Lexel wrote:

what about tint mixing on this board with separate LED pads for 2 colors like one column cw one WW and separate pads for both colors?

You can tint mix freely in any places with almost all 757 family except Optisolis. Mixing between CW/WW Optisolis can only be done between series strings.

[Clemence]

Lexel wrote:

clemence wrote:

Lexel wrote:

what about tint mixing on this board with separate LED pads for 2 colors like one column cw one WW and separate pads for both colors?

You can tint mix freely in any places with almost all 757 family except Optisolis. Mixing between CW/WW Optisolis can only be done between series strings.

[Clemence]

I am talking of indipendant LED positive pads for each tint

Not with this one. This MCPCB designed mainly for cost saving, voltage/LED choice flexibility, and easy application.

[Clemence]

AEDe wrote:

Is there any suitable сooler and optics for this dimension?at ali or other shops?

I use simple back frosted 10mm thick square acrylic glued to it

Pavlo wrote:

This is pretty cool! Any ideas on how this can be used in a portable setup?

Use buck driver with 36P or boost driver with 3S12P setup. You don’t need powerful driver with Optisolis, 65mA – 200mA per LED is all it needs

[Clemence]

When I said “cost effective” I really meant it. With 5steps R70 or R8000 GR-V3 you can get the MCPCB + 36pcs LEDs for as low as $5 ($7,15 for GR-V3 R95). R95 rank is already very close to Optisolis, especially at warm CCT.
GR-V3 R95 ranked above E21A and below Optisolis. With Optisolis the board will cost $16,24 – $22,4 (MCPCB + 36pcs LED), CCT dependent. Add + $5 per board if you want me to solder them all. But seriously, soldering 757 LED family is very easy unlike E17A or E21A.
You can save your money by using Optisolis only in certain areas. The cost/CRI point is still too high. I held back from trying the new Optisolis COB because it’s very expensive, and as per now, only comes in limited CCT (AFAIK, 3000K/5000K) and 36V only.

CRI comparison:

.
[Clemence]

The boards are still in the making, delivery schedule is 190515. Crash test will follow later. I still have no idea what is the maximum wattage this board can handle. Nothing fancy here, just normal good quality MCPCB.

I’m getting to like mid power LED these days. Spent few weeks designing the boards and browsing through hundreds of pages of data sheets for a project (that led to this VS35SP36). And as I dug deeper, I like the idea even more. I know the idea of high efficiency lighting is not widely appreciated in BLF but here’s my thoughts to share.

The beauty of mid power LEDs are: low power – low cost – high efficiency – power in numbers. Of course you can get the same output from a single XHP70.2 pushed very hard, but there’s no way you can get the efficiency (wattage/output, cost/module) of multiple mid power LED. I don’t see any benefit for manufacturers to use high power LED in general lighting unless smaller form factor is to be desired. Especially driven at below to max OEM rated current for reliable and predictable L70 life. This explains why more than 80% of Nichia’s LED sales come from low to mid power LEDs. I believe the same also applies to Cree and other manufacturers.
With high power LED, you can get slightly higher efficiency than mid power LED if they’re driven at the same low current. But what’s the point? At whooping 10x – 20x the cost? Development in high power LED is very similar to F1 racing. While mid power LED development is closer to 24 hours Le Mans racing. The latter has more real world applications. It would takes years before F1 technology applied to your public bus transport.
High power LED development from version 1 to version x doesn’t attract manufacturers to get the newest LED. For example the move from 319A to 319B, or XHP70 to XHP70.2 only gain marginal efficiency (within OEM spec), too small to think about for most luminaire manufacturers. Perhaps the most significant gain is the beam quality, not the efficiency. On the other hand, flashlighting is very similar to premium super car world. Anything faster or better, even those with decimal point advantages are still the better choice.

319B vs GR-V3 (both R70), 3000K, Target OTF output = 3500lm:

Nichia NF2W757GR-V3 R70 (6V)
Condition: 1400mA (116,67mA/LED), CC driver, 3S12P
Output = 3959 lm
Efficacy = 162 lm/watt
Wattage = 24,4 watt
LED# = 36pcs
LED cost* = $4,3
*$0,1176/pc @</=10.000pcs MOQ

Nichia NVSW319BT R70 (3V)
Condition: 1400mA (116,67mA/LED), CC driver, 3S12P
Output = 3914 lm
Efficacy = 179 lm/watt
Wattage = 21,5 watt
LED# = 73pcs
LED cost* = $59,86
*$0,82/pc @</=3500pcs MOQ

To get reasonable cost, we have to change existing driver to match 319B performance. To get at least the same efficacy as GR-V3, thus:
Nichia NVSW319BT R70 (3V)
Condition: 5760mA (480mA/LED), CC driver, 3S12P
Output = 4017 lm
Efficacy = 162 lm/watt
Wattage = 24,8 watt
LED# = 19pcs
LED cost* = $15,58
*$0,82/pc @</=3500pcs MOQ

See? There’s no way high power LED will win over mid power LED in terms of cost or efficacy. And FYI, almost all higher current LED driver are more pricey than their lower current output counterparts.
With such low current and power density, GR-V3 can use any good cheap laminated MCPCB reliably. While 319B with its higher power density is about to enter the DTP zone or at least very good laminated MCPCB
.
[Clemence]

Agro wrote:

The format is not very convenient for me but a dense linear module that I could use instead of a flexible strip would be useful. Smile

Can’t please everybody all at once. But don’t worry, I already made a design for linear 150mm x 15mm 757 MCPCB waiting to send to the factory.

SKV89 wrote:

Would it work in a sofirn sp36?

I have no idea what this is

JaredM wrote:

R95 in 3000K looks like a great value indeed!
I’d be curious to see a tint mixed R95+Optisolis in the low CCTs. There’s a small chance they may compliment one another in the blue/cyan region.

Don’t forget Optisolis higher Vf, can only be mixed in the series string

NWoodsman wrote:

It appears there might be a slim possibility to adapt to automotive 12v applications… it seems not designed for 3v emitters in series…am i wrong?

It can be configured to 3V or 9V with 3V LED. 6V or 18V with 6V LED, 12V or 36V with 12V LED

gchart wrote:

Oh my… Facepalm
Due to the orientation of the image of the MCPCB, I was reading those pad labels as “d” and “d/s”. I thought there was some cruel calculus joke going on (thinking d/dx). Nope… Those are probably “p” and “p/s” for parallel/series, huh?

No, no calculus involved in the design, s = series, p = parallel, s/p = series or parallel

.
[Clemence]

UPDATE 190617:
5700K (18x sm503 rfc00 & 18x sm653 rfa00) Optisolis module (9v configuration) prepared for testing in Finland

[Clemence]