How Can I Build an Effective DIY Helmet-Mounted HUD?

[xxtop]

TL;DR Summary

Trying to build my own HMD

Hello,

I'm trying to build a helmet-mounted display. Basically a HUD that is contained fully within a helmet. I am not experienced with optics though and can only sift through what the internet has to offer. I don't want it to be a "put your phone down next to an angled piece of polycarbonate" DIY HUD but rather an actual piece of engineered hardware. The FOV should be as large as the components (and the price) permit. Having the combiner be curved would be great but I assume that is not feasible.

Therefore my imaginative setup would look like this: A reasonably sized piece of teleprompter mirror (optical combiner, sourced from https://telepromptermirror.com/) angled at 45° with the lower side towards the viewpoint. Above a planar convex fresnel lens (sourced from https://www.edmundoptics.com) with its focal point in the middle of the combiner. To guide the beams towards the lens, a first surface mirror (sourced from https://firstsurfacemirror.com/) angled at 45° to reflect a LCD display. I have made a crude drawing of how I imagine this:

The lower is the drawing I based most of this on. In my usecase I have some space constraints though and would like to place the LCD in the same direction as the viewpoint. Therefore the first-surface mirror is be angled the other way. Does this interfere with how the LCD is perceived from the viewpoint?

Adding to this:
Many drawings feature a second lens in front of the first-surface mirror. What's the purpose of that one? The lenses used here look like lenses corrected for some sort of aberration, is this relevant? I expected chromatic aberration to be negligible for a monochromatic display.

What is the purpose of the lenses? I can find that they "collimate" the beams. But what does that do? Why doesn't the reflection of the LCD in the mirror and then on the combiner suffice?

Do the circular lenses limit me to a circular HUD?

How do I figure out the focal point(s) I need for a specific setup? How do I calculate the size of the lens? Is it correct to use a fresnel lense for this? I have read about condenser lenses being viable too.

Is the LCD magnified or does it need to be the same size as the reflection I see on the combiner?

Thanks for taking the time to read this, even though this is probably a nightmare from an optical engineers perspective. Please excuse my grammar and expression, this is not my first language so some of this may be phrased weirdly.

 

[berkeman]

Welcome to PF.

The optics in your diagrams look more like a vehicle-mounted HUD, not a helmet-mounted HUD. Are you familiar with more compact units like Google Glass and the one-eye HUDs worn by military pilots?

https://en.wikipedia.org/wiki/Google_Glass

https://www.shephardmedia.com/news/air-warfare/us-orders-integrated-helmet-spares/

 

[xxtop]

Welcome to PF.

The optics in your diagrams look more like a vehicle-mounted HUD, not a helmet-mounted HUD. Are you familiar with more compact units like Google Glass and the one-eye HUDs worn by military pilots?

https://en.wikipedia.org/wiki/Google_Glass

View attachment 346952

View attachment 346953
https://www.shephardmedia.com/news/air-warfare/us-orders-integrated-helmet-spares/

Yeah I was worried I would have issues with space.

I couldn't find much information on how these work, I haven't looked into the google glass and I assume Lockheed doesn't publish too much data on how a JHMCS or the F-35s helmet work . The F-35s helmet was part of the inspiration but obviously that is slightly above my level of expertise and manufacturing.
I saw that the newest one-eyed rotary wing ones use a waveguide but I can not find that up for sale and I'm not even sure I understand how it works.

If there's a better way to do this, I'm up for it, but so far I think I'll have to shrink down this admittedly aircraft-looking system and live with the compromise of a smaller combiner eventually.

Does this work in theory or did I mess up the concept already?

Thanks for your answer

 

[berkeman]

Can you say what your application is? What kind of helmet?

 

[xxtop]

Can you say what your application is? What kind of helmet?

I don't have a very specific application. I want to integrate it in outdoor equipment, have it display something like temperature, a compass, gps data etc.
It would be cool to integrate an infrared camera (like nvg or thermal) but that might be a bit far fetched.
Might use it for mountaineering, off grid hiking, something to that extent.

To add to this: it's not meant for any critical application. I won't be flying with it or anything. The helmet will likely have to be custom, probably from carbon fiber.

 

[berkeman]

Maybe consider joining the Google Glass Developer's Group. If your applications are novel enough, you may be able to get a Beta version of the glasses to develop with. I don't know the current state of that program, though, since they seem to have gone quiet the last couple of years.

https://developers.google.com/glass

 

[xxtop]

Maybe consider joining the Google Glass Developer's Group. If your applications are novel enough, you may be able to get a Beta version of the glasses to develop with. I don't know the current state of that program, though, since they seem to have gone quiet the last couple of years.

https://developers.google.com/glass

So in essence you are saying doing this DIY is infeasible? Part of the fun of this would be to design and engineer this, so it might just be a project I will scrap then. Where do you see the issue? Space constraints, complexity?

 

[A.T.]

...it's not meant for any critical application. ...

Neither is Google glass, which didn't take off either. One reason might be the question: If the information is not critical, what is the point being able to display it all the time.

 

[sophiecentaur]

TL;DR Summary: Trying to build my own HMD

What is the purpose of the lenses?

In a HUD, the display will be small and very near your eye. You want to be looking at an image that is at a 'comfortable' distance so that it will appear to be within the range of distances of objects in the scene in front of you (i.e. to appear at something like infinity.
This is the same principle as for the glasses needed for someone with 'long sight'. They can't focus on the newspaper so a convex lens of the right focal length will take the newspaper and 'put it' at a distance where their eye can focus. The lens in a HUD needs to be stronger because the difference between the distances involved is much greater.

That's the simple answer but, depending on the physical size of the display, more complicated optics are needed to change the angle subtended by the image of the displayed data to match the background scene..

If you want to get a feel for this then get hold of a magnifying glass (a suitable convex lens for this purpose) and experiment with 'focussing' on some text held very close to you.

 

[xxtop]

Neither is Google glass, which didn't take off either. One reason might be the question: If the information is not critical, what is the point being able to display it all the time.

As I said, part of it is the way there. I feel like the end product would be cool and the way there is an engineering challenge and something to constantly fiddle with and improve

 

[xxtop]

In a HUD, the display will be small and very near your eye. You want to be looking at an image that is at a 'comfortable' distance so that it will appear to be within the range of distances of objects in the scene in front of you (i.e. to appear at something like infinity.
This is the same principle as for the glasses needed for someone with 'long sight'. They can't focus on the newspaper so a convex lens of the right focal length will take the newspaper and 'put it' at a distance where their eye can focus. The lens in a HUD needs to be stronger because the difference between the distances involved is much greater.

That's the simple answer but, depending on the physical size of the display, more complicated optics are needed to change the angle subtended by the image of the displayed data to match the background scene..

If you want to get a feel for this then get hold of a magnifying glass (a suitable convex lens for this purpose) and experiment with 'focussing' on some text held very close to you.

This is some good eli5 :D

So that also answers the question wether my display is the same size as the reflection, which it is not then if the lens magnifies it. Where does the focal point need to be? Do I use a lens with both sides convex and have the focus point on the mirror and on the combiner? Or does a planar convex one suffice and I have the focal point on just the combiner?

 

[sophiecentaur]

Or does a planar convex one suffice

It's the net curvature of the sides of the lens that counts. There are different shapes which achieve different effects but the basic theory (relevant to you, at the moment) is that the strength of a simple lens is just measured by its focal length.

So that also answers the question wether my display is the same size as the reflection,

The apparent size of the image and where it appears are two different matters. Take a zoom camera lens. It can produce any apparent size within its spec but it can also focus near and distant objects. There is no simple answer for you but I can suggest that you look at some basic theory. This link shows how to work out the way a basic lens works. If you are happy with and can follow this then you can move on. But I can say that, without some maths, basic optics will always a closed book.

 

[xxtop]

It's the net curvature of the sides of the lens that counts. There are different shapes which achieve different effects but the basic theory (relevant to you, at the moment) is that the strength of a simple lens is just measured by its focal length.

The apparent size of the image and where it appears are two different matters. Take a zoom camera lens. It can produce any apparent size within its spec but it can also focus near and distant objects. There is no simple answer for you but I can suggest that you look at some basic theory. This link shows how to work out the way a basic lens works. If you are happy with and can follow this then you can move on. But I can say that, without some maths, basic optics will always a closed book.

Thank you, part of this is to learn how to figure out the components for such a system. I have no experience with optics and will definitely read up on it before buying any parts. Having resources for this is great and while I don't like math all that much I see why it's necessary :D

 

[sophiecentaur]

and while I don't like math all that much I see why it's necessary

I can sympathise with that. If you want to learn about optics then starting with a complicated project like a HUD is really not practical. When kids are taught about optics they don't start with HUDs. They start with simple systems and work upwards. There's a good reaon for that. ;D
Home optics soon becomes expensive. You could try simulations such as this one. this claims to be free.

 

[xxtop]

Home optics soon becomes expensive

Tell me about it... with these guys I found a company that offers pretty much anything (and online with pricing) but the prices are... startling at times. And judging from all the high-precision laser optics made from exotic materials and coatings, I might not be the target demographic :D

I'll bite the bullet and look into something easier before attempting this, maybe play around with some more basic lenses, mirrors and some focussing. A bit of a random question I came across though: Can you transmit an image via an optic fibre? I'm thinking in theory you should be able to focus a collimated image into a fibre and then do the opposite on the other side. Does this not work because the light is scattered unpredictably throughout the fibre?

 

[berkeman]

Can you transmit an image via an optic fibre? I'm thinking in theory you should be able to focus a collimated image into a fibre and then do the opposite on the other side. Does this not work because the light is scattered unpredictably throughout the fibre?

Do some reading about single-mode fiber optics and multi-mode fiber optics and you'll see why the answer is no for your version of the idea.

However, you can encode the image as digital information and send it through an optical fiber and decode it for display on the far end. That's commonly how images and video are sent (via fiber optic communication channels).

There is also a class of devices that use many parallel fibers to transmit images, but as you can imagine, the logistics of producing many parallel fibers and maintaining the ordering from the input plane to the output plane are quite difficult.

https://www.fujikura.co.jp/eng/products/optical/opticalfibers/07/2050110_12902.html

 

[Drakkith]

I'll bite the bullet and look into something easier before attempting this, maybe play around with some more basic lenses, mirrors and some focussing.

You should note that when I was an undergrad at the University of Arizona in the Optical Engineering program, we had senior students doing this as a senior group project. So it's definitely not an easy DIY project.

 

[xxtop]

You should note that when I was an undergrad at the University of Arizona in the Optical Engineering program, we had senior students doing this as a senior group project. So it's definitely not an easy DIY project.

I have a talent for choosing projects that are above my level of expertise. I will begin an engineering degree soon though so maybe this is something to just postpone

 

[sophiecentaur]

Can you transmit an image via an optic fibre?

You can with a bundle of fibres, once the coherence is sorted out but the resolution is limited by the density of the fibres. Afaiaa, images - as in endoscopy can be much higher resolution using a solid state sensor and transmitting the (TV) image down a wire. The only limit to the resolution with that method is set by the state of the art general purpose imaging (camera phones etc.)

 

[xxtop]

You can with a bundle of fibres, once the coherence is sorted out but the resolution is limited by the density of the fibres. Afaiaa, images - as in endoscopy can be much higher resolution using a solid state sensor and transmitting the (TV) image down a wire. The only limit to the resolution with that method is set by the state of the art general purpose imaging (camera phones etc.)

I was thinking I could move the LCD if I'm able to "send" the light through a tube basically. But building a fibre endoscope too is not what I intended