cxa3070s@350mA orientation

Scotch089 said:

Also, I messed up with the potentiometers- the c700A's have the screw (built in pots) and no wires coming out (looked at the data get for the wrong drivers (didnt type in HLG-185H, just 185) so I decided to mount them on the front of the panel for easy access to the dimmers.

Click to expand...

I guess that answers your pot mounting question.

How low can you dim with the internal dimmer? I think only 50% with the A version but haven't had confirmation yet.

Just for anyone else who reads this looking for the external dimmer, its the B version of the driver. The B version can be dimmed with PWM (automation) and down below 10% for the same price.

Examples....

HLG-185H-C1400B
HLG-185H-C700B
HLG-120H-C1050B
HLG-60H-C350B
HLN-60H-42B

Their are much smaller taps Than 4-40. A LOL SMALLER.

any way wd40 us perfect for tapping alum. You want a light weight oil... kerosene is very good as well.


for a quick and easy tap. Have two drills one with a drill the other with a tap.

drill the hole pick up the tap drill and run the tap threw on the slowest setting. If you feel it bind or get tight back it out a few turns and try again.

I personal like using roll taps instead of cut taps but that's just me.

For our purposes a hand tap should be plenty.

Not sure why you would want to use a drill to tap. The whole trick to not breaking taps in your heatsink is being able to feel it binding.

Cutting threads is a delicate process. Save the drill for the drill bit.

That looks like the hand-chuck I used (in a kit I got).

I did see a youtube, however, that showed if you have a drill with a slip clutch and variable speed trigger, it's easy. Just set the clutch to the lowest setting and tap forward for a second, backward for a half second until done. Didn't see a point in trying since I only had to tap 16 M3 holes.

I have a dewalt drill like this, and getting another drill (for driving lag screws into pre-drilled holes) is something I've been wanting to do for years. It would make building furniture so much easier for me. Drill, screw, drill, screw, etc. There are so many cases where waiting to drill all the holes before putting in the lag screws can be a pain in the ass. You want to see if your holes line up early, after the second hole, not after the last hole.

Positivity said:

For our purposes a hand tap should be plenty.

Not sure why you would want to use a drill to tap. The whole trick to not breaking taps in your heatsink is being able to feel it binding.

Cutting threads is a delicate process. Save the drill for the drill bit.

View attachment 3207588

Click to expand...

Because I tap holes 5x faster with a drill. I also have 20+ years machining so I can feel and hear when it's to much bite. I never tap by hand unless it's under a 4-40.

True that...I was just thinking some people are really skilled with their drill.

I've been using 1/4" screws so the hand chuck one works great for me

I drilled all mine. Broke a couple bits...only because I didn't realize how shallow the holes actually were and was binding it against the bottom, and din't reduce my torque. Then I bought a little hand chuck but it woulnt grab my 4-40 tap.So back tho the drill. Simple sharpy mark for hole depth so I knew where I was, and all smooth from there. So if anyone wants to use a drill, I think it's fine, just don't get super ripped right before...or do, your choice.

The dimmer on the "A" models is a setting for current limiting, which is actually the most inefficient part of these power supplies. (running in current limiting mode)

For example, lets say you have a 143V supply that goes up to 1.4A, but the diode string is only pulling 1A at 143V because of the Vf of needed for 1.4A is higher than 143V.

In this case, with the pot at full setting, the power supply will not limit current. It naturally flows at 1A because of the characteristic curve of the diode and how Vf correlated to current.

As you turn the pot down, the leds will not dim until the limit current goes below the natural current of 1A. Further turning will finally cause it to start limiting current, and thus dim. This is inefficient because the limiting transistors are running in linear region (i'm pretty sure the current limiting happens in linear mode and that it's just the voltage that's maintained with switch mode. It's a voltage supply with current limiting). A lot of heat is lost through these transistors when limiting current.

I actually bought an "A" model with the intention of setting it right above where it would be limiting. (If I want to run at 1.0A, I would turn it to 1.1A limit). There's really no reason to ever touch it, and thus it makes sense as a trim pot on the supply itself.

epicfail said:

I guess that answers your pot mounting question.

How low can you dim with the internal dimmer? I think only 50% with the A version but haven't had confirmation yet.

Just for anyone else who reads this looking for the external dimmer, its the B version of the driver. The B version can be dimmed with PWM (automation) and down below 10% for the same price.

Examples....

HLG-185H-C1400B
HLG-185H-C700B
HLG-120H-C1050B
HLG-60H-C350B
HLN-60H-42B

Click to expand...

churchhaze said:

The dimmer on the "A" models is a setting for current limiting, which is actually the most inefficient part of these power supplies. (running in current limiting mode)

For example, lets say you have a 143V supply that goes up to 1.4A, but the diode string is only pulling 1A at 143V because of the Vf of needed for 1.4A is higher than 143V.

In this case, with the pot at full setting, the power supply will not limit current. It naturally flows at 1A because of the characteristic curve of the diode and how Vf correlated to current.

As you turn the pot down, the leds will not dim until the limit current goes below the natural current of 1A. Further turning will finally cause it to start limiting current, and thus dim. This is inefficient because the limiting transistors are running in linear region (i'm pretty sure the current limiting happens in linear mode and that it's just the voltage that's maintained with switch mode. It's a voltage supply with current limiting). A lot of heat is lost through these transistors when limiting current.

I actually bought an "A" model with the intention of setting it right above where it would be limiting. (If I want to run at 1.0A, I would turn it to 1.1A limit). There's really no reason to ever touch it, and thus it makes sense as a trim pot on the supply itself.

Click to expand...

You are running drivers maxing out the voltage and as such the string can only pull 1A. I run them maxing out the current having room to spare on the voltage so I get the full 1.4 amps. When I turn my pot (100k resistor or PWM 0-10v) I notice the dim right away and so does my meter, I can fade smoothly from 1.4A to 0.14A+/- and automate with a led controller or arduino type device.

Your way works perfect if you want to set it and forget it, lets say drive 4 cxa3070 off 1 hlg-185h-C1400* (4x37.5=150V). The driver couldn't run them at 1400mA like you say the Vf is to high so you have to limit it and leave it alone. I would only run 3 cxa3070 on that driver so I can automate a fade from 5-10% cold start to 100% over 20-30 min in the morning and fade out at night.

Each works perfectly for how we want to use. With the "B" version you can add a (Dynamic) light controller later if you ever wanted and you can still use it the way your using it (Static). I left a pair HLN-60H-42B running a cxa3070 each, one driver steady at 1.4A (full) and the other fading from 10-100% up and down over 30 second in a loop. After 5-7 hours they were both cool to the touch so I don't think I should be concerned about transistor heat loss to much. I do understand that the drivers lose efficiency dramatically when under <50% load and I wont be running mine like that, only fading.

If anyone plans running static at 350mA a better driver match would be something like the HLG-60H-C350* (5 3070's), HLG-80H-C350* (7 3070's) or HLG-120H-C350* (12 3070's) so you can keep the load close to full therefore keeping the driver efficiency high (>90%) and not limit the current. Even the HLG-185h-c500* is better suited because when limited to 350mA is still 75% load and runs at >90% efficiency and you can give them the a boost to 500mA when needed. you can run 11 on the HLG-185H-C500*, 8 on the HLG-120H-C500*

I still haven committed to buying drivers for 8 of my COBs because I'm not sure how I want to run them yet. I thought of running 4 cxa3070 off 1 HLG-185H-C1400* but to get the Vf to drop below the 143v hard cap (143 / 4 = 35.5Vf per cob) they had to be run below 1A so the HLG-185H-C1050* is better suited. I want a 84V version to drive 2, the 120H cant fully utilize 3 and you run into the same scenario as the 1400mA version. Either have to run 2 at a lower load reducing efficiency or reduce the current so much its better to use the 1050mA version.

Awesome driver discussion here guys. Here are the results from an HLG-185H-C1050A

Current range:.48A-1.15A
Vf range: 34-190.9 Vf
When I go above 190.9 Vf the driver decreases the current very steeply. On the low side, I have not tried to go below 34Vf (1 COB) so 34 Vf may not be the absolute minimum but irrelevant because efficiency suffers with a low Vf (72-78%)

(5) CXA3070s + (3) XML2s, driver warmed up:
@1155mA = 93.6% (190.9 Vf maxed out, current maxed out)
@1050mA = 93.5% (187.9 Vf)
@700mA, = 93.5% (180.9 Vf)
@ 525mA = 93% (176.9 Vf)
@ 480mA = 92.5% (176 vF, current set to minimum)

(3) CXA3070s, driver warmed up:
@1155mA = 88.5% (100.5 Vf)
@480mA = 90.9% (108.5 Vf)

So the key to maximizing efficiency with these drivers is to keep the vF as high as possible without going over the cap. Now I wish I had checked the efficiency at 480mA with 191Vf.

I thought the efficiency would have suffered more, I guess 2 CXA30** on the HLG-120H-C1400B would run above 90%. I should just order some already, my heat sinks will be here this week. I need to get a new power monitor (last one melted) so I can check the efficiency myself. My main problem is the drivers I want (HLN-60H-42B) are on 7 weeks back order.

Anyone use one of these before?

http://www.amazon.com/ALEKO®-Electricity-Monitor-Reduce-Energy/dp/B00BGFRBTQ

I have a few power meters that look exactly like that but they are branded ensupra (no affiliation). They measure down to tenths of a watt which is very handy for smaller devices.

Id like to try this one, cheap and from the reviews seems to be able to measure down to 0.1W
http://www.amazon.com/Weanas®-Energy-Voltage-Electricity-Monitor/dp/B00DTMQ1S6/ref=pd_sbs_lg_5?ie=UTF8&refRID=06ET7G92Z518M9ZV2D6R

churchhaze said:

(i'm pretty sure the current limiting happens in linear mode and that it's just the voltage that's maintained with switch mode. It's a voltage supply with current limiting). A lot of heat is lost through these transistors when limiting current.

Click to expand...

The block diagram suggests otherwise. The power is throttled by the switching action. The intelligence is happening in the "PWM & PFC control" block. When dimmed it is still pretty efficient.

MrFlux said:

The block diagram suggests otherwise. The power is throttled by the switching action. The intelligence is happening in the "PWM & PFC control" block. When dimmed it is still pretty efficient.
View attachment 3208139

Click to expand...

I never really got past Ohm's law and Thevenin's theorem but isn't the switching why the drivers make a "squeal" when dimmed down? e.g. audible harmonics in the magnetics?

also, Supra's "no affiliation" joke never gets old

Observe & Report said:

I never really got past Ohm's law and Thevenin's theorem but isn't the switching why the drivers make a "squeal" when dimmed down? e.g. audible harmonics in the magnetics?

also, Supra's "no affiliation" joke never gets old

Click to expand...

Yes ...The audible "Squeal" is actually from the wire coil wraps *,that no matter how tight they are (and / or glued/epoxied/waxed ),they still 'move' / 'slide' in a perpendicular direction to the wrapping .

*( on inductors & transformers )
-----------------------------------------------------------------------------------------------------------------------
Remember the 'FBI ' / 'Dynamo' rule of thump ,where F=Force, B=Magnetic Field & I=current


The phenomenon is audible due to the 'switching harmonics' that somewhat fall inside the audible range of 20 Hz - 20 kHz .
The low duty cycle (dimmed down ,close to 100% ) just exacerbates the ' squealing ' ,
of the wire coil wraps ...


More info about PSUs :

http://www.adminstrumentengineering.com.au/media/pdfs/Catalogues/Manuals/Mean-Well-USER-Technical-Manual.pdf

http://www.xppower.com/pdfs/techguide.pdf

http://users.tpg.com.au/users/p8king/files/mwmanual.pdf

LED CC PSU :

http://www.powergatellc.com/pdfs/ledmanual.pdf

Alright. So reflowing is not my forté, got x2 of the XPEs to light up. The rest are bunk (on my part) dono what I did different to the others but they won't work. After a few hours of frustration I hit that "fuck it!" Level and decided to go back to my luxeons for now and use the XPEs I had going. Works out because I was thinking about mixing them anyways. But still a bummer. That said- it is DONE. (Never really done... Is it?)

Not very much red accent to the 3070s but noticeable blue from the xml2s (and two on the left are shot- the two I couldn't figure out wtf) so I may be firing an email or placing another order for a few.

Some sweet ass #s 12 and 18" away, but I am alarmed by the max draw (about 40 watts more than i figured it'd be, 272w) but dims down right to 119w (with fans [+8w]) anyone care to lay some insight of what I could have done wrong? Kindve worried of overpowering the smaller emitters.

picture #1 is for a good spectrum shot, #2 is just cool looking.. #3 is my 12" reading in the center, #4 is my 18" reading in the center, #5- edge reading at 12", #6 is my center reading dimmed down to minimum draw at 12", #7 is at 18"

I will say the spectrum shot gives it a bit more red hue than what appears to the eye, and the right side has more blue from the emitters being out.

Let me know about thy draw if anyone has some insight! Take care!

Edit: I am going to go back through and secure all the stars down though.

Scotch ,I just can't believe that reflowing was not easy to you ..
Great light ,there !

I know!! I had no issues gettin them mounted/soldered in place- but they would just beep at me when I tested them. Made sure the anode/cathode was on the right side and all. Used a skillet on my stove to do it. (Couldn't get em to heat up enough with the solder gun like the luxeons) killed me... REALLY wanna be able to use those XPEs.

sexy looking light brother scotch. Truly a great inspiration.