• Here is a link to the full explanation: https://rollitup.org/t/welcome-back-did-you-try-turning-it-off-and-on-again.1104810/

Rahz' notes and observations

Randomblame

Randomblame

Well-Known Member
lukio said:
well ive ordered one so that'll help...

It was actually £125 delivered, we normally get a lil more robbed in the UK.

@Randomblame ah right thanks!
Click to expand...
I hear ya!! It's usually the same with euros. But $US is quite low these days. Got 1,23$ per € currently, which is the best since a year or so..

Screenshot_20180327-154003.png

As long as Trump continues to threaten punitive tariffs for Europe's imports, the ECB will not print new euros to bring the price back to ~1 : 1. What's good for us is not neccessarily good for everyone, lol!
 
Last edited:
Randomblame

Randomblame

Well-Known Member
Slinging PAR said:
We need more people to buy them to bring the price down. They should be as cheap or cheaper than lux meters but they just don't have the volumes of sales to justify discounts.
Click to expand...
Yeah, mainly because they are essentially based on the same technology, only the values are interpreted differently and other/no filters are used. I've seen a thread here in the LED section where someone turned a lux meter into a PAR meter, even with reasonably accurate numbers.
But it was at least 2 years ago, so don't ask me for a link..
Some of the older members may remember ..
 
Rahz

Rahz

Well-Known Member
Testing 3500K 80CRI vs 3000K 90CRI is complete. This is the first win for low CRI in the testing I have done, with the 3500K 80CRI coming in at about 10% more yield which is pretty notable. Total run under 8 CLU058-1825s at 50w each 588g so not my best run under either sample, coming out at 1.47 GPW total. Works out to 1.4 GPW under high CRI and 1.54 GPW under the low CRI.

Same strain as before, plants stayed very healthy and end product looked nice. Nothing specific I can mention to explain the results.

I'm still leaning towards high CRI, but this makes the choice more ambiguious than before. I'm going to put my spectrum testing aside for now. For the next few grows at least I'm going to focus on pitting different nutrient formulas against one another, not necessarily brand vs brand, but I'm trying to get to the bottom of the optimal NPK elemental ppm ratios. I'm using 100/50/150 as a base line and will play around with boosting the PK to around 100/250 to see how much of a difference it makes if any. I don't plan on controlling other inputs (boosting NPK ppm/TDS will involve higher levels of cal/mag and micros) but for my purposes it's a good starting place for these types of tests.
 
OneHitDone

OneHitDone

Well-Known Member
Rahz said:
Testing 3500K 80CRI vs 3000K 90CRI is complete. This is the first win for low CRI in the testing I have done, with the 3500K 80CRI coming in at about 10% more yield which is pretty notable. Total run under 8 CLU058-1825s at 50w each 588g so not my best run under either sample, coming out at 1.47 GPW total. Works out to 1.4 GPW under high CRI and 1.54 GPW under the low CRI.

Same strain as before, plants stayed very healthy and end product looked nice. Nothing specific I can mention to explain the results.

I'm still leaning towards high CRI, but this makes the choice more ambiguious than before. I'm going to put my spectrum testing aside for now. For the next few grows at least I'm going to focus on pitting different nutrient formulas against one another, not necessarily brand vs brand, but I'm trying to get to the bottom of the optimal NPK elemental ppm ratios. I'm using 100/50/150 as a base line and will play around with boosting the PK to around 100/250 to see how much of a difference it makes if any. I don't plan on controlling other inputs (boosting NPK ppm/TDS will involve higher levels of cal/mag and micros) but for my purposes it's a good starting place for these types of tests.
Click to expand...
How will you get higher PK with Cal/mag?
 
Rahz

Rahz

Well-Known Member
OneHitDone said:
How will you get higher PK with Cal/mag?
Click to expand...
I'll get higher cal/mag when I add the PK. I might play around with MPK or other booster but I can get pretty close to those numbers using Maxi series and silica, which would cause cal/mag to be about 20% higher than running at 100/50/150. I don't think it's a big deal.

Problem with many PK boosters is too much P (imo). I'd rather not take P over 100, at least not until I have some other results.
 
OneHitDone

OneHitDone

Well-Known Member
Rahz said:
I'll get higher cal/mag when I add the PK. I might play around with MPK or other booster but I can get pretty close to those numbers using Maxi series and silica, which would cause cal/mag to be about 20% higher than running at 100/50/150. I don't think it's a big deal.

Problem with many PK boosters is too much P (imo). I'd rather not take P over 100, at least not until I have some other results.
Click to expand...
Would seem more valuable data if you made your own formulas or sourced some that maintained a stable micro, cal, mag ratio and only shifted the NPK's on top of that
 
Rahz

Rahz

Well-Known Member
OneHitDone said:
Would seem more valuable data if you made your own formulas or sourced some that maintained a stable micro, cal, mag ratio and only shifted the NPK's on top of that
Click to expand...
Perhaps. I haven't dived into DIY mixing from scratch so it's a little intimidating. I'm not advanced enough to know how those relationships should change at different NPK ppm levels, but one consideration is that if a plant is uptaking more NPK then it will benefit from/need more cal/mag and micro. I do know that cations compete with other cations, etc. so sticking with Maxi (both Gro and Bloom have micros) should be safe in that regard. It might be the most fair way to do it as well, but I need to do some more reading on the subject.

CobKits said:
Thanks for the feedback Rahz

seems at this point the results are still all over the map like we brought our shotgun to the rifle range.
Click to expand...
True. I would like to have included some 4000K results but they probably would have been similar as well. I think I've shown that 90 CRI can do just as good a job as 80 for full cycle. Switching from 70 or 80 CRI in veg to 90 CRI in flower will shave a few inches off total plant height. In some cases that alone will justify using 90 CRI in the flowering chamber when height is a concern. But that's all vegetative characteristics and doesn't have much to do with yield.

I'm sure someone with a better facility can/will eventually pin down exactly which spectrum produces best for the electricity used, but I suspect results are all over the map because the efficacies between the samples we've been testing are very small. Interesting to note that the current Vero29C in Bridgelux sim shows the exact same umol/j for different spectrums at a particular current.
 
Last edited:
ttystikk

ttystikk

Well-Known Member
Rahz said:
Perhaps. I haven't dived into DIY mixing from scratch so it's a little intimidating. I'm not advanced enough to know how those relationships should change at different NPK ppm levels, but one consideration is that if a plant is uptaking more NPK then it will benefit from/need more cal/mag and micro. I do know that cations compete with other cations, etc. so sticking with Maxi (both Gro and Bloom have micros) should be safe in that regard. It might be the most fair way to do it as well, but I need to do some more reading on the subject.



True. I would like to have included some 4000K results but they probably would have been similar as well. I think I've shown that 90 CRI can do just as good a job as 80 for full cycle. Switching from 70 or 80 CRI in veg to 90 CRI in flower will shave a few inches off total plant height. In some cases that alone will justify using 90 CRI in the flowering chamber when height is a concern. But that's all vegetative characteristics and doesn't have much to do with yield.

I'm sure someone with a better facility can/will eventually pin down exactly which spectrum produces best for the electricity used, but I suspect results are all over the map because the efficacies between the samples we've been testing are very small. Interesting to note that the current Vero29C in Bridgelux sim shows the exact same umol/j for different spectrums at a particular current.
Click to expand...
The reason I became such a convert to dry nutrient mixes was because of inconsistent results with water bottled nutes.

While it's possible to just mix big batches and use the same mix in both parts of a side by side, it's even better to be sure you know what's going into your res. The water bottle boys never tell you everything.
 
Rocket Soul

Rocket Soul

Well-Known Member
Thanks Rahz, its wonderfull to see theres still people doing stuff like this, side by sides etc and sharing results.

As you like your high CRI spectrums, what do you think about this one?? Its bridgelux Vesta strips, 2 channels: 5000k / 2700k both in 90 Cri. The red line is samsung 3000k 80cri. I like how the whole spectrum looks more flat than the standard 80/90 cri that have more of a peak. 660nm at 80% of max, 680nm at around 50% of max. The question id like to pose is rather than trying to fit in as much light as you can on the most efficient wavelengths, does it make more sense to try to distribute the spectrum more wide, over all of the different pigments, while at the same time maintaining high output in reds, even those 680-700nm which are normally missing in phosporbased leds.

I know the efficiency of these chips are low comparde to sammies and some cobs, but you can allways get more light with more power.
uyr0 (1).png
 
Randomblame

Randomblame

Well-Known Member
The curve looks so flat because they are compared to another strip and normalized. That below shows how each spectrum looks like without the other strip.

Screenshot_20180622-121737.png

There is nothing wrong with a ~3850°k/CRI90 spectrum, for sure very productive, but effiency matters more than spectrum. For this reason you will still get more weed from an F-strip compared to a Vesta strip at the same watts.
You need to run the Vesta's at only half current to reach almost the same effiency as with EB strips.
Compare two Vesta's at half current to one F-strip like above and the numbers would be much closer. Maybe ~52 vs. 54,7% efficiency. The curve would go almost as high as the F-strip curve but still more deep and far-red from higher CRI.
 
Rahz

Rahz

Well-Known Member
Rocket Soul said:
Thanks Rahz, its wonderfull to see theres still people doing stuff like this, side by sides etc and sharing results.

As you like your high CRI spectrums, what do you think about this one?? Its bridgelux Vesta strips, 2 channels: 5000k / 2700k both in 90 Cri. The red line is samsung 3000k 80cri. I like how the whole spectrum looks more flat than the standard 80/90 cri that have more of a peak. 660nm at 80% of max, 680nm at around 50% of max. The question id like to pose is rather than trying to fit in as much light as you can on the most efficient wavelengths, does it make more sense to try to distribute the spectrum more wide, over all of the different pigments, while at the same time maintaining high output in reds, even those 680-700nm which are normally missing in phosporbased leds.

I know the efficiency of these chips are low comparde to sammies and some cobs, but you can allways get more light with more power.
View attachment 4154392
Click to expand...
I've been wanting to see a radiant watt normalized blurple vs white for a long time, but if my results touch on the answer then spectrum isn't a huge factor in yield. To what degree the spectrum can be changed before it makes a difference I'm not sure. Vesta strips might be worth playing with in a short cabinet, but I would lean towards a low CRI for the veg phase. I didn't even know 5000K came in 90 CRI.

Like Randomblame says, efficiency matters. If you ran enough Vesta strips to equal the flux of F series... my results weren't conclusive enough to suggest anything other than "they'll work about the same". I would (do) want as much light for the least electricity so I would pick the F-strips over Vesta.
 
OneHitDone

OneHitDone

Well-Known Member
@Rahz, how come you never have any photo's to accompany the data? Would give a great visual perspective for the readers to go off of :peace:
 
B

Baura

Well-Known Member
Randomblame said:
The curve looks so flat because they are compared to another strip and normalized. That below shows how each spectrum looks like without the other strip.

View attachment 4154397

There is nothing wrong with a ~3850°k/CRI90 spectrum, for sure very productive, but effiency matters more than spectrum. For this reason you will still get more weed from an F-strip compared to a Vesta strip at the same watts.
You need to run the Vesta's at only half current to reach almost the same effiency as with EB strips.
Compare two Vesta's at half current to one F-strip like above and the numbers would be much closer. Maybe ~52 vs. 54,7% efficiency. The curve would go almost as high as the F-strip curve but still more deep and far-red from higher CRI.
Click to expand...
Agreed, but when the budget is short(my case), ppf/$ maters more than efficiency.

as randonblame posted,

The vesta are $9 and 112ppf = 12ppf/$
The f-series are $27 and 131ppf = 5ppf/$
and
The eb g2 are $6,5 and 32ppf = 5ppf/$


but if the high cri have more red and far red, the ppf number does not take in account the emerson effect right?


Like, if we supose that:
- a certain led 3000k 70cri has 100ppf
- a certain led 3000k 90cri has 100ppf


The ppf number is the same, but wouldnt the emerson effect make the 3000k 90cri "give" more energy to the plant?
 
B

Baura

Well-Known Member
nfhiggs said:
If thats the case then just buy the cheapest LED lights you can find. You can get 100W COBs for $2 each, and Acuity strips for the same price on Ebay.
Click to expand...
I don’t buy shitty tires, neither shitty brake components for my car. There are some things I won’t mess with.

I won’t buy chepo leds and drivers.
I’m using GE immersion led drivers(for refrigerated led lightning) for my 048-1812(50v 1.2a) I found them cheep enough(u$15 each driver here in Brazil) and GE a reliable brand, not a meanwell 96%eff, but good enough. This was the best $/w I found on a reliable driver locally sourced.

Wen I said the best ppf/$, I’m not going “full retard” on numbers. Quality and reliability are a must.

I’m seeking the best $/ppf with quality components.
Aiming high efficiency (2.5 umol/J) gets really expensive here in Brazil. I believe 2-2.2 umol/J might be a good target in third world countries
 
You must log in or register to reply here.
Top