P7 Testbed Pics(Future P7 Tiablo A9)

[Hazman]

Hey folks!

My P7 arrived just the other day :twothumbs, unfortunately they didn't ship the drivers I were to be using. I'll be using one 5 mode driver board and two slaves in a setup near identical to the one which NetKidz proposed(link follows).

http://www.candlepowerforums.com/vb/showthread.php?t=192677

In the meantime seeing as I don't have any drivers. I decided to have a fiddle with the P7 in a direct drive setup. To keep it cool I have used a spare Fan/Heatsink from a Core2 Duo, flipped upside down, which is running off of a modified computer power supply.

As many of you will be aware, there aren't any battery holders for the 18650's, however as I have done, you may be able to construct one by cutting up two single c-cell battery snaps, and then gluing both portions together using 3 pieces of perspex cut to appropriate dimensions(the perspex is about 3mm thick - I haven't actually measured it). The battery fits perfectly .

Anyhow, the P7 has been stuck to the copper section of the Heatsink with Arctic Silver thermal compound.

With my modified reflector resting atop of the LED.

To increase the diameter of the hole which was originally existant for the Cree XR-E I used a tapered grinding stone and slowly increased the diameter. After which i used a thin cylindrical sanding attachment to increase the hole further and faster(the tapered grinding stone was beginning to "stick"). Unfortunately some of the sanding had the effect of giving the inner-most region of the reflector a sandblasting with offcuts. However the effect of this is barely noticeable and probably adds toward reducing the slight donut hole.

Another shot, this time without the flash(I had to adjust the brightness/contrast a little so it's a bit grainy).

Anyhow, here are the pics you guys are after!

Unfortunately due to nature of the setup this is the only beamshot I have. There will be more in the near future, I promise!(You can't see it in this picture, but there is a slight donut hole in the centre)

One thing I have noticed is that the globe only seems to be running at about 1amp(the current drops slightly due to the internal resistance of resistance of the multimeter I am using). It does look brighter than the old XR-E at least.

Obviously the above is a pic of my multimeter: A cheapie.

Would anyone know why the amperage isn't as high as the globe is capable of handling(reading 870mA)? I checked the temperature of the globe and it's only around the 30 Degrees Celcius, so it's not getting too hot. However as my multimeter is rather cheap I'm naturally skeptical of the reading that it is getting(unfortunately I don't know anyone with a working one i can test this thought).

I've tried using two 18650's in parallel with near no change in intensity, and a marginal current increase of 90mA bringing the total to 990mA. So clearly the cause is not the batteries limited current output. The voltage across the LED is a recommended 3.6volts, and the LED is a C-Bin from DX.
http://www.dealextreme.com/details.dx/sku.12721

Later on I'm going to charge up a fresh battery and let the light drain it. Seeing how long it takes will naturally give me an idea of the current flow. So hopefully this will validate my assumption that my multimeter is damned inaccurate.

Any help on the issue(low amperage) would be appreciated, Cheers!

 

[sactime]

Looks good. I am planning to do the same mod when my A9 arrives. How many amps are you going to be running the P7 at?

Take plenty of pictures as you go along, make sure to keep this updated.

Good luck.

 

[TexLite]

Would anyone know why the amperage isn't as high...

Any help on the issue(low amperage) would be appreciated, Cheers!

Most likely an emitter with high Vf.Try using 4 NiMH.

Do you have any way to measure the Vf with the emitter at 2.8A?

-Michael

 

[KowShak]

Does the P7 seem brighter connected to the battery without the multimeter connected than with? The multimeter itself has a resistance which will affect how much current the LED will take direct drive, it may take closer to its rated current without the meter in series with it.

 

[ergotelis]

HI vF led. I had the same question till i made some tests.Try to make good wiring, reduce resistance to minimum and use a buck converter and higher voltage supply. Or at least test it for a while with 4x Nimh.

 

[Hazman]

Yeah, the LED was brighter with the multimeter out of the loop, although, only marginally. So i figured by guess that it would probably only be drawing about 1amp rather than the .870mA or so that I measured.

Knowing the internal resistance of the LiIon's i calculated the current being drawn from the battery from the voltage drop.

Internal Resistance(R)=180mOhms
Initial Voltage(V)=4.00Volts
Voltage Under Load(V)=3.60Volts

V=IR, I=V/R

If we consider the voltage of the battery both before and after a load is applied we have the following formula.

Vinitial=Vload+Vresistance

3.93=3.70+IR
3.93-3.70=IR
0.23=IR
I=.23/R=0.23/.180=1.28amps

I would trust this figure alot more than the previous reading as there was no drop in brightness when measuring voltages.

I also tried the globe with 4xNiMH's, which made the led alot brighter. I got a reading of around 2.2amps(from memory - this was the same as the old previous measurement, inline), and the voltage across the LED measured at 4.40volts.

This would mean that under this situation the LED would be drawing about 3.23Amps, which would be expected given that the LED is no doubt being over driven. Before reading someone else's comment on this, I too came to the conclusion that the LED is a high binned voltage, perhaps K-binned. Consequently I have ordered a C***I binned P7 and am eagerly awaiting it's arrival .

Next time I'll avoid ordering LED's from DX.

That aside. I've successfully managed to create a driver sandwich with the 3 DX boards that I mentioned previously, although I can't really test it out until I have my new LED here. Unfortunately in my rushed attempt to remove the LED from the aluminium board I managed to knock it's top off :mecry:.

 

[sactime]

any updates?

 

[Hazman]

Back again.

So today a DSWOJ binned SSC P7 finally arrived(well, a new one anyways, as I killed the old one). I actually ordered a CSXOI, so i'm not sure how well this will work considering I was planning on regulating amperage. Does anyone have any experience here? In direct drive my battery drops to 3.51v from 4.02v, and has an internal resistance of 180mOhm, which suggests it was pulling about 2.83amps, which is about the current that what I want, but all the time(with the option of dimming etc).

Perhaps someone could point me toward some li-ions with impedance much less than 180mOhm(Currently using TR18650 2500mAh 3.7v)? That way the voltage arriving at the emitter will have a tendancy to sit more within the voltage range of the emitter(J = 3.5v to 3.75v).

Anyhow, I naturally decided to continue my work on my mod, and I got back to work on this little project.

Above I have soldered in tanned wires which will serve to both provide electrical connection between the appropriate tracks of the boards and hold the boards in one stack.

The two outside(upper and lower) leads are -Battery(two wasn't necessary, but I added a 2nd anyways). The right lead is +Battery and +LED, left is -LED and finally the lead that runs nearly horizontal on the right hand side is connected to the base of the 7135 chip, which will allow the IC to control all 3 current regulators at once.

Note: It may look like the two left-most leads which are vertical and horizontal are contacting, however I can assure you they aren't.

With the 2nd board added, wires that protrude through each board have been soldered, and the "signal"(from the IC) wire has been looped back over to contact the same 7135 current regulator on both boards. The diodes from the top two boards have been removed so the IC on the lower board can control all 3 boards by itself.

Natively the board doesn't have a hole for the -LED wire, so i drilled one in each of the 3 boards right next to where I would otherwise connect this wire and loosely connect the wire running through the board. This helped save on tricky wiring.

Third board added.
Oh, and I also took out a portion of the top two boards so that I could run the "signal" wire between all the boards without having to sacrifice any of the sleeve to make room. Even though this cuts through the -Battery track on the board, the track still runs around the entire board so no harm done.

Contact Spring and Wire added.

This shows my new sleeve for holding the sandwich. I had a friend of mine, Jamie machine this up. As you can see he did a sweet job.

LHS: New sleeve, which is made to house 3 boards.
RHS: Old sleeve, which was made to house the original 1 board(note - I did bore out the centre to fit the new boards originally as they were slightly larger).

Complete driver "unit". Yes, i fiddled with the brightness/contrast with this last pic.

In the morning i'll run around to the shops to get myself some thermal epoxy so I can finish off this mod and get my P7 mounted, so there should be some more pics soon.

 

[Isthereanybodyoutthere]

And how is it going ??

 

[Hazman]

I've been a bit distracted for the past number of months, but anyhow, I did end up finishing my torch some time ago, and of course, here are the pictures.

The underside of the P7 is coated with computer heatsink thermal compound, specifically Artic Silver 5 or 7, I can't recall which one. The white stuff is thermal epoxy, i figured it was suitable to get a little more heat out of there.

Unfortunately I dropped the lens and there is a chip which can be seen in the left hand side

Holding the driver is three dabs of thermal epoxy, again, heat conduction ftw! Only two dabs can be seen in this pic.

Tada! Here she is in her entirety.

Of course, I can't forget the beamshot. Even though it can't be seen because of the exposure time, there is a doughnut there, I just couldn't be bothered adjusting the exposer. But when i get the spare time i'm going to fix that.

Enjoy!

- Haz

 

[Tatjanamagic]

Very nice...

Can U help me how to replace original tiablo driver for q5 led with something similar like that? Maybe with high mode only...

I would very appreciate that...

Or at least how to disassemble driver from housing? I removed the plastic cap, unsolder the emitter and now don't know what to do and with what tool to open it...

 

[Aircraft800]

Wow! He makes a nice sandwich! I hope the OP comes back!
Last Activity: 06-06-2009