Quad P7 custom thrower (edit outdoor video)

[IMSabbel]

Well, i finally got all the main pieces for my 4xP7 today, and already got first light.

Pictures will follow tomorrow (its already late at night).

I just want to share that i was surprised about the low Vf of those P7. (got the star ones drom DX)

4 of them in row (including 8 solder points, plus two crocodile clasps) reached 2.8 A at 13.5V and 3.5A at 14.0V.

That killed all my direct drive intents (because 4 18650 in row will KILL them, quickly. Especially hot of the charger). Good for my i already have a linear constant current driver for 2.8A around here.
Maybe direct drive with A123 cells would work. They should be a pretty good match with 3.3V.

 

[shakeylegs]

Re: Quad P7 custom thrower

What host do you have in mind and what reflectors have you decided to use?

 

[IMSabbel]

Re: Quad P7 custom thrower

What host do you have in mind and what reflectors have you decided to use?

I did use the 52mm reflectors from DX, so its only moderately throwy (but still better than the typical p7 light). As for the host... I still have an old handheld thrower in the basement that originally used 6V block batteries. I think i will dremel it to the right size (the reflector is ok).
Especially i am not quite fixed yet on the heatsinking. I have a huge-*** heatpipe cooler that should be able to take anything those things throw at it, for an indefinite lenght.
But on the other hand, a small HSF might be a better solution (the light isnt going to be very watertight anyways).

And first tries of the light including the reflectors are done now (the epoxy has cured enough).
Damn bright. Even indoors at daytime with sunshine outside.

 

[IMSabbel]

Re: Quad P7 custom thrower (edit first pics)

One thing right in front: I am no expert, and this wont be anything special.

The piece of aluminium i managed to get. I would have liked it a bit bigger, the 80mm are a tight fit with the safty margin for the positions.

Printout of the design. Lots of little M2 screws need holes

Well, that actually worked. I didnt have a P7 star when i made the layout, so i took a picture from DX and constructed it after it. Seems i didnt mix anything up. Everthing lines up nicely

Ok, after drilling and re-doing the finish... strangly, it _looks_ worse after etching and polish. But i dont need a mirror finish.
The big holes are for mounting a socket 775 cpu cooler on the backside. The contact surface of such a cooler is the square in the middle of the plan)

Ok, leds soldered and screwed in place, with PLENTY of pressure on those screws, so its a really thin layer of thermal grease. I dodged a real bullet as i remembered just in time that soldering _after_ screwing would be stupid... boiling thermal grease isnt good...

Same from the backside. I fixed those screws with soft epoxy to avoid them getting loose.

Ok, the reflectors arent fixed yet, its just so i dont become blind Surprisingly low Vf needs me to change my battery plans (seems like my 5xA123 is total overkill now, as the following baby showed)

Ghetto battery pack, made of shoe-string and battery clips. It really works when it doesnt create short-circuits (and nice sparks ). And its more than enough to drive the light at 2.8A using a linear buck driver.

The light is already a big further ( i used it to blind people at work already ), more (including the potentiometer control) will come later.

Edit:
just went outside with the barebone system.
Its not really night yet, and there are streetlights all around (lots of low pressure sodium).
But boy, this is bright. in a quick check it seems brighter than my (halogen) car headlights.
I have to replace those cheap plastic front windows, though. I just put them in place because i didnt want the reflectors to get dirty till the AR coated glass ones arive.

But out there, i could really see two spill cones: the "real" one, and a second, wider one, easily as bright as the spill of my P3D on high, which was caused by double reflection in the the windows.

 

[toby_pra]

Re: Quad P7 custom thrower (edit first pics)

looks very impressive so far!oo:

 

[Kid9P]

Re: Quad P7 custom thrower (edit first pics)

That looks so cool :thumbsup:

I would love to see an outdoor beamshot with that monster :naughty:

 

[Gunner12]

Re: Quad P7 custom thrower (edit first pics)

When you're done, I would love to see some beamshots.

37.6w, that's quite a bit of power.

You won't be using the light for longer periods right?

 

[IMSabbel]

Re: Quad P7 custom thrower (edit first pics)

When you're done, I would love to see some beamshots.

37.6w, that's quite a bit of power.

You won't be using the light for longer periods right?

Well...
Wait a few minutes, then i will show you the solution for the power problem. (videos and pics are just uploading)
I did all the calculations: With the heatsink attached, the light can run indefinitely at full power, overdriven to 4A (thats 65W including driver loss).

 

[IMSabbel]

Re: Quad P7 custom thrower (edit first pics)

A beamshot of a different kind....
Spill light of the quad vs hotspot of a P3D on Turbo.
Its a _little_ bit brighter

The current setup with new battery pack. 5*A123 is a perfect match in voltage (including about 1.5V loss in the linear driver and the PWM setup, 4*18650 are only good for full power the first 25% or so of the charge).
I meassured after the driver, and got 13.48V (compare to the picture with the psu, subtract loss at the crocodile clamps, and its clear its driving with the full 2.8A).
The linear poti runs a 500Hz PWM with "inifinite" (or at least quite a lot) steps.

Here the backside of the light.
There is a reason for those 2 additional holes, and the free area in the middle. This thermalright heatsink is able to keep a 80W CPU below 70C die temperature. Which should be enough to keep the P7 Dies below 100C (including all interface resistances). Indefinitely.
Actually running around outdoors should keep it a lot cooler (i estimate about 40-45C heatsink temperature.

Here another shot of the setup. The cooler is quite compact (for a heatpipe monster), but the fin spacings are optimized for passive running. Plus the fins are sturdy enough to be "grappable", i.e. they dont bend at the slightest provocation (hello, scythe ninja...).

On the right is the driver module, btw. Perfect fit there. Also note that the MOLEX connector is only temporary, as i didnt have anything suitable around. Any recommendations for nice connectors?

Edit: the reflector is not yet attached. The heat capacity of the plate is big enough to allow about 5 minutes on full power, which is enough for testing up to now.
And yeah, i can run it at 150mA. More throw than a p3D on turbo, 15 hours runtime and stone-cold operation

Videos:
(note: picture quality is quite rotten). Plus i managed to break a solder point on my light on the way down to the woods, so i had to go back without taking pictures down there. The only thing i got is a video from the stairs...
Also excuse by bad english...

http://www.youtube.com/watch?v=iuJKRWBdhng

Thats how the PWM works. You can see the duty cycle nicely in the overflow artefacts of the CCD. Its not really visible with the naked eye.

http://www.youtube.com/watch?v=jZ3gD6wXDeA

Ok, this is the afore mentioned video from the stairs. The trees at the end are about 50-60m away.

The cat i am talking about isnt really visible here in the video at all, but i could see it nicely. Very golden eyes reflecting. But the p7 was to bright, so its looked away (had no problem looking into the P3D, though )
The light is brighter then it looks like there... ( even the p3d on low and medium was usefully bright in person)

Btw: I made this video because its infinitely easier to do with a digicam without shutter control than good comparison shots. I hope its still useful.

 

[Gunner12]

Re: Quad P7 custom thrower (edit first pics)

I didn't read the part where you said you had a heatpipe cooler.

Now that I see it, I don't think there will be any problem with cooling.

 

[IMSabbel]

Aw poop.
The PWM regulator bought it. For some reason, its now always 100%. Started totally random (was connected for quite some time, but put to 0%. Then i wanted to start powering up... and ding... full power, never to go away again).
The regulator should have been wenn inside its specifications (7-30V, max 10A).
Damn, this will make it impossible to complete this light (at least _with_ dimming) before i have to go overseas for a while.

Well, maybe the concept was a bit _too_ neat to be true...

 

[Der Wichtel]

Is that the pwm regulator from Led-tech?

 

[IMSabbel]

Is that the pwm regulator from Led-tech?

Yes, it is. Are there any known problems with it?

 

[Der Wichtel]

No, I don't know any but maybe you should start a thread at ledstyles.de.


I just can figure out a ne555 and lm258.

If there is a mosfet on the dimmer it should not be defect since it still turns on.

Check the Ne555 and lm258
Check the slide resistor as well and have a look on the resistance

 

[IMSabbel]

Already took a look at the potentiometer.
Works fine (range seems to be 0-850 Ohm).

Is ledstyles.de affiliated with led-tech?

 

[IMSabbel]

I definitively smelled a little bit of magic smoke escaping. If i put my nose right against it, i can smell a little from the side with the op amps. But i cannot see any obvious damage. Maybe you can see something here.

 

[A123Powered]

I have been following your thread recently and saw something the other day that could make a decent enclosure for your light:

http://www.pilgrimwineandspirits.com/images/heineken-mini-keg-large-2.jpg

Just attach a handle and it could be the first quad P7 keg light

 

[IMSabbel]

Damn. Then i wouldnt need my heatpipe cooler.

Water^H^H^H^H Beer-cooling would be the obvious solution

 

[Bimmerboy]

Hehe... this thing is already a monster. Can hardly wait to see it in a portable housing. Love that sliding pot!

I'm planning out a similar idea (in some aspects), though nowhere near as professionally made. It'll be more of a "sleeper", and built like one too... lol.

If possible, can you report the temperatures you're getting across the aluminum plate at max power over a certain period of time? It would be very helpful to know if stays cool enough to not melt most low-temp plastics on contact. Yes, I said plastic. :naughty:

Damn. Then i wouldnt need my heatpipe cooler.

Water^H^H^H^H Beer-cooling would be the obvious solution

Giving new meaning to the term "light beer". :nana:

 

[IMSabbel]

Well, got a replacement for the dimmer. Lower PWM frequency, but much less heat, and a really _smooooooth_ potentiometer.
This reduces the lowerst possible duty cycle to about 1 lumen (you can look directly into the barrel without squinting).

Also, i got me some eneloops, and already did a test. Works perfectly nice with a 14s configuration. 12S might drop a bit too much under 3A (if you include the driver losses).

Is there any _good_ high current battery holder thats adviceable? I got me some cheap ones, but at 3A the thin springs ended up melting into the plastic, which tells a lot about the resistance...

My current (intermediate) plan was using PCV tubes. got some with 16mm internal diameter, a Eneloop just fits nicely. But interconnecting the tubes aint easy.

So anybody an idea how to not-ugly carry around 12-16 AAs? Doesnt need to be compact, it might also act as a grip in the final design ( i think i can encapsulate the whole electronics below and just have the heatpipes come out of the epoxy...)