Ultrafire sk98 clone mod (bling)

Ah, good to know. Still confusing that it increases throw without the reflector…weird.

Thx for advice about thermal path. I’ve noticed that the copper star form KD is a bit curved and i was thinking about sanding it for better thermal path but since i’ve applied thermal paste mixed together with bits of filed copper from rasping the copper pipe i’ve find it good enaugh. Body is warming under a minute on high mode which tells me that the thermal path is sufficient. I was impressed with yours and djozz’s T-twenties with dedomed xp-g2. First minute after reading that specific mod uf-t20 was immediately in my wishlist :heart_eyes:

EDIT: Yeah… zoomies are gr8 all-rounders :slight_smile:

Nice mod, I like the bling and the desk cloth most.

In former times I fell in love with the sk98 as well(when xmlU3 was THE THING) but today I seldom use it because I hate all the light what the aspherical steals and focused the beam isn’t really tight because of the XML, so I am a bit away from zoomies. Also the switches are most high resistance ones…
Check these 20mm adapterings for future mods
Mixing copper rasping in thermal paste is not ideal, the paste is just for bridging micro gaps between star and pill. Any rasp will be much thicker than useful…
But driven at 2.5A this is no big deal either.

Thx Werner for the link of adapters. About rasp… I thought about that but since the diy copper plate was not perfectly leveled with the pill so I thought that mixing would be useful but as you said the point of thermal paste is for filling micro gaps.I shouldn’t mix it… From time to time I take the switch apart and clean the contacts for eliminating as much resistance as possible. Btw emitter is driven at 3 amps(probably :stuck_out_tongue: ) .

These adapter plates save a lot of work, I have once soldered a wire around a nanjg to fit it to 20mm, very ugly and took a while until it fitted and made good contact.

These drivers fall out of regulation very fast with decreasing voltage, with the switch resistance you will faster have a <2.5A driven LED than you can imagine.

I’ve came back with tailcap readings. With short thick copper wires my cheap dmm showed 3.22A with fresh battery. I’ve tried to measure resistance of tailcap on 200ohm range. It showed 00.0 but after 5-6 seconds it showed 00.1ohm. I’ll try to modify it with some solder but I think there will be no much gain in terms of reduced resistance.

I am planning similar driver for my Raysoon TD-398. I ordered an adapter plate. Guess I need to solder it to the back of the driver. Your copper is better but maybe the plate will be easier. I like this style for mod because of the cooling.
Nice polish. I have diamond sand paper. Wonder if I could polish the black anodizing with it.

These adapters are made for nanjg, it’s very easy just stick the spring through the hole and solder it to the middle pad. Add a piece of wire from the outer ring of the adapter to the nanjg for ground contact.
—-
Nice to see that you now can measure current, despite 3.2A seems a bit high…

Have you soldered the brass part of the switch to the switch itself? This helps a lot. I would supply a photo but imageshack is sucking at the moment…
I am not sure if I have modified the switch innards but I am really sure that it was crappy on stock form in the sipik and in the ultrafire branded version I got. Maybe you got a better version.

Just before a few days I played again with it to test a direct drive driver, which puts 10A through a triple XML board…this is what I call hot.

Maybe my DMM is calibrated for stock leads and that’s the reason why my measuring result is bit higher than it should be. My version had 5 mode driver with dreadful next mode memory and screw plate for tightening emitter star to the pill. Thx for idea of soldering direct on the switch. Soldering iron is heating up :slight_smile:

EDIT: I’ve disassembled the switch and apply solder on all three contact plates.

These cheap DMM normally just have a piece of copper wire as shunt resistor to measure current. They “calibrate” them with pinching them with pliers, this can be done by yourself too. But you will need a known current to “calibrate” against.
—-
That is the best way to go on these switches. That reduces resistance a lot on cheap switches.
If you ever identify one which routes the battery negative contact through the spring, replace it immediately these will melt down on high currents.

The omten spare switches from fancyflashlights seem to be the best replacement they have near to zero resistance. It’s handy to have some around as these small switch parts can get lost fast…
I always plug switches/tailcaps to my bench power supply and push 3-5A through it, I measure the voltage drop to judge about them.

I will get some omtens for my future mods. I was planing placing order From FT but since China is celebrating lunar new year I have to be patient.

Nice polish and great mod. Thanks for sharing! :slight_smile:

Great Mod! Nice polish job on the light. Good to see anyone doing more modding.

Very nice. I loved the polished look. Could you elaborate on how you did the polishing?

Thanks Old-Lumens. You inspire me with your adroitness and adaptability! :slight_smile:

Scroll at post #11 …. After P500 sand paper two other steps (with P1500 and P2000) are done in washbowl with water. It’s was a lot of work mainly because machining on my sk98 clone was nasty and rough. Hours of work… Polishing is done with “Purol univerzal” since I didn’t have anything better. You’ll be better of with “Autosol” polishing paste.

I got my new driver working in my Raysoon TD-368, after various problems. It is greatly improved. Just now, I looked at the tail cap to see how to add copper braid. A retaining ring unscrews with needle nose pliers, but the plastic under it stays put. I suppose I need to force it out, but I hate to take a chance of breaking it now that the light is working so well.

I should be able to explain this. I know I don’t need to, but it will give me a rest from struggling with my less developed fabrication skills. The main thing I learned from DrJones is that the power density per phase space volume or luminance is increased by dedoming, making more throw for a given optics diameter possible, with ideal optics. This is because the dome allows light that would otherwise be totally internally reflected to escape and enlarge the phase space.
The dome is a focusing optical element, but it is so close to the junction and the phosphor that it interacts with them. So removing it makes the led appear smaller to the rest of the optics, spreads the light towards grazing angle and causes grazing angle light to be reflected back into the phosphor where some of it contributes to the throw and improves the color.
For a reflector light, it is clear that it improves the throw, because the grazing angle light is focused. For a lens light, the focusing dome is more useful, because grazing angle light is lost, so there is more loss in total output with dedoming. Apparently it is DrJones’ luminance argument that actually determines the throw, with the lower total output affecting only the beam spot size. One can probably prove that, but for the moment I rely on your experiments to show it.

From my experience these holders for switches are made from soft elastic-ish plastic. You can’t brake it easily. Just grab the same needle nose pliers and force it out with wobbling. You may end up with removed plastic holder without switch because metal plate of ground contact from switch can usually stuck in threads.

De-doming effects the brightness of the spill area as well (reduces it) - I've seen this many times.

Assuming that you are talking about reflector optics, the simplest explanation is that the done is optically converging, so eliminating it spreads the light out sideways where it hits the reflector and becomes part of the hot spot.