X mode).

Unfortunatly I do not have a light meter so I used a camera, aimed at the bulb(s) from 6' with an otherwise black room. The numbers on the cameras meter are useless except for comparison in this test (1-10 range).

The T8 output peaked at approx 2.5 times it's base when wired 4X which goes inline with the draw. The Philips however didn't fair as well by not quite doubling output (again, crude light meter and the actinic colour may be playing games too).

Now for the research side of it, to fire a tube, you basically either heat and vapourize mercury (normal ballast, conventional start) or you excite it with RF range voltage (instant on electronic), you then supply voltage to it to cause it to interact with the phosphores and produce light. The voltage required is a constant for a given tube length. The amperage is varied, to some degree, but the amount of mercury available (or load it can consume). This test clearly shows, like StirCrazy's, that you can overdrive NO bulbs easily, but we know that because even Icecap tells us that. They also point out that an NO bulb will run brighter on their ballast, but not as bright as a VHO.

These Sylvania ballasts have a max capacity of 120w total of tube (they state either 3 40W NO's or 4 32W T8's or anything less). I fully expect that wired together in 4x mode (4x32W), they will drive to full output, a VHO 110W 4 footer (and consume a wattage inline with the higher output). In a nut shell, it is my belief they can be used to run VHO's properly and without undo stress on either the bulb or the ballast.

2 addvantages become clear to me, 1 is cost, even at the normal CDN price of $45 each, 4 of them will drive 4 VHO bulbs for $180, at the eBay or other discount prices, the sky is the limit. In my case I can use 4 of them for 4 VHO bulbs for $10us :) The other advantage is one that kept me from using something like Icecap in the first place. I don't want all my tank lights to come on at once. I want an actinic on first, followed by a 6500K (ZooMed has) followed by 2 10,000K's and reversed at sun set. You cannot do this with the Icecaps, you can with 4 seperate ballasts..

Feel free to ask any questions, don't be foolish enough to try this with normal tar ballasts, they WILL burn! Try other electronics at your own risk, but know these Sylvania's do work. Also note, the wattages expressed are RMS using an RMS amp meter on the ballast feed. They are total consumed watts, not bulb consumed. For peak (what you are shown in adds) go roughly 20% more. Ie a T8 wired 1X consumes about 32.9w, wired 4x is about 75W. Following the listed lumin output for a T8 vs a T12 NO bulb, this would equal around 90W of NO required to match 1 T8 wired 4X

Imagine, VHO lighting for even $50 per 4 bulb ballast with the ability to turn on or off 1 bulb at a time. It's going to work for me. For now I will run the T8 bulbs, maybe someone can test this system with a VHO bulb? (I will eventually...)

Greg Moore

Greg Moore
January 8th, 2002, 12:14 PM
Forgot to add the heat. After running 45 mins, the T8 bulb was running 30% warmer then under normal wiring. The ballast was the same as normal and slightly cooler then if it were running 4 individual bulbs. This was with no fans running. That is as you would expect it to be. What also appeared (slightly) was ripple effect on the tanks floor. No where near what I get with my halides, however still visable. Since the T8's are only 1" diameter and were considerable brighter, I guess that is why. I ripples best in the same direction as the bulb runs of course though..

Greg Moore

StirCrazy
January 8th, 2002, 07:56 PM
hey that's cool Greg, but do you have the #'s for the efficiency and the power factor for those ballast? might be good to have a few tried out ballast to chose from.

Steve

Greg Moore
January 9th, 2002, 08:06 AM
Full product name is QT4X32120IS 120V 10/CS 1/SKU

It is Osram Sylvania.. Specs as follows.

Total Harmonic Distortion <20%
Power Factor > 0.97
Ballast Factor 0.90
Circuit Type, parallel
Sound Rating, A
Min Start -18C
Nominal Input Wattage, 114
Remote mounting max, 18F

Usable lamp list is massive but covers from 20" 14W T8 through 60" 40W T8. For our uses it covers 24", 36" and 48" in normal wattages for those bulbs and T8 wattages for those bulbs. Max wattage is 125, ie 4 32W T8 or 3 40W T12. Bulbs compatable with instant start are recomended. Research suggests that rapid start bulbs are ok too for our uses (on once a day for extended periods) but would not be a good choice for a room light that will be on/off several times a day for short periods.

My next target is our little 30G hex. It uses an 18" 15W power glo and that is all that can fit. Since this ballast does down to 4 17W bulbs in recommended use, I wonder if a 15W would work on an X2 or X3 :)

Greg Moore

Greg Moore
January 14th, 2002, 01:33 PM
As an update. I have since switched out 4 NO tubes in favour of 2 ZooMed tubes that are overdriven on my 90g. I did have 10K halides on this tank at one time also but that proved too much for most of the corals I keep (no clams, no SPS etc).

The 2 ZooMed's I am using are the 10k and the actinic and to be hounest, this is some of the nicest colour I have seen yet with all the different lights I have tried. Along with the 2 overdriven bulbs are 2 NO tubes (powerglow and 10k)

I did encounter a frustrating and odd problem with my DIY temp contoller. With the overdriven bulbs on, it read (and tried to control) 8 degree's low. After exhaustive playing and tinkering, I found out that A) it isn't due to overdrive but it is the ballast (did the same in NO) and B) the T8 ballast seems especially sensitive to grounding of the reflector. For some time the ground had been off of it and I hadn't worried about it. Re-attaching the ground to the reflector solved the digital readouts errors.

For those looking at doing this, here is the ZooMed bulb page, they don't sell directly, you need to find a supplier, but you can see the choices (better then either VHO or PC choices)..

http://www.zoomed.com/html/lites.html

Greg Moore

Greg Moore
January 21st, 2002, 07:47 AM
Attached is a picture taken with a Coolpix 950. It shows 2 Zoomed bulbs (03 and 10K) overdriven 4-1 and 2 PowerGlo tubes running normal output, in actual live views it is just as dificult to see if the PowerGlo's are on (they are). It is noticable in the tank if they are on or off, however this gives you a good idea of the difference in overdrive..

Greg Moore

artbatista
February 2nd, 2002, 09:06 PM
I just had similar results with a standard (electronic, instant start) Advance Transformer Co. ballast. This is a very common brand and is available at Home Depot. The part number for the ballast I used is REL-4P32-SC.

This ballast 120Vac input are the black and white wires, the two yellow wires go to one side of the lamps and the 2 red and two blue wires can be added together for higher output. This ballast can drive up to 4 x T8 and although it does not say so it can also drive 4 x T12.

I tried several combos as follows ( all were 48 inches, daylight bulbs):

1xT8 NO
1xT8 2x
1xT8 3x
1xT8 4x
1xT12 NO
1xT12 2x
1xT12 3x
1xT12 4x
1xT12 HO
2xT12 HO
1xT12 VH0

Since I don't have a light meter, I ran a normal HO and a VHO lamp side by side with the normal overdriven bulbs. The NO driven to 3x and 4x were slightly brigther that the HO bulb, but not quite as bright as the VHO. Also the HO and VHO ran essentially cool to the touch while the 2x, 3x and 4x NO bulbs were somewhat warm but not hot.

While driving the HO and VHO bulbs with the electronic ballast, the brightness seemed the same as the same bulb with the conventional ballast.

I am sorry I can not post amp readings as I left my Amp meter at home, but I will try to redo this with meter on hand in the near future.


Art

StirCrazy
February 25th, 2002, 12:22 AM
a while ago I had a therory on running two ballastes in parallel, DON"T DO IT unless you like buying ballasts :D
it would seam that the balasts are protected from feedback power throught there out put leads but not through there yellow wires. so when I flashed one up I fried the other.. then when I was playing with the other tring to figure out the problem I fried the first one.

so my recomendation is to stick to one ballast when overdriving bulbs hehe.

Oh well some one had to try it.. right :)

Steve

Dman
February 26th, 2002, 12:46 AM
Steve,
Thank you for trying first, as I was pondering the same thing. And Lord knows I've neither the time or the money, so thanks again
Dman
:spin: