DE Metal Halide Lamp

My build is finally starting to come together and since my nursery will only be completed in the fall I’m going to use one of the flower rooms for a nursery/ veg room for a few months. I’m using luxx de fixtures in both flower rooms and want to install 1000w de metal halides in the nursery room, I was wondering if anyone had recommendations on good lamps?


I ran luxx CMH 315’s, I tried out a like no name DE CMH and it was a garbage experience. So I would definitely avoid any no name.

Can I ask why your not considering LED? You waste a lot of BTU’s with high output lights. BTU’s that could be absorbed by your plants as light energy, and you wouldn’t have to pay to eliminate/balance. We are talking 30% wasted energy. You are also setting your self up to have no spectrum control.

Because cost to install LED’s is much more on the wallet than running hps. De hps is merely a stepping stone to doing it with LED’s


I’d refrain from ushio mh bulbs. I recently sent back a case of them. For some reason our gavita de’s couldn’t fire them properly.


I don’t think you are in ca but they are about to fund a hps to led swap for licensed grows. To stop our rolling black outs. They say ……

1 Like

A few points before answering your question:

MH for veg and HPS for flowering is a flawed paradigm based on a misunderstanding of seasonal spectrum shifts under the sun. Dr. Bugbee (and many other researchers) have found that a minimum of about 5% blue range light (400-500 nm) is sufficient for all growth phases of plants, with a maximum of about 15%. As the blue fraction increases, the net photosynthetic and growth rates decrease.

If you’re using MH to reduce internodal elongation and plant height, there are other more appropriate and effective methods. Including negative DIF (-5’F to -10’F) for a week or so, about a week after plants transition to the veg phase. Or better yet, use a DROP (“cool morning drop” aka “cool morning pulse”) instead of DIF. Plants begin responding to negative DIF or DROP within a day or two. And run P at 20 mg/L.

I recommend using DE HPS for veg because most DE HPS lamps provide sufficient blue light, a good amount of green light, and a nearly perfect amount of red light. The blue and red light fractions are why DE HPS’s photosynthetic photon efficacy (μmol/J) is almost 1.5x that of DE MH. High-quality DE HPS lamps have good spectral characteristics for all growth phases. Much better than DE MH but not as good as LED (because DE HPS has a substantial IR spike at about 825 nm). The IR spike from DE HPS is why they increase leaf temperature compared to MH and LED.

Here are the spectral characteristics of a Gavita DE HPS 1000W lamp in a Gativa Pro DE fixture measured directly below the lamp (0° angle). These values are typical of high-quality DE HPS lamps:

UV-A (320-399 nm) = 0.19%
Blue (400-499 nm) = 5.12%
Green (500-599 nm) = 41.74%
Red (600-700 nm) = 48.01%
Far-Red (701-750) = 4.94%
Blue:Red Ratio = 0.11
Red:Far-Red Ratio = 9.71


MH is a poor choice for lighting for many reasons:

  • DE MH has low photosynthetic photon efficacy (PPE) from 400-700nm as μmol per joule. The best DE MH is from Ushio, and its PPE (μmol/J) is only 1.8. That means the plant usable photons (1 micromole = 602 quadrillion photons) produced by DE MH per input watt is low when compared to HPS (about 2.4). And it isn’t even in the same ballpark as high PPE LEDs (currently around 2.7-3.0 μmol/J, with a theoretical max of approximately 4.6-5.1). So you’re throwing money into the toilet when using MH in terms of your electric bill.
  • MH has way too much blue light, much more than is needed or appropriate.
  • MH has way too green light, much more than is needed or appropriate.
  • MH has far too little red light, not nearly enough as is needed to drive a high photosynthetic rate.
  • DE MH lamps produce about a third (33.333%) of the PPF as DE HPS, so you’ll need to reduce the distance between plans and lights to provide the same PPFD.
  • MH DE doesn’t work well with any DE ballast. They often misfire; if you have 100 MH DE in a room, at least 2-5% of the blubs won’t fire each time you have a lights on event. Then, the next time you have a lights on event, the same approximate percent of lamps won’t fire, but they will be different lamps. There’s no rhyme or reason why various lamps won’t fire each time.
  • The first time DE MH lamps are powered in a DE HPS ballast, IME nearly all lamps fire. For that reason, you should only use DE MH lamps on a 24-hour photoperiod.
  • DE MH has a very short lifetime, only around 5,000-6,000 hours. Compared to around 10,000-12,000 hours for DE HPS and 30,000-50,000 hours for LED. Ensure to track the PPFD at the same distance from the lamp every 30 days and replace it once you see a 10% reduction compared to the PPFD measured after the first 100-hour burn-in period.
  • DE MH has a long warm-up period of 3 minutes (compared to DE HPS).
  • DE MH has a long re-strike period of 10 minutes. That means DE MH lamps shouldn’t be powered within 10-minutes after they are depowered. So, do not use DE MH lamps if your facility is prone to electrical brownouts where you’ll lose power for a couple of seconds or minutes because most DE HPS ballast controllers don’t have a re-start delay.
  • If you’re running a 24-hour photoperiod at a high PPFD, your DLI will probably be too high. So your PPFD shouldn’t exceed 600 (=51.84 DLI).
  • Regarding the flower DLI:veg DLI ratio, which should be 1:1 or greater (more DLI in flowering), if you’re running a 24-hour photoperiod in veg, you’ll need to keep veg PPFD between 400-500 (=34.56-43.2 DLI) if your flowering PPFD will be 1000 under 12 hours (=43.2 DLI).
  • DE MH brands are a safety concern because most are not “open rated.” Meaning they are not rated for open fixtures that lack a glass insert. The Luxx fixtures are open fixtures, so you shouldn’t use any lamps that aren’t “open rated.” Lamps that aren’t open-rated lack an inner glass shield around the arc tube. So, in the event of a lamp explosion, glass shards are thrown wide and at great speed, considering the arc tube operated at temperatures approaching 1000°C and pressures of >3 atmospheres. So, most DE MH lamp brands are a safety issue regarding glass shards in employees’ eyes and all over your plants and igniting fires.
  • If you’re a licensed and insured facility, you should talk to your insurer or attorney before using lamps that lack an open rating in the Luxx fixtures.
  • And other reason but I have to go…

To answer your question @Medicine.grower :

The only open-rated 1000W DE MH lamp I’m aware of is the Ushio Super MH. But if I have convinced you to avoid DE MH, I strongly recommend the Ushio DE HPS. I love Ushio because they are the best lamp company and produced in an ISO:9001 German mfg facility.


The CapEx/OpEx balance is typically covered in under 12 months, literally through cost savings on HVAC and Dehumid. It’s a no brainer to any investor if presented correctly, and should be a huge consideration if your funding yourself, you give yourself 6 months to climb out of a technological deficit in an extremely competitive market. Cost to produce… Cost to produce… You can cover your investment very quickly when your lbs cost you $120 or less.


I would add that when choosing LEDs, the photosynthetic photon efficacy (PPE) as μmol per joule threshold should be 2.5. In other words, if the PPE (μmol/J) is less than 2.5, don’t buy it because the ROI isn’t optimal. For example, the new Fluence SPYDR 2h we’re buying has a photosynthetic efficacy of about 2.7.

When comparing LEDs, the primary considerations should be photosynthetic efficacy, uniformity, price per photosynthetic photon flux (PPF), and performance and spectral characteristics (for PAR, according to Dr. Bugbee, 10% blue, 20% green, and 70% red). A bonus is whether the fixtures allow for remote drivers operating on DC power using digital electricy, which dramatically reduces the power distribution, installation, and electrical transformer costs.

Unfourtanetly, the PhysioSpec Indoor spectrum from Fluence isn’t optimal according to dr. Bugbee’s recommendation. But, Dr. Bugbee is workign with Fluence and their spectrum is better than other commeical LEDs I’ve seen:

Blue (400-499 nm) = 19%
Green (500-599 nm) = 41%
Red (600-700 nm) = 40%

For our phase two build, we’re using VoltServer for remote LED driver racks on DC power using digital electricity for energy distribution to the fixtures. With custom-designed LEDs by Libra Design. That allows for drastically reduced power distribution cost (when using alternating photoperiods between sets of rooms), and Libra Design can provide practically any performance and spectral characteristics, including PSS, PPE, and UV/blue/green/red/far-red profiles and ratios.

Libra Designs was co-founded by the former CTO and co-founder of Fluence. The other co-founder of Libra Designs is the former CMO of Fluence.

1 Like

I don’t think it’s a Gavita DE fixture and ballast issue; no DE MH I’m aware of works well with any DE HPS fixture/ballast. DE MH is a solution in search of a problem. Check out my post above with info on running DE MH in DE HPS fixtures.

From Gavita: (they’re wrong about the open fixutre rated DE MH, Ushio is open rated)

I am most concerned about him getting left behind on “out of spectrum” radiation that is soooooooo important it’s just starting to be counted…

1 Like

True, although I think people are overstating the importance of including far-red (700-750 nm) in PPFD measurements as “ePPFD” (in terms of net photosynthetic rate).

MH has far more spectral deficiencies than only the lack of 700-750 nm. Assuming you’re referring to ePAR from a 2021 paper by Dr. Bugbee and co-authors, and not referring to EPAR (which includes UV-A, FR, and NIR).

Check out the following papers if you haven’t read them. The first is the paper by Dr. Bugbee and co-authors, where they argue for including 700-750 in the PAR range, termed “extended-PAR” (ePAR). Also, check out my comments in this thread: PAR meters

Hps cost me 41k led would of been over 120k . My estimated power cost per year is about 100k, so the return is far longer then a year, and if I would of went with led the reduction of wattage is only 5000w max per 1050 sqft, some led company’s are recommending more wattage then hps now. And also some of the cooling cost is going to come at the expense of more propane re heat for my hvac.

@ralf my nursery will be completed this fall and I will be running 4200k cmh in there, I might just have to try and see how the de hps veg to get by.


Well my gavitas happily fire the new MH bulbs I got. I’ll send you a pic when I get in tomorrow.

My apogee meter is in as well. Will we be able to know based on ppfd if the light is firing correctly?


Weird that it say that they don’t have a hot restrike timer on the hps de. It definitely has it. So… someone on the gavita team is talking Bologna

You still running the hortilux ceramic hps? Little more blue and lots of far red. I’ve never put one in veg, but I know the Luxx will fire them.

1 Like

You actually veg under hps? I always see odd growth doing that

1 Like

Not regularly, but it works.

HPS vs. LED is a 55% difference in heat emissions. So you could literally reduce your heating/hvac cost by roughly 55%. That would put you pretty close to LEDS in one year based off some quick math, when you consider bulbs, fans and a few other things… Good luck with the output lighting though.

That’s not how that works. 1 watt of any light emits 3.412 btu of heat. Simple physics. The only way you’ll reduce cooling need is by reducing wattage.


I’m not since I don’t have the ceiling height for de’s at home, but one of the wholesalers I deal with have them, of I don’t go metal halide they would probably be a better option then hps for veg.
I really like the 315 cmh for veg but I’m not putting them in that room since it’ll be a flower room in six months.

1 Like