Originally posted by JerseyReef:

I noticed in the 400W Iwasaki test, the test results for the same ballast (Magnetek#1110-247SC-TC) are different. I know the tests were done two years apart. Has anyone received a reason for the difference?

In this case the ballasts are clearly identified by part number in both articles. It appears to be a case where the individual ballast that was tested for the Marine Fish & Reef article put out only 85% of the PPFD of the individual ballast that was used two years earlier. Again, only a single sample ballast was used in each test.

Another question, what was the spectral shift of the 400W Iwasaki bulbs in refference to MV Vs. MH? The PPFD numbers are fairly close and efficiency being nearly the same as well. Is the bulb more yellow on MH than on MV.

If you click on the first link in the original post, you can examine the various graphs in the article. There appears to be no perceptible difference between the various ballasts tested in regards to "shifting." But check out the graphs for yourself and see what you think. I have seen a few posts by hobbyists who claim that, in their opinion, Iwasaki lamps run on MV ballasts are less yellow than when run on MH ballasts. And, of course, Iwasaki Corp. says that these lamps should be run on MV ballasts.

Ninong [img]/ubb/smile.gif[/img]

------------------
Irrational Exuberance!

Thanks Ninong,

The part II of the Aquarium Frontiers report (on the 250W bulbs) has the following table:

TABLE III
PPFD and CCT for the new 250-watt lamps

Lamp Type PPFD CCT

6500 K Iwasaki 104.50 7457
5000 K Coralife 58.34 4585
10,000 K Coralife 51.57 not applicable (na)
20,000 K Coralife 37.24 na
10,000 K HQI 128.80 11,723
6000 K Tungsram 56.00 8152
6500 K Coralife 53.38 5339

These all use the M58 ballast as stated in the article. There is no mention of the 124 PAR result in this article, so that is really what I am wondering. Maybe a different test with a different ballast produced the higher Iwasaki result? It must have come from the Marine Annual 2001 report? I wonder what ballast they used. I don't have a copy of that report so I can't see it for myself. I think Magnetek makes other newer ballasts. Maybe they are improved. And maybe aren't out in the "population" yet but reported on in that 2001 article? Just guessing here now.


[This message has been edited by MtnDewMan (edited 04-06-2001).]

Just wanted to add, since the subject of cost of a new ballast vs. a diy ballast was brought up, the cost of my MH setup came in just under $60 for the ballast, wiring, and switch assembly. It is a safe setup now enclosed in an electrical box,which is vented (though it doesn't need to be since it is located away from the tank in a remote location). It runs 2-250 watt Iwasaki bulbs. The cheapest I was able to find a PFO dual ballast was at Marinedepot for $143. All wiring and switches used were top quality and approved by an electrician I know (who I was able to ask these questions about the specifics needed). I would not recommend a DIY setup for someone who doesn't have the ability to set it up correctly or at least have someone set it up. Going PREmade is definately the way to go in that case. But for anyone with basic wiring skills, it's easy! When I have it all set up...(like things actually IN the tank [img]/ubb/smile.gif[/img] ); I'll post pics! Then you can see what you think. HTH someone out.

[img]/ubb/biggrin.gif[/img]
Joanne

------------------
That which does not kill us makes us stronger...like lifting those salt buckets.

Originally posted by MtnDewMan:

The part II of the Aquarium Frontiers report (on the 250W bulbs) has the following table:

TABLE III
PPFD and CCT for the new 250-watt lamps

Lamp Type PPFD CCT

6500 K Iwasaki 104.50 7457
5000 K Coralife 58.34 4585
10,000 K Coralife 51.57 not applicable (na)
20,000 K Coralife 37.24 na
10,000 K HQI 128.80 11,723
6000 K Tungsram 56.00 8152
6500 K Coralife 53.38 5339

These all use the M58 ballast as stated in the article.

All were tested on the same individual Magnetek M58 ballast except the 10,000K HQI double-ended lamp, which was tested on an HQI ballast.

There is no mention of the 124 PAR result in this article, so that is really what I am wondering.

The 250w Iwasaki tested at 104 PPFD in this test. The measurement of 124 was made a year later.

Maybe a different test with a different ballast produced the higher Iwasaki result?

It was a different test but we are not sure if it was a different model ballast or not. I e-mailed Sanjay last night asking that question.

It must have come from the Marine Annual 2001 report? I wonder what ballast they used.

I'll let you know what Sanjay says.

I don't have a copy of that report so I can't see it for myself.

I do. [img]/ubb/wink.gif[/img]

I think Magnetek makes other newer ballasts. Maybe they are improved. And maybe aren't out in the "population" yet but reported on in that 2001 article? Just guessing here now.

Who knows. I'll post what Sanjay has to say about this. One interesting point about the recent ballast test is that the Ice Cap electronic ballast that they originally tested on the 250w Iwasaki lamp ran it at 25% more PPFD than the Magnetek ballast, but was only able to light one out of five lamps tested without flickering. They contacted Ice Cap for additional ballasts and it appears that the ballasts Ice Cap sent them that were capable of running all of the 250w Iwasaki lamps, without flickering, were now underpowering them, instead of overpowering them as the first sample ballast did.

You will notice that the Magnetek ballast that was used to test the 400w Iwasaki lamp in both studies was the same identical model number and yet the result in the most recent test was only 85% of the earlier test. Could be that what we are dealing with here is just normal deviation in ballast performance between various individual samples of the same model. Might even be variation in the lamps themselves.

Will be interesting to see what Sanjay has to say about it.

Later,

Ninong [img]/ubb/smile.gif[/img]




------------------
Irrational Exuberance!

I am beginning to think you are right about variation in results. I am inclined to use the lower numbers or the average of the numbers than the higher "single" number until more data is released. Even the HQI PAR number is suspect in my mind because there is no other data to back it up. Hard to make definitive conclusions based on one data point. The only conclusion I can make on the HQI right now is that it is in the same ballpark as the Iwasaki, which I don't think is too far fetched of a conclusion.

Thanks for asking the questions to Sanjay. Looking forward to hearing what the response is.

[This message has been edited by MtnDewMan (edited 04-07-2001).]

What/where is the test for icecap? The on sanjay did reported 99.8par on iwaski(IC). But test can vary. IMO if you have at least 250w halides or any halides in general it makes for better corals growth. The k is preferance. IMO water quality still is more important.

[This message has been edited by Ironreef (edited 04-07-2001).]

Originally posted by Ironreef:

What/where is the test for icecap?

From Marine Fish & Reef USA 2001 Annual, article by Sanjay Joshi and Dave Morgan, page 92:

"The only 250-watt lamp tested was the 6500K Iwasaki lamp. The ballasts used were the Magnetek magnetic ballast, the Magnetek mercury vapor ballast, the Blueline 250 and the IceCap 250.

None of the ballasts performed as well as the magnetic ballast using this lamp (see Table VI). More specifically, the electronic ballasts (IceCap and Blueline) seemed to significantly 'underdrive' the lamp. The IceCap ballasts previously available 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 five lamps we tested without considerable flicker. This ballast, however, did provide lamp output with ABOUT 25 PERCENT MORE PPFD than the magnetic ballast."

What Sanjay is telling us is that the first IceCap ballast they tested (the one that was able to light only one of five lamps tested without considerable flicker) produced 25% more PPFD lighting that one lamp than the Magnetic ballast. The Magnetic ballast tested at 124.3 PPFD in this test, which means the first IceCap ballast tested at 155.3 PPFD on the one lamp it was able to light without considerable flicker. They requested, and received, additional ballasts from IceCap. The results of the first IceCap ballast are not shown in the chart, although he clearly tells you that it produced 25% more PPFD than the Magnetic ballast. The figure of 99.8 PPFD was obtained using a ballast that IceCap sent them that was successful in lighting all five of the lamps without flicker. I believe they went through three or four ballasts before they were satisfied that they had a ballast that would light all five of the sample lamps without flicker. It is obvious that the first IceCap ballast was designed to "over drive" the lamp and the last ballast IceCap sent them was designed to "under drive" the lamp.

Ninong [img]/ubb/smile.gif[/img]


------------------
Irrational Exuberance!

Ninong,

Any reply yet to the Magnetek ballast that provided the PAR of 124.3?

What are these Iwasaki's rated at? If it were able to run at a PAR of 155 as mentioned above? Would the life of the bulb not be compromised and reduced because of it?

Originally posted by MtnDewMan:

Any reply yet to the Magnetek ballast that provided the PAR of 124.3?

As a matter of fact, yes. Sanjay just e-mailed me and I will post his reply in a minute.

What are these Iwasaki's rated at? If it were able to run at a PAR of 155 as mentioned above? Would the life of the bulb not be compromised and reduced because of it?

Your question is, "Would the life of the lamp be affected by 'over driving' it with the ballast?" I don't know the answer to that question but you might inquire of IceCap. I know that they do over drive fluorescent lamps with their 430 and 660 ballasts, so they should be able to answer your question at least as far as fluorescents are concerned. I'm not actually sure if they are attempting to over drive metal halide lamps or not, but you could always ask them.

Ninong [img]/ubb/smile.gif[/img]

------------------
Irrational Exuberance!

E-mail from Ninong to Sanjay Joshi:

Dear Dr. Joshi,

One more question, if you don't mind. In the Dec. 1999 Aquarium Frontiers article the 250w 6500K Iwasaki was tested as 104 PPFD on a Magnetek M58 ballast.

In your Marine Fish & Reef USA 2001 Annual ballast comparison article, you show the same lamp (250w 6500K Iwasaki) tested as 124 PPFD on a Magnetek ballast.

Were these the same Magnetek ballast? Or were they different Magnetek ballasts? And, if different, how do we identify the one that tested with the higher PPFD?

If they actually were the same, do you have any idea why they might have tested differently?

Thanks again for your help,

Ninong

Reply from Sanjay:

Ninong,

There are several factors that an explain for the discrepancy.

1) The lamps used were different from what I used in 1999 and last year
2) The ballasts although Magnetek were not the same ones either
3) When I did the ballast test I made sure that i was using a voltage regulator and all
ballasts were exposed exactly to 120 Volts. When I first tested the lamps, I did not do
that. There is some variation whent he voltage fluctuates too.
4) when I tested in 1999 there were some lamps that did test higher than 106.

My rule of thumb is that there is enough variation in product that small numbers (<20% )
may be a good tolerance band.

Hope this tries to explain the variation.

sanjay.

Thanks for the response.

Basically what his response tells me is ... Only use these charts as a rule of thumb. Do not use them as absolutes. So if the chart says 124, that is not an absolute and may not be a fact for that particular lamp set. Could be as low as 20% lower from lamp to lamp, ballast to ballast, power supply to power supply.

When someone argues that a PAR is significantly higher based on the charts, it may or may not be fact unless more tests with more controlled equipment is used with a sample rate much higher than 1. You can however use the charts to get a "ballpark" idea of a PAR level.

Does that pretty much sum it up? That there are no absolutes in the charts?

OK, maybe this will help. When he says that the lamps used were different, I believe he means that even though they were 250w 6500K Iwasaki lamps, they were not the exact same samples that he used in the earlier test.

When he says that the ballasts, though Magnetek were not the same ones either, I believe he means that he did not reuse the exact same ballast that he used in the earlier test. In the 1999 test he used a single Magnetek M58 ballast, and I believe he is telling us that that exact same ballast was not used in the most recent test. I do believe we are talking about the same model, however.

And he says that he used a voltage regulator for the recent ballast test and he did not use one in the 1999 test.

He is also telling us that even though he used a figure of 104.50 PPFD for the 1999 test, there were some lamps that tested at 106 point something. He usually uses several sample lamps and I would assume he chooses the figure that he concludes is the most accurate, or the most representative.

And finally, he considers that there is enough variation in product that variations of less than 20% should be expected.

I suppose if you acquired 10 samples of each ballast to be tested and at least 10 samples of each lamp to be tested, and then averaged your figures after discarding the high and the low figure, you might have a more reliable number. But it still wouldn't mean much if you ended up buying a ballast that matched the worst one tested, or a lamp that matched the weakest one tested.

Ninong [img]/ubb/smile.gif[/img]

------------------
Irrational Exuberance!

Yep, that's what I was thinking also ... variations between lamp to lamp and ballast to ballast can range from low to high PAR ... somewhere in between the low and the high is the real "nominal" number (assuming a different test wouldn't pop out a new high or a new low PAR level).

I don't want to take up more time on the subject, but just shows that variability in the hardware and voltage can make the PAR results somewhat misleading without a larger sample size of ballasts and lamps to make the data statistically relevant. For example ... 20 different ballasts of the same model with 20 different lamps of the same model and run tests on all the combinations of them (like a design of experiments study). Then the resultant data would be more statistally relevant. But the data is still useful to get a good idea of what the output can be.

People tend to look at the data as hard and fast data .... and it is clear that it is not hard and fast. What you buy can be far off from what the test results indicate (per Sanjay, up to 20% maybe). I also got caught up in the numbers as well until I started to think harder about it, so now I am not caught up in those numbers as much because I realize they can be much higher than what I possibly may have with the combination ballast/lamp I bought. Others should keep that in mind as well if they think they absolutely are achieving the PAR levels that are described in the reports.

Thanks for your time on this Ninong.

[This message has been edited by MtnDewMan (edited 04-08-2001).]

hard data does mean a difff. But if you keep a range . Like absolote may not be needed but when 250w iwaki and 250hqi put out more par than 400w 10k then I belive data come to effect. Anyway I have a nice softball size purple tip that came from 250w 10k. not hqi and its green or turning green and purple tip and hadve a brwn millipora thats turning green. This is what I go by. When anyone put corals in there tank and have good result this is what they go by. I would like to see how MTD frag do after the change. I belive good alk/cal is most important. even docmac says he had some sps do better under vho. Just depends. i prefer Iwaski cuz they will last 2 yrs compared to hqi 1 yr. even when you addd the varible thats alot.If my girls didn't take private violin and piano I probally would have 1000w griesiamen(sp) I'm what you call buy all the good stuff but try to be cheap. How do you do that with a reef? Buy it one = buy all good stuff

Ninong,


Is it possible that the double ended 10k new Coralife lamps from Germany are the BLV Hitlites under the Coralife brand name ?

Just a thought...........unless you know otherwise ?

------------------
Cheers

Marc
All credit for the webpage goes to Aquanut.... Marc's reef


One of the few worthwhile UK Reef BB's....HERE
Now with added CHAT

Marc ~

I cannot say for sure, but it is certainly possible, maybe even probable. Ushio's BLV division is known for manufacturing identical lamps that are sold by other distributors under their brand names, in addition to the obvious brands of Ushio and BLV.

Ninong [img]/ubb/smile.gif[/img]

------------------
Irrational Exuberance!

Whew sounds like 250 watt HQI are the way to go [img]/ubb/smile.gif[/img] now to start saving up for them [img]/ubb/smile.gif[/img]

Schottman [img]/ubb/smile.gif[/img]

Ironreef,
I would like to know where you get your data that the HQI bulbs only last 1 year?

Ninong,

I've tried all sorts of digging about looking for a link between the two manufactures with no luck!
It just seemed to me the way the new caralife's where described it could almost be too much of a coincidence[img]/ubb/biggrin.gif[/img]

Correct me if i'm wrong but you use(plan to0 use the Hitlites 250w 10k lamps dont you ?
Have you set anything up yet or are you still planning ?
Curious to know how you find the lamp when you do set up......IMO these lamps give the best colour out of all the lamps i've seen and tried. I was wondering what criteria made you chose the lamps yourself ?

------------------
Cheers

Marc
All credit for the webpage goes to Aquanut.... Marc's reef


One of the few worthwhile UK Reef BB's....HERE
Now with added CHAT

Marc ~

I have all of my equipment except for the sump. My primary light fixture came with 250w AB 10000 HQI double-ended lamps, which are actually 13,000K; and I purchased a couple of 250w BLV 10,000K HQI double-ended lamps, which tested at 11,723K according to Sanjay Joshi.

I chose the 250w HQI double-ended lamps because they are equal to, or slightly better than, the 250w 6500K Iwasaki lamps and, in addition, are not yellowish looking. They are also superior to most 400w metal halide lamps (with the exception of the 400w Iwasakis). This means I can get by with less actinic supplementation and with less electricity consumption. I was influenced by Sanjay's test results and by viewing pics of so many German and Japanese tanks lit with HQI double-ended lamps.

Ninong [img]/ubb/smile.gif[/img]

------------------
Irrational Exuberance!