High Output Cree Street Lamp LED Spec Help

[a.baker]

So Ive looked quite extensively around the web for more information on these. They were given to me, aparently the pole they were on had gotten wiped out and these were salvaged. Curious what I would need to do or know to be able to drive them. Not sure if these require a driver which can maintain cc/cv or if that is a (partially) built in function of the mcpcb these are built into and therefore could be powered through maybe a simpler means. Any assistance is greatly appreciated. In the attached pictures you will see some data sheets I believe is on the same style light and I had found this on cree's website, however im not sure they provide much insight. If you need any better images or more details please let me know and Ill try to respond as quickly as possible. Thank You!

[/URL] [/IMG]

 

[lightfooted]

Umm...maybe you shouldn't be trying to do ANYTHING with them if you don't understand the documents you found for them. I'm really not trying to come off like a jerk or whatever but, they are meant for mains voltage...as in plug it directly into the supply you find inside a lamp post. That can kill you if you don't know what you are doing. And well... I'm really not sure that you do.

 

[DIWdiver]

There must have been electronics in the lamp head that are not part of the board you have. There is no way that board attaches directly to mains voltage (the inputs are marked + and -), and there's no driver circuitry.

They would definitely need some sort of driver, but without proper documentation or testing, there's no way of knowing what the voltage and current should be, though you can make a pretty good guess that the total input power should be limited to 100W.

 

[PhotonWrangler]

Deleted to reduce confusion regarding topic.

 

[lightfooted]

There must have been electronics in the lamp head that are not part of the board you have. There is no way that board attaches directly to mains voltage (the inputs are marked + and -), and there's no driver circuitry.

They would definitely need some sort of driver, but without proper documentation or testing, there's no way of knowing what the voltage and current should be, though you can make a pretty good guess that the total input power should be limited to 100W.

I dunno...that seems to be what the Cree documentation is saying = 120V to 277 volts. But really that was my point. There might be a driver board that the red and black wires are connected to, or there may not. But the data from Cree says that they operate on 120 to 277 volts regardless of whether they are done so directly or through a driver board. The LEDs themselves could be seeing from 24-55.5 volts if they were all in series, but the other diodes and the traces on the board suggest that they are not in series. They look to be in parallel which means full voltage, divided current. Each LED is seeing the full mains voltage...even if it is supposed to be coming from a DC converter. Which I am not certain of. Electronic Switch? Yeah sure...that would make sense. That would explain why the wires are black and red instead of black and white. Also...where is the ground?

Did you try using the QR code on the board?

 

[Dave_H]

If you follow the PCB traces, appears LEDs are wired in series. Looking closely at the LEDs, they appear to have
four chips, which may or may not be in series.

The line-powered driver is obviously missing, which would be operating in CC mode. There is no obvious
sign of current limiting on the LED board.

In absence of detailed docs on the board itself, some probing and careful testing with a variable power
supply could answer a lot of the questions. Even a fixed dc supply with appropriate series resistor(s)
could be used. I can give further details.

As suggested, total LED voltage may be approaching 60v dc, however if each LED uses paralleled chips,
total voltage may be as low as 15-16v (at 6-7 amps for full brightness). Testing at reduced current is not
hazardous if done with care.

The small diodes are wired across each LED which suggests that if a LED fails open, the diode will
carry current and the other LEDs will stay on. Probably a small zener with breakdown voltage
not much above the LED forward voltage e.g. 5v for parallel-wired chips per LED. It could also
protect against transients.

Once you find what the total forward voltage is, a proper driver could be identified. I would avoid running
these anywhere near full power without proper heatsinking that the enclosure would provide.

Dave

BTW quick check on LEDs is probing across them with DMM set on low "ohms" or "diode check".
In one direction you should get a dim glow. If not, LED is bad, or several in series which the DMM
can't light up. Might be a bit difficult with these due to clear conformal coating on the LEDs.

 

[Hooked on Fenix]

In my experience doing electrical work, those lights are usually set up for 3 phase 277 volts. When I was wiring them up, two phases were wired to the light wires and the remaining phase was wired to the next light over. (Brown and orange wires to the first, yellow and brown to the second, orange and yellow to the third from two sets of brown, orange, yellow wire). You try to wire them in sets of three lights to balance the load. Fixture gets grounded to the metal head down to the ground wire sticking out of the base column. A fuse is wired to one of the phase wires at the base. Wiring only one light should be interesting.

 

[JustAnOldFashionedLEDGuy]

dave_h is mainly bang on. Not sure what the rest of you are going on about.


They are wired in series, but that little device across each LED is probably an application specific LED shunt protection device and will conduct at low voltage if the LED fails. If it didn't, it would go up in a puff of smoke if it operated at the LEDs forward voltage and current.


The forward voltage is at least 30V for the whole string, could be 60V, but there is no guarantee those LED die are single junction devices. They could be 2, or 4 junction devices (or more), hence the forward voltage per LED package could be 6V, 12V, 24V or higher. You cannot tell that from the pictures. A multi-meter will not have enough voltage on ohms to light these in all likelihood.


As noted, the LED driver is missing (CC) and without knowing the LED properties, there is no way to know to pick a replacement. Not sure the OP has the skills or equipment to find that out. Takes minutes if you know what you are doing and have the equipment.


Do you have the actual fixture or just the LED boards? If just the LED boards I would be inclined to just scrap them. You will spend more money and time than it is worth getting them running.

 

[Hooked on Fenix]

Can't say for sure what the black thing labeled D6 is (possibly the driver) but D7, D8, D9, D10, and D11 are clearly rectifier diodes. I can tell by the silver band on one end. They keep the current flowing in one direction (cuts off the negative part of the voltage sine wave) making ac current to DC current so the power doesn't destroy the l.e.d.s.

 

[Dave_H]

AC is not being supplied directly to the LED board...there is a separate driver module with 100W capability connected in the complete product. The diodes are not there to rectify ac.

Size of most of the shunt diodes except D6 look too small to carry full current if this is a low-voltage (16v) LED configuration, which would require 6A or so.

As a shunt bypass, a normal zener diode would dissipate way too much power for that size of diode, and be very inefficient. The technique works for Christmas tree light strings at low current.

However, these diodes may be low-voltage SIDAC or similar device which when triggered by voltage across them rising, drop to a low voltage, which could carry the LED load current with tolerable power dissipation. Otherwise, diodes may be protection from static (ESD).


Dave

 

[jtr1962]

Can't say for sure what the black thing labeled D6 is (possibly the driver) but D7, D8, D9, D10, and D11 are clearly rectifier diodes. I can tell by the silver band on one end. They keep the current flowing in one direction (cuts off the negative part of the voltage sine wave) making ac current to DC current so the power doesn't destroy the l.e.d.s.

D6 is just another diode to protect the entire LED string from reverse voltage. If the driver is hooked up backwards, D6 will carry the current, thereby limiting reverse voltage to probably under 0.5V.

I'm also going to hazard a guess that each LED is 4 dies in series, so total string voltage is ~60V. It's much easier to design a higher voltage, lower current driver than lower voltage, higher current one. I'm sure that's what was done here. That makes even more sense given the fact the rectified AC voltage is very high, so might as well take advantage of it. The only time it makes sense to go with a high current, low voltage system is when the raw supply voltage is limited, say because it's a battery-operated system.

 

[Dave_H]

D6 is just another diode to protect the entire LED string from reverse voltage. If the driver is hooked up backwards, D6 will carry the current, thereby limiting reverse voltage to probably under 0.5V.

I'm also going to hazard a guess that each LED is 4 dies in series, so total string voltage is ~60V. It's much easier to design a higher voltage, lower current driver than lower voltage, higher current one. I'm sure that's what was done here. That makes even more sense given the fact the rectified AC voltage is very high, so might as well take advantage of it. The only time it makes sense to go with a high current, low voltage system is when the raw supply voltage is limited, say because it's a battery-operated system.

D6 reverse-polarity makes sense and it's not a huge device so the low-current
high-voltage configuration makes sense. At 60v, current would be around 1.6A.
Doubtful D6 could carry 6-7A continuously for a high-current driver connected
in reverse, if it would allow such a low output voltage (almost a short).

Any idea about the smaller shunt diodes? The SIDAC idea is speculation. I
looked at specs for low-voltage part and at 1.6A its clamp voltage would still be
2-3v, which is a lot of power for a small device to handle indefinitely.

To the OP:

Do you have capability to do some simple testing: variable power supply, DMM,
various power resistors (or rely on current-limiting in the supply if available).

Also, you could run on lower current, lower brightness and probably slightly
higher efficacy, if you don't need 10k lumens. BTW you will need good heatsinking
as more than half the power you feed to it will go out as heat (at full current,
roughly 60W, think of an incandescent with that power).

Dave

 

[jtr1962]

Any idea about the smaller shunt diodes? The SIDAC idea is speculation. I
looked at specs for low-voltage part and at 1.6A its clamp voltage would still be
2-3v, which is a lot of power for a small device to handle indefinitely.

They just seem to be reverse protection diodes. Not even strictly necessary unless D6 fails but not a horrible idea given that a 10 cent part could save an LED costing a lot more. Also, I don't think the intention is for any of these diodes to carry 1.6A continuously. The idea is just to briefly protect in case the driver is hooked up backwards. The person connecting the driver will realize the LEDs aren't on, and connect it the other way. The streetlamp certainly isn't going to be installed with the driver connected backwards.

While the SIDAC idea is interesting, fact is once one of the LEDs goes out the street may be lit less than adequately. For a whole host of reasons if one LED fails you would rather the entire string goes out so the defective light gets the immediate attention of the local DOT.

 

[JustAnOldFashionedLEDGuy]

These modules are not 100W, they are 52 watt (rated) at the fixture, so likely 40-45 at the module. 20 LED in series, probably around 0.7A operating current. They will want to keep <60V likely to stay Class2 electrical.

Those are LED shunts in all likelihood, not reverse protection. When they conduct after breakdown they are about 1V at 700mA. It is better to have 80% output and still have output versus a light out almost always. This was a requested feature.

 

[Dave_H]

These modules are not 100W, they are 52 watt (rated) at the fixture, so likely 40-45 at the module. 20 LED in series, probably around 0.7A operating current. They will want to keep <60V likely to stay Class2 electrical.

Those are LED shunts in all likelihood, not reverse protection. When they conduct after breakdown they are about 1V at 700mA. It is better to have 80% output and still have output versus a light out almost always. This was a requested feature.

Interesting, the product power internally delivered is around 100W, so there must be two of these LED PCBs?
100W jives with total flux and efficacy numbers. Only one board is visible in the image, perhaps one more is hiding
under a translucent cover? One board itself looks a bit small to deal with more than 50-60W or so not converted
to photons, coming from 100W input power.

So the small diodes could be:

(1) Reverse polarity protection,
(2) Bypass, or
(3) ESD protection

One Bourns SIDAC device I looked at could probably manage 1v at 700mA (3v max.at 5A) so maybe it's that,
or similar device.

Is the OP planning to do any tests, possibly use this board, or are we mostly speculating?
BTW could OP check the small diodes for number/letter markings?

Dave

 

[JustAnOldFashionedLEDGuy]

Interesting, the product power internally delivered is around 100W, so there must be two of these LED PCBs?
100W jives with total flux and efficacy numbers. Only one board is visible in the image, perhaps one more is hiding
under a translucent cover? One board itself looks a bit small to deal with more than 50-60W or so not converted
to photons, coming from 100W input power.

So the small diodes could be:

(1) Reverse polarity protection,
(2) Bypass, or
(3) ESD protection

One Bourns SIDAC device I looked at could probably manage 1v at 700mA (3v max.at 5A) so maybe it's that,
or similar device.

Is the OP planning to do any tests, possibly use this board, or are we mostly speculating?
BTW could OP check the small diodes for number/letter markings?

Dave

Yes, the 100w lights have two of these boards.

They are almost definitely bypass shunts, LittelFuse, Bourns, etc. It is not a SIDAC if Bourns, it is LSP, though similar.

 

[Dave_H]

Here is an interesting technical note from Bourns on their LSP devices:

https://partnership.bourns.com/bu/bu_ledsp.shtml

Dave