So, you're saying that all the diodes would sell easily? But... it's a bad idea, because someone else might mark the low-end ones up? Although, I see your point, it seems like the benefit would be worth it... People would love to buy the diodes with specs.
No. I'm saying that practically speaking, that first seller simply
wouldn't list the low end diodes at $15. There'd be no reason for them to. Demand would always keep the value of those diodes up above $15, and somewhere pretty close to what we pay for normal diodes now. There are lots of sellers selling 445s for $50 and still labeling them as 500 to 1,000 mW diode. You could truthfully label a diode at 1.3W, and still get every penny of that $50 price tag (especially selling to the public, like on eBay, etc). So simply put, while I can see a seller marking up the tested high end diodes, they would never substantially discount the low end. (Unless you're talking like 300mW zombies). As much as that model may make sense in your head (or in a closed system), it doesn't work in reality, where each seller is just one of many harvesting the same diodes.
I also think that while 1.3W to 1.9W is a noticeable difference, those other gradations are relatively meaningless for their price. Certainly 100mW power differences are unnoticeable, and I know that I would end up just buying diode from the cheaper tiers because I could put that money into a better lens or something longer lasting. In the end the going rate of $45 for an M140 diode is quite reasonable.
Agreed. It's one thing for people to say that they can spot the different between 200mW and 250mW of DPSS like 532nm. It's an entirely different (and much more absurd) proposition for people to suggest that they can recognize 25% power differences in "445nm diodes". Using my spectrometer for the last week or so has really revealed to me the patent absurdity of any claim that someone can "eyeball" the differences.
Why? Because these "445s" all differ so hugely in wavelength. Granted, I've only spectrometered a couple dozen 445s so far, and that's not nearly as many as some of the guys on PL have tested, but even accross the 30 or so I've measured for wavelength, I've see a range of 441 to 453nm. That has HUGE implications for how bright a beam appears.
Now, there's a lot of debate over how well the various "relative brightness" calculations work for vastly different wavelengths (like comparing red to green), but I think the figures are fairly trustworthy for relatively close wavelength comparisons. So, imagine you run a multi-mode blue at 1,250mW of output. Not too impressive right? Well, if that's 1,250mW of 453nm, it would appear as bright as
OVER 2W of 441!
Think about that for a second. Two diodes that you thought to be essentially identical, and yet one of them (the 453) running at 1,250mW would appear brighter than the other at 1,250, or 1,500, or 1,750 - in fact, they wouldn't appear to be on par until you got the 441 to ~2,018mW.
Beam: (453nm 1250mw) vs. (441nm 2018.21mw)
The assortment of diodes I have probably reflects the variety in most people's collections. Whether you've got two multi-mode-blues or twenty, chances are good that you've got different wavelengths, and perhaps wavelengths different by a fair amount. While you may not perceive a difference in colour between say 445 and 450, there's going to be a substantial difference in brightness. Even that little 5nm jump accounts for an extra ~20% in brightness.
So, basically what I'm saying is that anyone who claims a talent at spotting power differences between different "445s" is full of bologna, unless they know that the wavelengths are identical.
Perhaps the takeaway is really that we should be spectrometering all of our diodes. I'm certainly doing that with every diode I use from this point onwards.
