Smartphone to Laptop transformation (no sponsor)

Hi Matt, great video. I wonder if you can give the sizes of the aluminium sheet you used and all of the sizes for this project (the aluminum profile and plate, in mm or cm) and another question- how did you connected the 45 degree aluminum profile? With epoxy?

But to do that, you need to get a better understanding of where the losses are. And then plug them with a bit of $$$.

I think the overall goal would be to make a 4K projector you can see in daylight. Or at least without bright roomlights on.

Did you call up Edmund's Optics when you were researching this? They supply mirrors and fresnel lenses, and have good tech support. And are BRITISH....

Anyways, thanks for the idea. Now I can see from Mr. Google this is a VERY popular idea. I never knew.

In your case, you can fix it by adjusting the LCD's colour balance. But I think that for $100, one can get a better LED.

But they likely feel safe in claiming that, because how many people actually read the spec sheets? Well, except me. And it's not so easy to measure a colour spectrum. But to measure lumens, well, for £20 ANYONE can....:-0

So their advertised specs about 6000 lumen at 5600K and 4amps are not substantiated by their actual lab test.

So if you run at 25V, you get about 4500 Lumens (quite a bit less than the rated 6000 lumens), and a colour temp of 3300K. Quite red.

You can see from the spectrums why: check the nominal Spectrum of the 5600K test: it's VERY low (45%) excpt for the one peak. Whereas the first test is much more uniform.

And to get their claimed 5600K, they had to cut back on the volts a LOT, to reduce the reds in the output. And sure, it was the claimed temp, but VERY low lumens: 82 lumens!! With with 21V, and 0.05A.

Check the spectrum.VERY red.

Did you check the Spectrum Test Report on your LED? It's VERY revealing. There are two tests, done at drastically different voltages. The first is at 25.78V, and is what they use to justify their claim of 100W. And 6000 lumen. However the TEST report admits the truth. They only got 4200 lumen. And colour temp was 3300K.

Aha.....I see what the problem is with your projector, and not being very bright, even tho it's SUPPOSED to be a 6000 lumen LED. But those are Chinese specs.

But I don't care if you bought it, I do care if you could do the measurements though. That would be a big help. Good for you too.

This is what I have.

If you wanted it, I could sell it to you for whatever its's going for now, and send it for free to you. I had it on eBay I think, but didn't really persevere. I thought maybe need two at some point. Was very good price for a nice bit of kit. I see it sells for under £20 now on Banggood UK.

But that in iteslf is useful to evaluate if your cooling itself is optimum.

From my tests with powerfull LEDS, they REALLY drop for the first 10-20minutes. Might be a function of cooling though: I was testing projector LED's without active cooling.

It has a standard tripod type threaded hole at the back of the case, so easy to hold in place. And with BT, you set it up, see output on your phone, and decide when steady state is reached.

It uses two LED's at two base wavelengths. And averages the results to subtract the IR, and give just visible in Lumens. Says it is +/- 4% of reading at <9,999 Lumens. Think that's honest value too.

I said I have a UNIT-T UT383BT (BlueToth) light sensor (reads in lumens). in fact I have TWO: my thumb ordered TWO instead of one off eBay....sigh?? The one is still pristine in the sealed box etc etc.

Hmmm...seems to have swallowed a post.

Is handheld, and with the BlueTooth you can mount on a tripod/stand etc, and then leave it all to stabilize, and capture te reading by BT on your phone. It has a standard camra type female threaded hole on the back that a tripod grub screw can go into.

You can get Mr. Googles to explain, but basically it uses two LEDs to measure light, and averages the outoput from the two to get just visible light. It claims +/- 4% of reading at < 10,000 Lumen and I think that's probably a fair value.

Still in sealed box actually.

As far as the light meter is concerned, I have TWO....I did a mistake on eBay, and ordered a double. From China. Is VERY good design. Very accurate from what I've see. Is the UNI-T UT383 BT (Blue Tooth model).

With some time in between.

And sorry, if this all seems like a barrage, but I was VERY interested in this, and went through your design, and thought abiout it some more. Was MUCH easier to critique an EXISTING design: it's always the case. To REALLY make something well, you need to make at least THREE of them.

In fact, I would be interested in dropping by and helping. I'm Covid free: I got a test a week ago because I was curious.

If you don't have a nice thermocouple for a multimeter, I have a spare I could lend you. I'm assuming you have a multimeter. If not, well, the Warthog instrument loan service could no doubt add that...:-}

OK, so lots of input, I hope it all seems constructive. I'm in Sheffield. It seems you are somewhere in the Land of Brexxit Boris the Covid Killer....sigh.

Often the FRONT of LED's doesn't get very hot, it's just the BACK of them.

And if you did that, the SAME fan would cool the LCD better. Blo directly across that, and then circulate out below the Fresnels. You could draw from an inlet up near the mirro, then down to the fan. Blow it across the LCD, then exit ABOVE the LED.

In fact, I a not sure it's such a great idea to use the same air from the fan exhaust to "cool" the rest of the system. How hot does it get? After it cools the LED. If it's only a degree of so, fine. But if it rises above ambient much, you not cooling the rest as much as if you had a 2nd fan. Fans aren't expensive.

The LCD is the main one.

Because if you cool the LED, and stop THAT heat from getting to the rest of the system, you shouldn't actually NEED to get a big heat build up if you use low loss components.

Could then couple that to do an enegy balance with the light loss. Not exact, but easy, and useful to see what needs better cooling.

To be honest, as well as measuring the lumens, it would be nice to measure temperatures at steady state. Easy to insert a thermocouple from a multi-meter (with the sides installed, obviously).

And if it does'nt get hot in the space by the mirror, I'd be tempted to seal that off from the main air circulation to stop dust.

And dust? I see you have a fairly crude filter, but maybe could use better FEMA type foam: not expensive. And for between the Fesnel's, I'd be tempted to isolate them with FEMA foam to stop dust.

If one were to have several cases, where you used basic components, and then same sizes, but higher quality/low loss stuff, that would be very useful. And then could do a different one using a big LED, and bigger overall system.

I am going to look up the various material data. And will make a basic model to estimate losses. So when have that, I can give it to you in exchange for the lumens measurements....:-}

You interested? Would really add to the design. You could make a nice video just based on that actually.

Would need measurent at say (1) halfway between LED and Fresnel, (2) halfway between Fresenel and LCD (3) halfway between LCD and standard lens. (4) maybe two measurents between the lens and the mirror (or just one if can get it dead centre). (5) In fact, need one WITH, and WITHOUT the mirror to get the losses. (6) last but not least, at a set distance from the mirror centre spot. (7) Bonus one, measurements at the projection surface, at centre, and edges.

Anyways, would be REALLY nice to measure your system at each point. And then can calibrate the model, and then predict what happens using different materials.

Next is the plastic cover. Depending on what you use, can lose there too. Easy to get something with low loss. Is surprising what works sometimes. Some types of cling film are VERY good at transmission of IR, but I don't have data on hand for visible. Although it's so thin, should really matter

Next is the mirror. I've never made one, but you lose light there too. Edmunds Optics says cn go from 85% transmittance for standard front surface mirror, and get up to 95% on higher quality ones. That's a pretty big loss right there.

So I'd say you want as big an aperture lens as you can afford. But again, need to measure the loss, and compare to calculations.

Next loss is the lens. I need to work out the optics, but I'd say you want it at open aperture if you can. After all, you are projecting onto a flat surface, so depth of field is no issue.

But i think in the least having a fan there would help.

In terms of cooling, yes, but not liquid cooled as a layer below the lens. You'd have MASSIVE losses then I think. Need to check, maybe water isn't really bad.

Next is why do you have a plastic sheet UNDER the LCD? That just seems o me to be another way to reduce transmission, and make heat. Yeah? Like why not just mount the LCD "bare"?

I didn't catch your comment about removing the polarising coating on the LCD. I'll read up on that. Perhaps that is the culprit.

Next is the LCD. Be nice to know what the transmittance loss is. And compare it to a stand-alone type that is designed for though vision imaging (or at least i assume some are).

Heat too.

Next are the Fresnel lenses. These are plastic, and maybe not best type for low transmission loss. Need to check that. Sure, ones with less loss are more expensive I'd assume, but might have a major impact on losses.

I'll need to check the LED output, but an IR filter might help in terms of heat.

But that mainly means a BIGGER one, which impacts other items. Not a big deal though.

Like you can get a brighter LED, and with a tighter output in the visible region. And not send IR.

And once you know that, for a given design, you could put in prices of components, and see what makes the most sense.

I'd like to revisit the design in terms of efficiency, in terms of looking at the LED output (at least the useful output), and see where the losses are.

Would it be possible for you to make a few measurements of the light intensity at a few places? Ideally in Lumens. I have a very nice Chinese meter, and if you don't own one, or want to buy one right now, I could lend you mine to do it. I'm in Sheffield.

Couple questions, please. I'm a materials engineer, and so I see some of this in that context.

Whoops. Start again. Nice video and idea, I'm going to use it as the basis for my own.

Hi Matthew,

I got my screen today. I plugged the control board into the back of the screen, plugged a USB-C 3.1 cable into the daughter board and my phone. Nothing happened. I'm going to keep trying, but I'm stumped right now.

Be nice if you provided more links to as what you used

Great video with lots of scope for those broken or damaged screens,,, Thanks Matt

I like the idea of being able to connect even a fully functional to this setup. I don't know if I've ever seen anything like this as a purchasable item. Very well done, Matt!

Mike Allison

Having the phone mount on the back of the screen was actually quite a brilliant solution. I was curious what the back USB cable was for but that works very well.

Graham

Thanks! I love the way you built it to turn on and off the phone. Very unique

Thank you!