Spectral Analysis of MH Lamps - Do ballasts make a difference

Sanjay Joshi & Dave Morgan

Reprint of the Article from 2001 Annual Marine Fish and Reef USA, Fancy Publications.

Introduction

Lighting of reef aquariums is still an area of considerable debate among reef hobbyists. One missing component in this debate was the knowledge of verifiable light output by the lamps, however, in recent years, there has been some research conducted on the lamps used for reef lighting to provide quantitative data in order to make assessments and comparisons(Dana Riddle [1], Richard Harker [2,3], Sanjay Joshi & Dave Morgan [4,5,6]). Based on this research we now have a much better understanding of the spectral quality of light output by various lamps used in the hobby.

More recently another variable has been introduced - the choice of ballasts to drive the lamps. Electronic ballasts have been introduced e.g. by Icecap and Blueline, and special ballasts have been targeted towards specific lamps:. PFO-HQI ballast for the 10000K Ushio lamps and 20000K Radium/Osram lamps, mercury vapor ballasts for the Iwasaki lamps. The aquarist is now faced with yet another decision in the complex world of aquarium lighting - "Which ballast should I use?" The intent of this article is to evaluate the various ballasts and lamp combinations and hopefully provide some data to help the aquarist make an educated decision.

Criteria for Comparison

To evaluate the ballasts several different criteria were selected:

1. Input Power - this is of importance to the aquarists since it ultimately impacts the electric bill
2. Light output - the spectral distribution of the lamp output and the PPFD (Photosynthetic Photon Flux Density) is used to determine the quality and quantity of light output by the lamp
3. Efficiency - which is defined here as the ratio of the PPFD to the input power.
4. Set of allowable lamps - not all ballasts will work with all lamps. One criterion that may be of concern in choosing the ballasts is the allowable set of lamps that can be used with the ballasts. This is of importance, since it provides the flexibility of changing lamps in the future without having to change the ballast. Table 1 shows the list of available ballasts and the set of allowable lamps for each ballast.

Table 1: Ballast and Lamp Combinations

 

Ballast		LAMPS
		400W				250W	175W
		Iwaski 6500K	Ushio 10000K	Radium/Osram 20000K	GE 6000K	Iwasaki 6500K	Venture 5500K	Ushio 10000K
400W	Magnetic	X	X	X	X
	MV	X
	PFO-HQI		X	X
	Blueline	X	X	X	X
250W	Magnetic					X
	IceCap					X
	Blueline					X
	MV					X
175	Magnetic						X	X
	Blueline						X	X

 

5. Economic Considerations - The cost of the ballast will impact the initial cost of the lighting system. This should be weighed against the potential running costs to determine the economic impact.

Description of Test Setup

Figure 1 shows a schematic of the circuit and equipment used to test the ballasts. In setting up the test apparatus several parameters were controlled and measured. The first necessary control is voltage. Controlling the line voltage insures that all ballasts tested are operating at the same input level, and any variations due to input voltage fluctuations are eliminated. This was accomplished using a Powerstat variable autotransformer set at 120 VAC. The next function of the test apparatus was to measure current and power. Both of these values were measured using a Victor true RMS Watt meter. It measures voltage, current, and power. These parameter were used in comparing the performance of the ballasts. Next the lamp operating voltage was measured to give some indication of operating differences between the tested ballasts. The final parameter measured was spectral output of the lamp. Measuring the spectral output provides all the necessary information about the quality and quantity of light output. The spectral measurements were taken using the same setup that was used in our earlier articles (References 4,5,6).

The instrument used for measurement of spectral distribution is called a spectroradiometer. The authors used an instrument made by LiCOR, the LI-1800/12 portable spectroradiometer (www.licor.com) with a standard cosine receptor, capable of measurement from 300-850 nm at 2 nm intervals. The specific instrument used was calibrated for use between 310-850 nm.

The test lamps were mounted in a fixture, such that the center of the arc tube was 18" from the cosine receptor. The lamp was mounted without a reflector, in a dark room with walls painted dull black to minimize the impact of reflection, and eliminate any stray light. The same setup was used with all the test lamps.

The lamps were allowed to warm up for about 20 minutes before taking the readings. A few readings were taken during this period to make sure that the light output from the lamps had stabilized. The spectroradiometer was programmed to take the average of 3 consecutive readings, and automatically record the data. The IBM PC Interface program (LICOR proprietary software) was used to analyze the data collected and compute the photosynthetic photon flux density (PPFD) for each lamp.

Based on the data collected and measured, several parameters were defined to further help evaluate the ballasts. The magnetic ballast was used as the baseline, and the following was computed:

1. Percent change in the PPFD - to compare variation in PPFD
2. Efficiency - defined here as the ratio of PPFD to input power. This determine how effective the lamp is at converting the input power into PPFD
3. Percent change in input power requirements - to compare the changes in input power
4. Percent change in efficiency

Figure 1: Test Setup

400W Lamps run with different ballasts

400W 6500K Iwasaki Lamps

The ballast recommended by the manufacturer is the ANSI H33 ballast, since this lamp is designed to replace the mercury vapor (MV) lamps in MV light fixtures. However, most aquarists run this lamp on ballasts rated for both M59/H33 (e.g. Advance and Magnetek ballasts). The magnetic ballast used here for the tests is the Magnetek ballast part # 1110-247SC-TC, the Blueline 400W, the Magnetek MV ballast, and the PFO-HQI ballast. The spectral distribution of the light output with the various ballasts is shown in Figure 2. As seen from the figure the variation in light output between the ballasts is fairly small, with most of the variation occurring in the 450-650 nm range.

Table 2 shows the comparison of the input current (amps), and input power, along with the lamp voltage and the PPFD of the light generated, as well as the percentage change in PPFD and power when compared to the magnetic MH ballast.

Table 2: Comparison of Ballasts for 400W Iwaski Lamps

It appears from this data that there is very little difference between the light output as well as power consumption of the ballasts. The Blueline ballasts is slightly more efficient in terms of PPFD/watt of input power, but its light output is also lower. The Blueline ballast does use less power and correspondingly has less light output. For the 6500K Iwaski lamps it appears that there is no significant difference among the ballasts.

Figure 2: 400W 6500K Iwaski Lamps with different ballasts

10,000K USHIO

As seen below in Table 3, in the case of the 10,000K Ushio lamp there seems to be some advantage to using the PFO-HQI and Blueline ballasts over the conventional magnetic ballast. Interestingly, although there is not much difference between the PFO and Blueline ballast based on Table 3, spectrally the PFO ballast shows higher output in the lower end of the spectrum as well as higher peaks in the 525-575 nm region. The Blueline ballast tends to have a higher increase in the 600-700 nm region.

Table 3: Comparison of Ballasts for 400W 10,000K Ushio Lamps

Figure 3: 400W 10,000K Ushio Lamps with different ballasts

20,000K Osram/Radium Lamps

For this lamp, as shown in Table 4 there was more than a 20% difference in light output using the PFO and Blueline ballast. Spectrally the output of the PFO and Blueline ballasts are almost identical (see Figure 4) the PFO ballast however has a higher power consumption and current draw.

Table 4: Comparison of Ballasts for 400W 20,000K Osram/Radium Lamps

 Figure 4: 400W 20000K Osram/Radium Lamps with different ballasts

6000K GE Lamp

Again as seen from Table 5 and Figure 5 , there is a significant difference in light output when using the Blueline ballast. For the same power consumption, there is 28.6% increase in light output. Based on this data we checked the specs for the GE 6000K lamp and found that the recommended ballast is the ANSI M-135, not the ANSI M59 as was used in the test. Unfortunately, we did not perform tests with this recommended ballast, and it may explain the poor performance of this lamp with the M59 ballast.

Table 5: Comparison of Ballasts for 400W 6000K GE Lamps

Figure 5: 400W 6000K GE lamp with different ballasts

250W Lamps

The only 250W lamp tested was the 6500K Iwasaki lamp. The ballasts used were the Magnetek magnetic ballast, Magnetek Mercury Vapor ballast, Blueline 250, and IceCap 250.

As can be seen from Table 6 and Figure 6, none of the ballasts performed as well as the magnetic ballast. More specifically the electronic ballasts (IceCap and Blueline) seem to significantly under-drive the lamp. The earlier IceCap ballasts for this lamp were reported by several users to be unstable, and this was our experience as well. The first ballast we tested was successful in lighting only one of the 5 lamps tried without considerable flicker. This ballast however did provide lamp output with was about 25% more PPFD than the magnetic ballast. The replacement IceCap ballast was more stable and did not have problems lighting the lamps, so it seems that the manufacturer subsequently addressed this problem. However, it would seem that this resulted in the lamp being under driven as seen from the data. The Blueline electronic ballast also performed poorly with this lamp.

Interestingly, the recommended MV ballast did not perform as well as the MH magnetic ballast.

Table 6: Comparison of Ballasts for 250W 6500K Iwasaki Lamps

 Figure 6: 250W 6500K Iwaski Lamps with different ballasts

175W Lamps

The 175W lamps tested were the 5500K Venture and the 10000K Ushio. The ballasts used were the Magnetek magnetic ballast, and Blueline 175. As seen in Table 7 and Figure 7, the Blueline electronic ballast performs much better than the standard magnetic ballast when driving the Ventura lamp, although the efficiency of these ballasts with this lamp is pretty low. For the 10000K Ushio lamps the situation is reversed - where the magnetic ballast performed much better (see Table 8 and Figure 8).

Table 7: Comparison of Ballasts for 175W 5500K Venture Lamps

 Figure 7: 175W 5500K Ushio Lamps with different ballasts

Table 8: Comparison of Ballasts for 175W 10000K Ushio Lamps

 Figure 8: 175W 10,000K Ushio Lamps with different ballasts

Economic Considerations:

The magnetic ballasts are the cheapest option, however the electronic ballasts claim to be more efficient and hence cheaper to run. This claim can be evaluated to some extent by the data presented in this paper.

Another consideration that plays an important role in the selection of the ballast is the economics - weighing the initial cost of the ballasts against the operating costs.

To determine this we can compute the Pay Back Period, using the following formulas:

Running cost per day = (Power consumed in KWH) * cost of electricity per KWH

Power consumed in KWH = (Power consumed by ballast * hours of use per day)/1000

Do this for each of the ballasts being compared. Then compute the,

Pay back period = Difference in initial cost of ballasts / (Savings/Day)

As an example, let us consider the situation where I would like to evaluate the 400W magnetic ballast vs the Blueline 400W electronic ballast. The numbers used for the example pertain to my operating conditions, yours may vary so please choose the numbers that represent your particular situation. All pricing was obtained off the web, (after searching for the best available price).

Cost of PFO magnetic ballast = $105

Cost of Blueline 400W ballast = $159

Difference in Initial Cost of Ballast = $54

Power Consumption of Magnetic Ballast (from Table 2) = 463W

Power Consumption of Magnetic Ballast (from Table 2) = 419W

Cost of electricty/ KWH = $0.06

Hours of use per day = 10 hrs

Running Cost/day of Magnetic Ballast = 463*10/1000 *0.06 = 0.28

Running Cost /day of Electronic Ballast = 419*10/1000*0.06 = 0.25

Savings/day = 0.28 - 0.25 = 0.03

Pay Back Period = $54/$0.03 = 1800 days or 4.93 years

If your cost of electricity were double (say 0.12) then the pay back period would be 2.46 years, any further use beyond the payback period would result in a saving of approximately $22 per year, hence the Blueline electronic ballast may be an attractive option, depending on your electric rates.

However, when using 2 6500K Iwasaki lamps, you can get a dual PFO ballast for $159, but would have to use 2 Blueline Ballasts at a total cost of $318, leading to an intial difference in cost of $159. Using the previous electricity cost of $0.06/KWH would yield a pay back period of 7.26 years.

You should perform similar analysis based on your ballast/lamp combination and cost of electricity in your region to perform your own economic analysis when choosing a ballast.

Conclusions

Based on the data presented it seems that the ballasts do make a difference, but it depends on the lamp being used. Based on versatility and performance the magnetic (Magnetek, Advance, PFO) and Blueline ballasts are the best choices since they run all the lamps.(The Blueline ballast is not recommend for the new Sunburst 12000K lamps). The Blueline electronic ballasts are attractive since they performs quite well is most situations and are far more versatile ballast than the IceCap, since they do not require retuning for specific lamps. Based on economics the PFO dual magnetic ballast is an outstanding value.

Manufacturers of electronic ballasts have also claimed that lamps when with electronic ballasts will have a longer life because they are being driven at higher frequencies. These tests were performed using new lamps, and currently we do not have any data to support any claims on improved longevity of lamps when used with a particular ballast. Furthermore, at this point we have not ascertained the amount of variation between ballasts. All these tests were performed using a single ballast of each type.

Acknowledgements

We would like to acknowledge PFO lighting, Champion Lighting and IceCap for providing some of the ballasts used in this study. Finally, we would like to thank Dr. Paul Walker of Penn State University and LiCor Inc., for the use of their spectroradiometers and dark rooms for testing the lamps.

References

1. Riddle, D. The Captive Reef : A Concise Guide to Reef Aquaria in the Home. Energy Savers, Harbor City, CA. Pp. 297.
2. Harker, R. E. 1997a. Lighting for the reef tank. Aquarium Frontiers May/June:29-37.
3. Harker, R. E. 1997b. A comparison of metal halide bulbs. SeaScope 14(Fall):2.
4. Joshi, S. 1998. Spectral Analysis of Metal Halide Lamps Used in the Reef Aquarium Hobby Part 1: New 400-watt Lamps, http://www.animalnetwork.com/fish2/aqfm/1998/nov/features/1/default.asp
5. Joshi, S. and Morgan D. 1999. Spectral Analysis of Metal Halide Lamps Used in the Reef Aquarium Hobby Part II: Used 400-watt Lamps http://www.animalnetwork.com/fish2/aqfm/1999/jan/features/2/default.asp
6. Joshi, S. and Morgan, D. 1999. Spectral Analysis of Metal Halide Lamps Used in the Reef Aquarium Hobby Part III: New and used 250-watt Lamps http://www.animalnetwork.com/fish2/aqfm/1999/dec/features/2/default.asp