After some googling, the space seems inundated with pseudo-science snake oil garbage. Any recommendations for high powered broad spectrum lamps?
Do consider therapy, a new relationship and/or a new job though.
I keep it on the medium setting for 10 minutes and the rest of the day feels so much better than days where I don't. I can go outside in 0°F and it still feels warm. It totally fixed my energy and motivation issues with SAD.
Here's my Ask HN thread about SAD treatments, for more ideas: https://news.ycombinator.com/item?id=18257011
They cite Terman that dawn simulation is not quite as good as light box therapy but better than placebo (true!), however..
They also claim that dawn simulators are "really nothing more than a timer and a rheostat (a device to slowly change electrical current) hooked to your bedside lamp." This is FALSE, and Terman himself hates such devices, explicitly calling them out in the 2013 review I detail in my main reply. The devices used by Terman in his trial that demonstrated dawn simulation efficacy were on tripods so that they projected light downward, from a meter above the bed [1].
This is exactly the sort of misinformation that enables the snake oil the OP complained of. But I don't mean to make you feel bad; the presentation is of high enough quality it would've fooled anyone who hadn't actually read the original research. Your recommendation of the Daylight by Carex/Uplift is terrific; it has indeed been used in studies and is probably the most reasonably priced of the reputable light boxes.
[1]: https://www.researchgate.net/publication/6648580_Controlled_...
The main benefit I've noticed that it doesn't matter what season it is, the sun always rises at the same time, which makes for a happier circadian rhythm (for me at least).
(I live in the northern U.S where there are long winters and daylight saving is observed, which makes for variable sunrise times over the course of the year)
I don't use the alarm feature though -- my phone is my alarm.
A major recent paper on the use of light therapy is the 2013 review by Terman & Terman [1]. If you're mildly scientifically inclined, it's an easy enough read and includes valuable information on determining the optimal timing window for light therapy, which varies (sometimes substantially) from patient to patient, as well as an overview of successful light therapy regimens and the use of melatonin.
(Note on melatonin, from that paper - optimal doses are on the order of 0.2 mg. Pills at the drugstore are usually 3-5 mg, so strong they'd actually further disrupt circadian regulation. Even 0.5mg is usually too much.)
Terman himself authored the main paper exploring the use of dawn simulation as an alternative to traditional lightbox therapy. He too is incensed at the snake oil you're finding: "These findings led to a spate of untested commercial 'dawn alarm clocks' that differ, however, from the clinical trial device ... which [are] easily missed by turning away on the pillow. Such alarm clocks have no demonstrated efficacy." The 'dawn alarm clocks' he is railing against squarely includes such devices as the Philips Wake-Up HF3520 and the Lumie Bodyclock Starter mentioned by other responses here.
In addition to the techniques discussed in Terman's review, the AutoMEQ Morningness-Eveningness Questionnaire [2] can help you determine the ideal time window for light therapy.
One way to fight sketchy lux claims is by calculating things yourself. The issue is that lumens is a measure of output, while lux is a measure of received light. Converting between the two requires specifying how the output is spread (ie everywhere/spherically vs more directionally).
The lower calculator here [3] allows you to convert lumens to lux. You can determine the type of bulb used in any reputable light box and you can look up its properties, like lumens, on any decent bulb webstore, e.g. 1000bulbs. Fluorescent especially put out less and less over their lifetime, so the listed 'initial' lumens isn't really appropriate, you want the mean lumens if possible.
For example, my lightbox [4], a reputable one that's been used in clinical trials, claims 10000 lux at 12". So, we calculate: Three 36 W 2g10 bulbs. 2400 lumens (mean) per bulb. Sanity check: 66 lumens/W, nicely within the 60-80 expected [5] range. Now to the calculator - 7200 lumens, 0.3 meters, and we start with a beam angle of 180 (hemispherical). The calculator returns 12732 lux. In other words, yes, this lightbox is capable of hitting 10000 lux at 12", as long as it sends all the energy forward hemispherically, which is a close enough approximation to truth.
Finally, once you're laying out $100 on a therapy lamp, you might as well just drop $13-$17 more [6][7] and get a simple lux meter and then you can be sure of exactly the amount of light received at your retina. I have the second [7] meter and can attest to its ease of use and accuracy. It's also handy to be able to measure after several months of use the light now produced by your fluorescents, so that you can replace them or lengthen your therapy accordingly as they drop.
Oh - one last tip. Depending on your morning routine, you may want to do light therapy at your computer. The 12" distance needed by many lightboxes is very short. I mounted my lightbox onto a vesa monitor arm (just screwed through the back housing) that is installed behind and above my monitor. For my light therapy, I can swing the light into place, and when I'm done I tuck it back out of the way.
1: https://www.researchgate.net/publication/264037149_Chronothe...
2: https://www.researchgate.net/file.PostFileLoader.html?id=56a...
3: http://www.ledstuff.co.nz/data_calculators.php
4: https://smile.amazon.com/Carex-Health-Brands-Day-Light-Class...
5: https://en.wikipedia.org/wiki/Luminous_efficacy#Examples_2
6: https://smile.amazon.com/Proster-Luxmeter-Illuminance-0-1-20...
7: https://smile.amazon.com/Digital-Luxmeter-Illuminance-Displa...