DIY LEDs - The write-up

[brassmonkyballs]

BMB,


Not if you want them to keep running. I thought RB were limited to 1000ma by design.

Sounds like a lot less PAR than is recommended with MH, which is what I was beginning to think.

I love the bubble coral, please let me know when it needs to be fragged.

That's where I disagree on the rating. They are indeed rated for 1000 ma.... constant 1000ma. That is nowhere near a failure point and the failure comes with junction temp which if kept down is not an issue. This is how I have been reading the specs. Though I now subscribe to running much lower if for no other reason than to keep nice corals colorful

I also can say that my very 1st fixture I ever built ran XREs at 1.3 A for over a year without any failures. This was my test fixture which I used to determine if LEDs work and how best to care for my corals under this lighting. That 14g cube is still going strong at a friends house . Though it's hard to keep acros in it without bleaching.

Yes my bubble coral..... It actually closed up for 2 months after I got it about 3 years ago. When it came back it split itself into 2 colonies. It's a double bubble! And I know of no good way to frag these things. It's a 'bute

My PAR levels I'm running at are interesting. I have search without luck but I would like to know what the PAR level are in the ocean at the reef to get some correlation.

I forgot to mention I run the RBs for 11 hours and the white, green, blues for 6 hours.

I have an ORA Blue Voodoo that I am not happy with the color. It's at the top of the tank where the 250 is. Since I changed and added the greens and reds and dimmed it down 2 weeks ago I am getting some blue back on the tips. We'll let it go for awhile and if it doesnt color up properly I'll move it down. Everything else is pretty happy though.

 

[brassmonkyballs]

Im not questioning you just trying to learn from your experience, I am having some color issues and was suprised by your findings. I may dial things way back. I am mixing T5 and LEDs
I have 60 RB and 20 NW Cree leds, the 60 run at a max of 850 ma and the NW run at 1amp.
This is over a 180 and I think with all the lights on I get 280 ish on the top and 80 ish on the bottom. With the t5's and the LEDs on I get 800 on the top and
250ish on the bottom but my colors are a little pale.

I would try to dial the whites way back and shorten the time they are in. Monitor for month and continue tweaking. In my experimenting I've come to the conclusion that color mix with LEDs is the key to coloring up. Too much white and you lose the color. That just my opinion based on experiments. I've bleached an ORA tri-color just using 50-50 CW-RB with the white dimmed 50% and the coral 14" away. When I switched 2-1 RB with the whites still dimmed the color came back in a month.

 

[TheFishMan65]

When I switched 2-1 RB with the whites still dimmed the color came back in a month.

So you increase the light by adding more blue. Or did they get dimmed from the original of 1:1?

 

[SWF]

Hi FishMan65,

Thank you so much for your response.
I do not have a link to the LED's that i am considering... I will try to find it.
Its a DIY fixture made by a person whose getting out of the hobby and is basically selling it.
its a 84 LED fixture which fully automated and has dimming features.
its gonna cost close to $1300. Would it make sense to go ahead and get it.

Only reason i am not doing it myself is .. am not that comfortable with electronics.

Thanks again,
SWF

 

[brassmonkyballs]

So you increase the light by adding more blue. Or did they get dimmed from the original of 1:1?

No I replaced half the whites with RB

 

[der_wille_zur_macht]

SWF, we can't tell you if it is worth it without more info. Hopefully, make, model, and bin info for the LEDs and relevant info for the drivers and other hardware.

Otherwise it is like asking if a used car you are looking at is worth $20k and not telling us if it is a Kia or a Benz.

 

[SWF]

Hi DWZM,

I get your point... i'll try to collect those information.

Thanks again for responding.

Regards,
SWF

 

[TheFishMan65]

Another question you need to ask yourself. Exactly how comfortable with electronics are you? You may need to replace an LED, replace a driver, or some other work. While not a full build you still need some skill (assuming the seller is not around). Also consider his dimming controls. Is this also a DIY that may need some support? Get the code if it is a microcontroller of some sort.

That is about $15 per LED. That sound a bit heavy to me.

 

[jkopp36]

Didn't think we'd let this thread go a whole day without a post did you?

Ok, I'm stumped. I received my new HLG-240-48B yesterday and wired it all up. It is doing the exact same thing on the dimming circuit as the original one. I need some troubleshooting input. Here is a quick recap of the issue:

I've tried dimming via 1.5v-10v (1.5 is as low as my bench supply will go) and 100k pot. The setup will be 0-10v via Apex but I "played" with the pot just to see what happens. With the pot there is a very narrow band of adjustment - like literally less than a couple degrees of rotation of the shaft - not even a possibility for dimming really.

Here are some better results using the 1.5-10v supply (from memory but VERY close):

1.5v = 940mA
2.0v = 900mA
3.0v to about 6.0v = 3-4mA (LEDs are visible glowing though)
from there the current barely goes up with each volt:
7v=6mA
8v=7mA
9v=8mA
10v-9mA

Notice the 940mA at 1.5v and 900mA at 2.0v? The odd thing about that is that at 100% the driver should only be around 740mA per string. BTW - All the above numbers are for the 54v version - the new one I got yesterday is a 48v version per FishMan suggestion since total vF is 46.8v. The thinking was that I could get a little more current per string if needed. The 48v tested out almost identically with the only difference being the currents were a little higher.

If I run the drivers with the dimming circuit open the driver is right at rated current with all the strings adding up to 4.47A on the 54v and 5.03A on the 48v.

I've only got a couple ideas for troubleshooting right now:

1) Re-rig the wiring on the XPG CWs (36) to three strings of 12 instead of 6 strings of 6. Test the HLG-240-48B on them. Since the HLG-185 running them is acting perfectly, this should rule out any trouble with the strings - right?

2) Test each string of 12 XPE RBs with a ELN. If they operate and dim correctly this should also rule out string trouble.

Any more ideas? Sorry for the long post but I want to make sure I'm covering all bases (or at least trying).

 

[TheFishMan65]

Can you measure the total current coming out rather than just one string? On the off chance the on circuit or a short is hogging the current.

The lower voltage should have the lower current so that makes it really confusing. 1 volt should be about .5 amps and 2 volts about 1 amp. Have you double checked the voltages for the dimming circuit with the meter.

Also can you get the string voltage for each of the 1 volt steps.

 

[jkopp36]

Can you measure the total current coming out rather than just one string? On the off chance the on circuit or a short is hogging the current.

Sure can. What's the best way to do that? Meter inline between driver and resistor block?

The lower voltage should have the lower current so that makes it really confusing. 1 volt should be about .5 amps and 2 volts about 1 amp. Have you double checked the voltages for the dimming circuit with the meter.

Not sure I follow but the total voltage was 46.8 for both drivers.

Also can you get the string voltage for each of the 1 volt steps.

Yep, you mean total voltage right?

 

[TheFishMan65]

Sure between the driver and block is great.

No, was the bench supply really supplying 1.5 - 10 volts to the driver? Probably, but it is easy to check.

Yes, hopefully the string voltage is increasing as you increase from 1.5 to 10 volts. If not well it might be a clue.

 

[emora]

The HLG-240-x-B are dimmable by three methods: 0-10v, PWM, and resistance.

Try connecting one of your pots directly to the dim wires.

If the dimming range on your potentiometer is too narrow (few degrees of adjustment) have a look at the resistance in that range and find a pot with that resistance value.

 

[jkopp36]

Sure between the driver and block is great.

No, was the bench supply really supplying 1.5 - 10 volts to the driver? Probably, but it is easy to check.

Yes, hopefully the string voltage is increasing as you increase from 1.5 to 10 volts. If not well it might be a clue.

Ok, I follow now - duh. Yes, I watched the voltage on the dim circuit like a hawk to be certain.

I'll get more and exact mearsurements if I do not turn something up testing other drivers out on the strings.

Thanks for all your input!

 

[jkopp36]

The HLG-240-x-B are dimmable by three methods: 0-10v, PWM, and resistance.

Try connecting one of your pots directly to the dim wires.

If the dimming range on your potentiometer is too narrow (few degrees of adjustment) have a look at the resistance in that range and find a pot with that resistance value.

I've tried a 5k, 10k, and 100k pots. All are so sensitive between full dim and full on (barely have to rotate shaft) that they are unuseable. Besides, being able to dim via my Apex was a fairly important positive to converting to LEDs.

 

[jgwinner]

I've tried a 5k, 10k, and 100k pots. All are so sensitive between full dim and full on (barely have to rotate shaft) that they are unuseable. Besides, being able to dim via my Apex was a fairly important positive to converting to LEDs.

How exactly is it hooked up?

== John ==​

 

[jkopp36]

How exactly is it hooked up?

== John ==​

One wire on center (wiper) and another wire on one side. blue and white dim wires from driver hooked to those.

 

[jgwinner]

That's probably why it's so sensitive. I'm at work, I don't have the DS in front of me, but if you can set it up as a voltage divider it would be much bettter.

If you can, run a 10v line from the driver to the high side of the pot, low to the other side, and the wiper to the DIM line. If the current sourced by the DIM line is high you might get some arcing longer term at the low end but that shouldn't be an issue I'd think.

== John ==

 

[jkopp36]

That's probably why it's so sensitive. I'm at work, I don't have the DS in front of me, but if you can set it up as a voltage divider it would be much bettter.

If you can, run a 10v line from the driver to the high side of the pot, low to the other side, and the wiper to the DIM line. If the current sourced by the DIM line is high you might get some arcing longer term at the low end but that shouldn't be an issue I'd think.

== John ==

Ok, back up here just a bit, we may be talking about two different things. I am testing the drivers using both the 0v-10v method of dimming allowed by them and also the pot only method where the pot is just a simple rheostat. There is no voltage involved when using just the pot. For the 0v-10v method I am using a regulated bench supply which is capable of 1.5v-12v DC.

 

[jgwinner]

I am testing the drivers using both the 0v-10v method of dimming allowed by them and also the pot only method where the pot is just a simple rheostat. There is no voltage involved when using just the pot.

As I mentioned, I didn't have the data sheet handy. If the pot was attempting to act as a voltage dim, that would explain the problem, but if the wires can do resistance only dimming then it should work. Not sure what's wrong then.

I checked the data sheet and it says use a 100k resistance. It also says don't connect the DIM- and V- lines, so don't follow my advice! :eek2:

== John ==​