Your CXA3590 Mission Should you Choose to Accept It !

Kuifje76

Well-Known Member
They only go up to 32 volts on the output and so they wouldn't work for the cobs we are using. The AC input models require low AC voltage 7-24 volts. So not something your average DIY'er is going to want to deal with. They do have a boost puck that goes up to 48 volts but you have to power it with a DC power supply. Once you go that route it make is more inefficient since most DC power supplies are are in the 80%'s efficiency. So you are much better off getting a Driver that has 120 volt AC input and high efficiency.
I thought these would work too, indeed matched with an AC-DC power supply (does it need to be CV or CC ?) There even is a board to wire them all up.
If so, if we could find 36V buck pucks (for the cxa3590 36V version), is it more efficient then those puck above ? I found some here (spec. sheet); I find this also very interesting because they're so small and you could build it into any heatsink

Can't we use an efficient power supply like HLG-185H-48B to connect to those pucks ?
How does efficienty works with AC-DC & DC-DC, is it also multiply ? So tot. eff. = eff. cob x eff. power supply x eff. DC-driver ?

Sorry, if i'm rambling again, let's just say stoned again :bigjoint:
 
Last edited:

Kuifje76

Well-Known Member
Well put Java.

You can dim to almost zero with the Mean Wells that use external dimmers (B versions), although I recall someone warned not to take it below 10% for driver safety? I can verify my HLG-185H-C1050B dimmed a pair of CXA3070s down as low as 20 MA smoothly.
I thought for example the HLG-185H-C1050B only dim's to 50% ? 525mA to 1050mA ? What am i missing here ?
Fantastic news btw :razz:
 
Last edited:

Kuifje76

Well-Known Member
Thanks for the link Erlking.

If you can dim that low with the meanwells...
Ok, so again the HLG-185H-C1050B seems most wanted for cxa3070 and cxa3590 (36V) if you want to drive 5 cobs 8)
and HLG-120H-C1050B seems most wanted if you want to drive 4 cobs 8)
and running them lower all 9 together becomes more interesting :D

if you don't mind high voltage...
 
Last edited:

JavaCo

Well-Known Member
When you talk about buck puck , you mean this things like this ?
http://www.taskled.com/techhboost.shtml

Becuz the thing can outp 80V @ 1.5A it seems
Well you still have to power it with a AC to DC PSU. It is a boost module so you will need a high amp PSU. Unless you have solar panels and a battery bank then you could go straight to that module. Most folks dont have that option. So you have to factor in both the PSU and the led Driver as far as efficiency. But yes those Drivers would work but you could end up in the high 70% on efficiency when you add both together. Here is the buck puck i was talking about in my previous post http://www.luxdrive.com/products/buckpuck-3021-3023-led-driver/

I thought these would work too, indeed matched with an AC-DC power supply (does it need to be CV or CC ?) There even is a board to wire them all up.
If so, if we could find 36V buck pucks (for the cxa3590 36V version), is it more efficient then those puck above ? I found some here (spec. sheet); I find this also very interesting because they're so small and you could build it into any heatsink

Can't we use an efficient power supply like HLG-185H-48B to connect to those pucks ?
How does efficienty works with AC-DC & DC-DC, is it also multiply ? So tot. eff. = eff. cob x eff. power supply x eff. DC-driver ?

Sorry, if i'm rambling again, let's just say stoned again :bigjoint:
36 volts is cutting it close these cobs usually run a couple volts higher till they worm up so you might have a start up problem. Plus the 36 volt CXA 3590 runs @ 2.4 amps so running it @ 1050ma would be around 30 watts.
No you cant power a puck with a constant current driver , Needs to be a constant Voltage power source.

I thought for example the HLG-185H-C1050B only dim's to 50% ? 525mA to 1050mA ? What am i missing here ?
Fantastic news btw :razz:
It will dim lower than the that. You looking at the numbers for the A version of the driver with the built in pot. The B version can go all the way dark if set up correctly. But really you dont want to dim it past 20% since efficiency takes a dive.
 
Last edited:

Kuifje76

Well-Known Member
Well you still have to power it with a AC to DC PSU. It is a boost module so you will need a high amp PSU. Unless you have solar panels and a battery bank then you could go straight to that module. Most folks dont have that option. So you have to factor in both the PSU and the led Driver as far as efficiency. But yes those Drivers would work but you could end up in the high 70% on efficiency when you add both together. Here is the buck puck i was talking about in my previous post http://www.luxdrive.com/products/buckpuck-3021-3023-led-driver/


36 volts is cutting it close these cobs usually run a couple volts higher till they worm up so you might have a start up problem. Plus the 36 volt CXA 3590 runs @ 2.4 amps so running it @ 1050ma would be around 30 watts.
No you cant power a puck with a constant current driver , Needs to be a constant Voltage power source.


It will dim lower than the that. You looking at the numbers for the A version of the driver with the built in pot. The B version can go all the way dark if set up correctly. But really you dont want to dim it past 20% since efficiency takes a dive.
I know it would be an expensive lamp with cxa3590 at 30W/piece... and those pucks add up to the cost, I'm just looking at the most flexible and efficient (also silent) setup, but i guess we don't need those pucks...
those meanwell drivers will do 8-)

imo cxa3590 will eventually take the place of cxa3070 when things get cheaper

But an ultra efficient 2x2 foot coverage lamp with 9 cobs mixing spectrums better (ex 5x2700K and 4x5000K) and dim them preciesly or mimick sunrise-day-sunset etc. sounds tempting...
In a few years goal will be 3g/W :blsmoke:.
 
Last edited:

SupraSPL

Well-Known Member
I am with you on that. The mythical CXA3590 3000K CB bin could be a game changer if we could get them for a good price. At 270mA (19W) they are 58% efficient (typical ~ 35C) and still cost less per PAR W than I recently paid for my deep reds.

Unfortunately CREE seems to be getting its butt kicked in the market currently so who knows how that might affect the CXA3590 production.
 

Kuifje76

Well-Known Member
I am with you on that. The mythical CXA3590 3000K CB bin could be a game changer if we could get them for a good price. At 270mA (19W) they are 58% efficient (typical ~ 35C) and still cost less per PAR W than I recently paid for my deep reds.

Unfortunately CREE seems to be getting its butt kicked in the market currently so who knows how that might affect the CXA3590 production.
How so ?
 

dimebagor

Well-Known Member
He is wrong the LPF-90-48 will not put out 2250mA . It might put out 50 mA more then the rated 1880mA but thats about it. Voltage has nothing to do with the current output. These are Constant Current Drivers. So A lpf-90-48 Will put out roughly 1880 mA between 28.8 to 48 volts.

I agree with Observe on the LPF-90-42 it is cutting it really close on the upper voltage. You could have start up problems when cold. The Vero 29 data sheet has 42.5 volts as a maximum Vf @ 25 C and 43.5 Vf @ -40 C. But they Also list 38.6 Vf as the typical forward voltage for a Vero 29 @ 25c . So looking at the data sheet i would say most Vero 29's would work with the LPF-90-42 @ room temperature, But i also would not be surprised if a few didn't, especially if it is cold. It is right there on the edge.
For the vero29, this meanwell looks ok right ?
http://fr.onlinecomponents.com/mean-well-cen10036.html?p=37809805&ref=octopartfeed

you learned me something about the temperature at start up^?
So if its cold , the vero needs more voltage than 36V ok ,
but this driver first, has an Voltage adj range 33-40V
Second, the protection over voltage is 41-46V .
So at cold start , the driver can provide the necessary voltage for the led right even if its the 36V output series

At worst there is the cen-100-42 , with this one its sure , no problem when its cold right ?
But this one , is more difficult to find at good price , as well as that :)
 

SupraSPL

Well-Known Member
That is a good question DB. I have a similar one on order to see if I can get it to work and how much current I can get out of it.
 

nogod_

Well-Known Member
The cxa3070 ab bin (also cree) is actually more efficient. But only by a few lumens per watt at the same power consumption. Other than that nothing can compete (yet).

Cree is getting walloped in the markets but so is everything else except my beloved gwpharma.

Run pharma run.

oh, and who is doing well that also produces great chips ? Is there any competition against the cxa3590 in terms of efficienty ?
 

JavaCo

Well-Known Member
For the vero29, this meanwell looks ok right ?
http://fr.onlinecomponents.com/mean-well-cen10036.html?p=37809805&ref=octopartfeed

you learned me something about the temperature at start up^?
So if its cold , the vero needs more voltage than 36V ok ,
but this driver first, has an Voltage adj range 33-40V
Second, the protection over voltage is 41-46V .
So at cold start , the driver can provide the necessary voltage for the led right even if its the 36V output series

At worst there is the cen-100-42 , with this one its sure , no problem when its cold right ?
But this one , is more difficult to find at good price , as well as that :)
The Vero 29 is more like 38 volts driven at 2.1 amps when warm. Here is the chart from the data sheet.

I would use this driver if your wanting to run them @ 2.1 amps http://www.jameco.com/webapp/wcs/stores/servlet/ProductDisplay?langId=-1&storeId=10001&catalogId=10001&pa=2214977&productId=2214977&CID=octopart the CEN 100-42 would work also but it would be 2.28 amps at full power. You would more then likey have a start up problem running the CEN-100-36
 

epicfail

Well-Known Member
Yep O&R nailed it. I am a absolute fool for doing this, but I work on live wires all the time, to save time. I am much more careful working on live 120V AC wires than live 12V DC wires, even though car batteries are capable of delivering massive amounts of current. I have been shocked dozens of times by 120V, but only once from 12V, for example.
"Do as I say not as I do"

I am notorious for working with live electricity and have zapped myself numerous times throughout my life. Recently I was working with a live fasttech open style driver and it blew up in my hand, sparks flew everywhere freaking out my buddy. I have been conscious to make sure my HLG drivers are unplugged when working with them and have decided that all further builds I do will be done with low voltage parallel strings.

Instead of using the HLG-185H-C1050B to drive 4 CXA at 1050ma running close to 150v I should have gotten the HLG-185H-42B (21-42v 4.4a) and driven 4 in parallel at 1100ma as it would only be around 37v. Thermal runaway could be prevented with a pretty simple circuit


http://www.ledsmagazine.com/articles/print/volume-6/issue-2/features/led-design-forum-avoiding-thermal-runaway-when-driving-multiple-led-strings-magazine.html
 

Kuifje76

Well-Known Member
"Do as I say not as I do"

I am notorious for working with live electricity and have zapped myself numerous times throughout my life. Recently I was working with a live fasttech open style driver and it blew up in my hand, sparks flew everywhere freaking out my buddy. I have been conscious to make sure my HLG drivers are unplugged when working with them and have decided that all further builds I do will be done with low voltage parallel strings.

Instead of using the HLG-185H-C1050B to drive 4 CXA at 1050ma running close to 150v I should have gotten the HLG-185H-42B (21-42v 4.4a) and driven 4 in parallel at 1100ma as it would only be around 37v. Thermal runaway could be prevented with a pretty simple circuit


http://www.ledsmagazine.com/articles/print/volume-6/issue-2/features/led-design-forum-avoiding-thermal-runaway-when-driving-multiple-led-strings-magazine.html
Really like the idea of low voltage and high efficienty, but this is chinees for me :oops:, is this difficult for someone with a low level electricity degree ? i think i know someone.
I've looked at the link, but it still remains chinees, if it's to difficult i'm gonna take my chances (and precautions) with the high voltage...
 
Top