Can a light get too efficient

[terran]

Light is just the absence of dark. Lights don't really put out light, they just suck up dark - that's why electronic schematics always label lights as "DS1, DS2, etc". DS stands for "Dark Sucker". Dark has weight too - it sinks to the bottom of lakes and oceans and makes it all dark down there.

Its TRUE, I read it on the internet.

That makes perfect sense. Like an AC sucking up heat, taking it outside and converting it into light or rather anti dark matter.
I thought that they stored it at the end of tunnels.

An electrician friend told me years ago. "It's all fire."

Not wishing to interrupt this thread anymore. It's turned itself around. Good info here.

 

[Rahz]

The benefit of 70 CRI is that you get more photons for your watts. So when you keep the power input the same then it would beat the others on higher light intensity

Understood but what I'm saying is, that hasn't been my experience. I'm specifically running lamps with equal power input. Obviously a couple test runs aren't good proof but results are consistent so far. High CRI is performing beyond expectations.

 

[PurpleBuz]

Understood but what I'm saying is, that hasn't been my experience. I'm specifically running lamps with equal power input. Obviously a couple test runs aren't good proof but results are consistent so far. High CRI is performing beyond expectations.

specifically what "high CRI" lights are you comparing ? e.g. 3000K 80 cri vs 3000K 90 cri ?

 

[wietefras]

Understood but what I'm saying is, that hasn't been my experience. I'm specifically running lamps with equal power input. Obviously a couple test runs aren't good proof but results are consistent so far. High CRI is performing beyond expectations.

Depending on the difference in photon output that could also be the case. In general high CRI would score higher on McCree RQE since it has more red and less blue. Which could offset or overcome a small difference in photons.

Although the Samsung LM561C 3000K SPD's I checked, the 80 CRI scored higher than 90CRI and 70 CRI was lowest of the three.

 

[alesh]

Although the Samsung LM561C 3000K SPD's I checked, the 80 CRI scored higher than 90CRI and 70 CRI was lowest of the three.

This would be true for most manufacturers. High CRI LEDs usually have too much far red to score any better.

 

[Rahz]

specifically what "high CRI" lights are you comparing ? e.g. 3000K 80 cri vs 3000K 90 cri ?

Just got through doing 3000K 70 CRI vs 3500K 80CRI. In the past 90 CRI has beaten both 70 and 80 CRI. I'm currently finishing up a 3000K 90CRI grow. It's not a side by side and It's Vero 18g7 so it might not be the best GPW I've every had but it's looking like it will be the best GPPW I've ever had.

Edit: and it's in soilless. If I beat my GPPW record I'll be second guessing the benefits of hydro.

 

[Rahz]

Depending on the difference in photon output that could also be the case. In general high CRI would score higher on McCree RQE since it has more red and less blue. Which could offset or overcome a small difference in photons.

The photon difference between 3000K 70CRI and 3000K 90CRI is significant. Adjustment does favor the high CRI to a small degree but not nearly enough to offset the PAR reading.

 

[Rahz]

Here's some data

PAR test results
3000K70CRI 950
3500K80CRI 921
3000K80CRI 905
3000K90CRI 856
2700K90CRI 780

Conversion factors for plant response from 400-700nm (based on Mcree curve)
3000K/70CRI 78.034
3000K/80CRI 78.229
3500K/80CRI 76.566
2700K/90CRI 80.460
3000K/90CRI 78.355

You can see high CRI benefiting, but not by much. It lines up more by Kelvin.

Par values multiplied by the conversion factor
3000K/70CRI 741
3000K/80CRI 708
3500K/80CRI 705
2700K/90CRI 627
3000K/90CRI 670

 

[PurpleBuz]

an interesting side by side would be the 2700K90CRI vs 3000K80CRI

 

[wietefras]

Conversion factors for plant response from 400-700nm (based on Mcree curve)

McCree RQE curve is much wider though. I though something like 320nm to 780nm. YPF is calculated over 360-760. Wouldn't it skew the conversion factors if you're using only 400-700 and then especially for high CRI SPD's?

Depending on the PAR meter (like the previous version Apogee meter) you could also miss out on measuring a lot of red in your PAR readings. Again working especially against the high CRI SPD's.

Also, is GPPW really a useful metric? I would think it gets rather slanted because you get different amounts of photons from the same PAR watt depending on wavelength.

Perhaps something like g/PPF? Or maybe YPF would be better still.

I use g/PPF myself and I find that for me it's even reasonably comparable between leds and HPS.

 

[Rahz]

Wouldn't it skew the conversion factors if you're using only 400-700 and then especially for high CRI SPD's?

A small amount, but again not enough to alter the predictions. I would factor it in but the PAR meter drops off to almost zero at 700.

Depending on the PAR meter (like the previous version Apogee meter) you could also miss out on measuring a lot of red in your PAR readings. Again working especially against the high CRI SPD's.

Measurements were taken with an Apogee SQ-520 and spectral response was corrected from 400-700.

Also, is GPPW really a useful metric? I would think it gets rather slanted because you get different amounts of photons from the same PAR watt depending on wavelength.

It's a useful metric when it comes to efficacy of spectrums by energy output.

 

[ANC]

I've only seen one 80 vs 90 CRi side by side grow... the 90s were significantly larger.

 

[wietefras]

@Rahz, I see quite big differences between YPF and PPF conversion factors though. On the 90 CRI at least.

It's a useful metric when it comes to efficacy of spectrums by energy output.

Using PAR watt is not about efficacy but about efficiency though.

The point is all these errors are accumulating and skewing the comparison. For a comparison where there isn't a whole lot between the things

 

[wietefras]

I've only seen one 80 vs 90 CRi side by side grow... the 90s were significantly larger.

Well Rahz had one done and the 90 CRI stretched out of the top of their box and then took twice the area to grow.

Dawg did a test and the 90 CRI yielded less.

Greengenes reported about some of these tests done where they didn't see any winner in these comparisons.

So I'd say we have rather conflicting anecdotal evidence.

 

[ANC]

Yeah I don;t know if there was a yield dif, only the 90 side was about a foot taller and much bushier.

 

[Rahz]

So I'd say we have rather conflicting anecdotal evidence.

No disagreement there, but my own work has shown a pattern. If it's in error time should point it out.

 

[Rahz]

Using PAR watt is not about efficacy but about efficiency though.

The point is all these errors are accumulating and skewing the comparison. For a comparison where there isn't a whole lot between the things

I primarily use GPPW to determine the overall efficacy over various grows/mediums/lamps along with my skill in tending the plants. I don't know for sure but I suspect I'm bumping up against the plants limitations in this most recent grow. GPPW was higher at 60 g/sqft than at 73 g/sqft. That seems reasonable.

But none of those numbers matter when doing a side by side. Wall watts in, yield out is the primary indicator, may the best lamp win and what most people are interested in. The tests I do are most suited for a fair comparison of the chips efficacy rather than spectral efficacy, and while I don't have perfect numbers to look at spectrum alone in the same controlled manner all this anecdotal evidence is worth considering. I don't think contradictory evidence is wrong. I'm still in the I don't know phase, but I have noticed a pattern and it's worth mentioning. It's not unreasonable to consider the possibility that a yield factor based on plant mass may not translate perfectly into flowering mass.

 

[nevergoodenuf]

The biggest difference I notice is stretch, until I get around 4000k. I had 3000k/90cri, 3500k/80cri, and 4000k/70 or 80 (can't remember) over a 4'x8' tray next to another tray with 2 gavitas. 3000k/90 was almost identical to the gavita side in height and leaf/branch density. The 3500k/80cri was a step shorter, but had similar leaf/branch density. The 4000k/??cri was another step shorter, but the leaf/branch density was 75% of everything else. We pulled the 4000k and replaced it with 3000k/90cri @ around week 3, but that end never quit caught up.

 

[wietefras]

I primarily use GPPW to determine the overall efficacy over various grows/mediums/lamps along with my skill in tending the plants. I don't know for sure but I suspect I'm bumping up against the plants limitations in this most recent grow. GPPW was higher at 60 g/sqft than at 73 g/sqft. That seems reasonable.

I just don't get why you use two metrics based on watts. Why not use the usual g/W and something that relates g to umol? The latter already corrects a lot for spectral differences where PAR W doesn't.

Average PPFD is a standard way of checking what the plants can deal with.

I personally doubt I'd be able to have consecutive grows yield within 5% of each other. Temperatures can differ, humidity, perhaps some root issues or plant issues. Growing is not an exact science unless you have full control over all parameters. Even then. There is a large statistical error on tests like this.

 

[Kassiopeija]

For purposes of photosynthesis blue has more energy than a plant needs. When a pigment is excited with a blue photon it must lower the energy level before transferring the energy to the carbon reaction center.

the rest of what's being "left-over" is converted to heat inside the leaf