DHP Presents Citizen Cob 1812 3500k 80CRI Vs 1812 3500k 90 CRI

The Dawg

Well-Known Member
Wanted to ask if you found a way around this problem? is this the reason why you went with the 1050mW drivers for two lights in this side by side thread?
I Went With The 1050mA Drivers With The 1812 Cause Their 50 Volt. Now The Above Statement Was Made When I Tried To Run 4 1212's Which Are 37.5 Volt On A HLG-240-1750mA Driver. My Solution Was To Drop To A HLG-185-1400mA Driver For My 1212's :hump:
 

MediheaLed

Well-Known Member
I am very disappointed. Right when I upgrade to COB, you guys decide, now, after the chips have been around for years, to test CRI. :cuss:Somebody tell me why it took so long? :shock:
 

HighOnDIYLife

Active Member
I am very disappointed. Right when I upgrade to COB, you guys decide, now, after the chips have been around for years, to test CRI. :cuss:Somebody tell me why it took so long? :shock:
The efficacy of the cri90 chips has improved significantly over the time. I think people never considered them as an option before because of that. But they put out an acceptable to good amount of light now.
 

BuddyColas

Well-Known Member
Thanks for the effort documenting this. It shows similar results to the testing I did. High CRI = better even though it provides lower PPF.
I know the short version is: "Redder is better!" I agree. Please share the LER and QER you use for the 3000K/80 and 3000K/90 calculations. Thanks.
 

Rahz

Well-Known Member
Someone mentioned 272 as LER for 3000/90. I suspect if accurate it's based on the v2 spectrum. I haven't calculated it myself because there's no graph to digitize. The LER and QER for 3000/80 gen7 is 332 and 4.68. I have lost the spreadsheet for the calculations and experienced some confusion because I had calculated the LER for 3000/80 Vero v2 at 322, but I'm positive 332 is the correct figure for gen7 because the efficiency figure at 700ma came out to be 57%. Compared with Bridgelux product sim indicating 189 LPW at that current 332 LER is the correct figure. Also, someone else digitized the gen7 3000/80 graph and came up with 333.

I did take PAR measurements for the 1825 samples I used. Figuring the LER for them based on the Vero sample as ratios of the PAR readings will suggest:

3000/90: 314
2700/90: 285

How accurate that is (for Vero or Citizen) I'm not sure. With a 10 point spread between v2 and gen7 it's very possible the Citizen samples are lower than the math suggests. It's very sloppy and shouldn't be used as established figures. All the Citizen spectrums are available if anyone wants to digitize them 8) The QER won't scale as well as a ratio so I'm not going to bother posting those figures.

Anyway, redder is better within limits. 3000/90 did better than 2700/90 but whether it's because the red saturation point was reached or because the 2700 sample lacked the efficiency to keep up I'm not sure... but either way the real world test so far is indicating 2700/90 isn't the best. It will have a chance in the next test to provide a different result.
 

BuddyColas

Well-Known Member
Someone mentioned 272 as LER for 3000/90. I suspect if accurate it's based on the v2 spectrum. I haven't calculated it myself because there's no graph to digitize. The LER and QER for 3000/80 gen7 is 332 and 4.68. I have lost the spreadsheet for the calculations and experienced some confusion because I had calculated the LER for 3000/80 Vero v2 at 322, but I'm positive 332 is the correct figure for gen7 because the efficiency figure at 700ma came out to be 57%. Compared with Bridgelux product sim indicating 189 LPW at that current 332 LER is the correct figure. Also, someone else digitized the gen7 3000/80 graph and came up with 333.

I did take PAR measurements for the 1825 samples I used. Figuring the LER for them based on the Vero sample as ratios of the PAR readings will suggest:

3000/90: 314
2700/90: 285

How accurate that is (for Vero or Citizen) I'm not sure. With a 10 point spread between v2 and gen7 it's very possible the Citizen samples are lower than the math suggests. It's very sloppy and shouldn't be used as established figures. All the Citizen spectrums are available if anyone wants to digitize them 8) The QER won't scale as well as a ratio so I'm not going to bother posting those figures.

Anyway, redder is better within limits. 3000/90 did better than 2700/90 but whether it's because the red saturation point was reached or because the 2700 sample lacked the efficiency to keep up I'm not sure... but either way the real world test so far is indicating 2700/90 isn't the best. It will have a chance in the next test to provide a different result.
Thanks for your prompt detailed response.:mrgreen:
 

DankaDank

Well-Known Member
Someone mentioned 272 as LER for 3000/90. I suspect if accurate it's based on the v2 spectrum. I haven't calculated it myself because there's no graph to digitize. The LER and QER for 3000/80 gen7 is 332 and 4.68. I have lost the spreadsheet for the calculations and experienced some confusion because I had calculated the LER for 3000/80 Vero v2 at 322, but I'm positive 332 is the correct figure for gen7 because the efficiency figure at 700ma came out to be 57%. Compared with Bridgelux product sim indicating 189 LPW at that current 332 LER is the correct figure. Also, someone else digitized the gen7 3000/80 graph and came up with 333.

I did take PAR measurements for the 1825 samples I used. Figuring the LER for them based on the Vero sample as ratios of the PAR readings will suggest:

3000/90: 314
2700/90: 285

How accurate that is (for Vero or Citizen) I'm not sure. With a 10 point spread between v2 and gen7 it's very possible the Citizen samples are lower than the math suggests. It's very sloppy and shouldn't be used as established figures. All the Citizen spectrums are available if anyone wants to digitize them 8) The QER won't scale as well as a ratio so I'm not going to bother posting those figures.

Anyway, redder is better within limits. 3000/90 did better than 2700/90 but whether it's because the red saturation point was reached or because the 2700 sample lacked the efficiency to keep up I'm not sure... but either way the real world test so far is indicating 2700/90 isn't the best. It will have a chance in the next test to provide a different result.
Bridgelux need to get there shit together and take some lessons from Citizen when it comes to SPD's
All we get is some shitty hand drawn graphs for 3000K 80 CRI; 4000K 80 CRI; 5000K 70 CRI and 6500K 70 CRI

Maybe @Stephenj37826 can hit up a GEN7 90CRI under the spectrometer.
 

HotBunz

Active Member
I have been meaning to expand my current cxb3590 fixture running a hlg-185h-1400b driver. I first built it with 2 3000k 80cri COBs. After seeing this thread I order up a pair of 3000K 90cri's and finished out the fixture last night.

Given the numbers the Citizen COBs are getting most likely any future COB fixtures I build will be Citizen based and not Cree. Only thing I don't like with the Citizen's is the higher voltage and the need for a bigger driver.
 

Stephenj37826

Well-Known Member
I have been meaning to expand my current cxb3590 fixture running a hlg-185h-1400b driver. I first built it with 2 3000k 80cri COBs. After seeing this thread I order up a pair of 3000K 90cri's and finished out the fixture last night.

Given the numbers the Citizen COBs are getting most likely any future COB fixtures I build will be Citizen based and not Cree. Only thing I don't like with the Citizen's is the higher voltage and the need for a bigger driver.

Higher voltage but not bigger driver. Also the 1212 is 36 volt
 
Top