The Nightsword project

[get-lit]

Here's my solution for an easier to identify "caution" color function with the meters. This is shown with all meters illuminated. However, all meters begin in at the black line at the bottom of the green, and increase upward with gain, for seeing how much you "have". The yellow/red bars below the black line are separate meters for cautionary monitoring of how much you "need". This makes it so that the caution colors are only displayed when there is a "need" condition. As "need" conditions grow, the bars illuminate downward from the black line, into the yellow and then into the red...

Here is display of initial startup, with power near bottom and power caution initiating in the red. As power increases, the green bar grows upward, and the caution bar also diminishes upward toward the black line...

Here is display of normal operation, with no caution colors showing...

Here is display of an an impending air flow fault, the "FAN" meter is extending downward into the yellow and a bit into the red. It is nearing it's maximum capability to power against an air-flow restriction (head wind pressure, dirty filter, blocked intakes). Since it is not yet completely max out, it is still within the threshold of maintaining air flow within bounds, and the air flow meter is not yet beginning to indicate a shortage. It will if the fan is pushed beyond max red...

Here is display of an air flow fault, the "AIR" flow meter is extending downward into the yellow and a bit into the red. Once it hits red, the controller will shut off the lamp to preserve the retro-reflector and lamp seals (BTW, retro-reflector will burn up before the anode seal overheats). The fan had already max-ed to the red just before the air flow began to diminish, the point at which an air-flow restriction (head wind pressure, dirty filter, blocked intakes) began to overtake the fan's maximum ability...

What's nice about this method is that you still get to see what you "have" while also seeing the yellow/red caution colors ONLY while they exist.

EDIT - What's especially helpful with this method is that this accommodates VARYING "caution" conditions. For instance, during startup, air flow requirement is very small, and that requirement itself increases as the lamp powers up. During startup, the green "AIR" flow bar will be at a lower point than when under full lamp power, but still within bounds of its current requirement. So having a stationary marked caution color as with the previous method would be a problem, since it could not handle varying caution conditions like this method does.

 

[London Lad]

Excellent :-)

 

[get-lit]

I've been experimenting with the display and realized something different has to be done here. There's some issues with the display. First off the thumb doesn't like extending to the top outer buttons because the screen is so wide. That in of itself is not a problem because I could move those top buttons inward. In fact all of the buttons are much better off moved inward. I could then move the meters to the outside. However, the overall feel of the thing is just so clunky with the screen being so overly wide for its purpose. The icons are all much too large and the meters are borderline ridiculous, which brings us to the next problem... This thing is too bright!! It washes out your vision in the dark. And come to find out, dimming is not possible with these. So I've been attempting ways to mitigate the brightness. Switching to red doesn't make a whole lot of difference, but using much smaller icons, and making the meters as small as functionally possible went a long way. Being an LCD/TFT screen, it emits a fair amount of light even with an all black screen. Areas of the wide screen that could be otherwise unused only contribute to the luminance even when those pixels are black.

To make this have a better overall look and feel, a less wide screen is needed, and would help with the brightness issue by having less luminous area. The wide screen is 400x240 pixels, and the standard format is 340x240 in the same height. So I experimented with 340x240 screens by letterboxing them into the wide screen display I have to get a feel for how it would work out. Definitely much better. The smaller icons look so much more refined in their simplicity. I removed the circles from the play and Low Power buttons, and the small meters are just perfect. I also widened the thumb slider. Everything looks the part now...

 

[BVH]

The inability to dim the display sure changes things! Would getting rid of a little more of the white text help reduce overall brightness a little more - maybe change the "time" field to a color? When thumb-sliding the focus to minimum, as long as your thumb is firmly on the screen and you swipe your thumb across and beyond the slider vertical border, will that ensure that focus is at max focus or do you sort of have to find the right spot to stop? Will the speed of the focus follow your thumb movement speed or will there be mechanical lag due to motor speed?

 

[London Lad]

How about a toggle switch and a warning light ;-)

 

[HKJ]

Why can you not dim the background light?

 

[get-lit]

Here's what I've found when experimenting with different brightness reduction methods.. The biggest improvement is in reducing the amount of illuminated pixels. Changing color hue away from Cyan/Magenta/Yellow helps, but it is much better to make an object smaller than to reduce it's color brightness, because making it smaller reduces its net luminance while also maintaining pixel luminance for legibility when viewing at the distances we're at. Dimmer pixels tend to get lost in the screen because the background is somewhat luminous. This effect occurs more with indirect viewing angles. It's simply not worth it to dim the pixels when the background doesn't dim with it as well, you just lose contrast.

Our last version we sort of settled on, had 38840 luminous pixels, and the new one has 6680, which is almost 6 times less. That alone pretty much takes care of the entire problem.

How about a toggle switch and a warning light ;-)

That's much too complicated ;-)
EDIT - actually it's more difficult to assemble and wire each light with separate buttons for power, focus, strobe, dim, play, and indicator light(s), etc than a single screen.

Why can you not dim the background light?

This I don't fully understand either. I guess it depends on the display control board or display itself. It needs to be transistor connected to PWM dim the backlight, but most don't. I would have to go with some other product line with another manufacturer. That is the last thing I want to do because this company has all of the development steps clearly outlined, and has excellent development tools, all speeding up development tremendously. Additionally, their products integrate an incredible little video processor that takes the display processing resources off the hands of our primary control board, freeing it up to have reserve for other areas. It also reduces the amount of communication pins consumed of the control board from around 12 to just 2, again allowing for much better future expansion capability with the main control board.

At this point I'm still looking into OLED with touch. I sent an inquiry with this company about any anticipated future OLED touch screen offerings. I've also seen separate touch panels made to be mounted over displays and I may be able to use that over this company's OLED displays.

 

[get-lit]

Currently the largest OLED display is 1.32"Wide by 1.06"high (160x128 pixels). That's small but just large enough for what's needed.

I'm also toying with the idea of not using a digital screen, but rather use a touch digitizer over a screen printed film with button cutouts for light to pass through from an LED backlight.

 

[get-lit]

Forgot...

When thumb-sliding the focus to minimum, as long as your thumb is firmly on the screen and you swipe your thumb across and beyond the slider vertical border, will that ensure that focus is at max focus or do you sort of have to find the right spot to stop? Will the speed of the focus follow your thumb movement speed or will there be mechanical lag due to motor speed?

If the thumb is slid past the sensor area, the focus will go to the maximum position since it would be the last position touched on the sensor area. The speed of the focus will track thumb movement speed. Mechanical lag would not be apparent unless swiping fast, but it would quickly catch up and land at the last position touched. The motor is powerful enough to handle gearing for very fast movement, probably under 1/4 second for full travel. But it can also go as slow as the operator wants to slide the thumb.

 

[HKJ]

This I don't fully understand either. I guess it depends on the display control board or display itself. It needs to be transistor connected to PWM dim the backlight, but most don't.

Doing a pwm on backlight is usual very simple, do it have some other problem?

 

[Patriot]

The motor is powerful enough to handle gearing for very fast movement, probably under 1/4 second for full travel. /QUOTE]


:faint::ooo::drool: I would have never guessed! That's really quite amazing! I guess I was expecting something more along the lines of a Maxabeam. It would figure that Get-lit would be using a far better system.

 

[BVH]

And the Maxabeam focus system is very loosie-goosie due to the large, kind of clunky 4-way rocker switch and there is some type of time delay between moving it and seeing a movement.

 

[get-lit]

I don't have any ambition to out-do or knock the Maxabeam, it's a respectable light having been the staple workhorse for decades. I don't even know anything about it's focus and alignment system, and never held one myself. Wanted to come up with my own ideas.

I have a lot to say about this screen at this point and will I'll be replying about it after I do another test here.

 

[get-lit]

I used a Lux meter to see the numbers behind why I don't like the screen. They're worse than I expected...

->With all red objects, the luminance of the background accounts for 84% of the screen's net luminance. That's 1.2:1 contrast ratio. Terrible.

->With all red objects but colored buttons, the luminance of the background accounts for 73% of the screen's net luminance. That's 1.4:1 contrast ratio. Terrible.

->With all white objects, multi-colored buttons, the luminance of the background accounts for 62% of the screen's net luminance. That's 1.6:1 contrast ratio. Terrible.

The manufacturer has posted multiple times on their forum that the brightness can not be changed. Maybe it can be done with a hardware mod. But even if there's a way to lower the brightness, the horrible contrast ratio would still apply, as the brightness of the active pixels would also reduce.

I measured the on/off contrast, a test with all white screen compared to a test with all black screen, and it came to 42:1. That's pretty bad. That's also at an optimal measure angle for this screen, which is a fairly small viewing cone. Go too far out of the viewing cone and the contrast actually becomes negative!! Yuck.

To get the most out of the screen's limited contrast, we would have to make a screen with half the pixels lit full white and reduce the brightness way down. That means large icons, large lines, and no tiny text, everything that would negate the benefit of the screen's resolution.

The reason the screen has such bad black level is because it's a high resolution for a small backlit screen with full color. Backlit screens have more space between pixels when pixel density is increased and with full color.
The screen has 2.5 times more pixel density than my 1920x1200 LCD computer monitor. I'm guessing this is also the cause of the display's small viewing cone.

Even though smartphones have extremely high pixel density, they are OLED/POLED/AOLED which isn't backlit and doesn't have light passing through their screen door. A high density color LCD is absolutely the worst combination for night viewing.

I tried real hard to like this screen and frankly I'm fed up with it. I would hate to have to look at this while driving. Of course there are decent LCD screens on car stereos, but they are negative mode, very low resolution and usually one color at a time across all pixels, so they can be made with much better black level. My Garmin Edge 800 color screen is ok at night, definitely not great, but it is a very low resolution.

Very high resolution OLED screens for smartphones are dominated by just a few manufacturers, primarily Samsung I believe. Other smartphone manufacturers have to buy Samsung's screens in high numbers through contract. For us little guys, the best we can currently do for OLED is from manufactured by Densitron and Midas. Other companies offer packaged displays from those two manufacturers. Those packages can include a "shield" with embedded video processor to take the video processing load off the main controller board and reduce the number of com pins required of the main control board. For our application, a packaged display with shield is required.

So here's what I'm thinking. Use a transparent glass touch panel, not a display screen but just a transparent touch panel, wired directly to the main controller board to send touch and swipe signals directly to the main controller board independent of any display. It would be a part of the light housing, embedded for the best water-tight seal possible. These are good for millions of hard presses, and have no electronic components. Separating the touch component out give all kinds of flexibility we otherwise just don't have. It also gives the flexibility to use newer displays as needed in the future without making mods to the light housing.

Here's two great screen possibilities behind an embedded transparent glass touch panel...

1. OLED display of our choice. These are simply the best digital displays you can get, black pixels are actually black, zero back-light glow. I'm currently looking more into packaged OLED displays with shields. I wasn't looking before because we needed them with touch screen before, and they don't come that way.

2. Physical display... I could use a screen-printed transparent film with the a back-light. The buttons would emit light through the transparent areas of the film. This is similar to how car dashboards work and is the best looking, most natural, easiest on the eyes, sharpest lines, not formed with pixels. I could use an RGB LED for changing the color and brightness of the display. I could also include a 2-line character OLED display for hour meter & focus angle, menus and functions, feedback, notices, etc. Character displays don't need a shield because they use no resources like graphic displays. This is by far the easiest to do, but would not provide for animated meters. Meters were never a necessity for this project, just a bonus with the digital display. The screen printed transparent film could have the same icon buttons I've been using, for Flash, Low Power, and Power button, square outline for thumb slider, plus and up, down, and enter to cycle through a menu on the character display.

Also, when going with a housing-embedded transparent glass touch panel, the two methods above could be interchanged with one another without changing the housing. I'm very much displeased with this LCD display and would like some feedback on my alternative ideas.

 

[get-lit]

Here's the "Physical" display I mentioned, using a single transparent glass touch panel, over a screen-printed transparent film, over an RGB LED back-light. The transparent screen-printed film would allow light through for touch buttons and slider and would block all light through the un-lit areas unlike LCD/TFT screens. We wouldn't be constrained to the limited availability of graphic OLED displays. An OLED "character" display would be at the bottom center. OLED "character" displays are readily available, easy to use, and don't require a shield like "graphic" displays do. Buttons on left would include, power, flash, low. Buttons on right would include up, enter, down, for cycling through menus on the OLED character display.

 

[BVH]

First off, I love flashy, high performance video displays. I have one of the last Pioneer Plasma Kuro Television sets they made - the 8G. They made one more, the 9G before they chose to stop production of the best black-level Plasma ever made. The demonstrated the fantastic black levels at CES by having the audience enter the room with the TVs on (the audience did not know this) but with no picture what-so-ever. Then they turned off the lighting to a black room. Then they did nothing until the people began to ask when the sets would be turned on. To which they replied that the sets were already on. The black levels are phenomenal on the 9G's you really can't tell their on with no picture. Anyway back on subject. I probably let myself get to deep into the screen on the NS. It's all about what this light will do with its' Lumens and focus. You describe the problems very clearly with the TFT screen. If I understand correctly, option #1 is using a great quality OLED with a touch-sensitive piece of glass over it which is embedded into the case. In this case, you can still have multiple "pages" of displays such as the HOME screen and then the next screen requesting a password whereas this would not be the case with option #2. Animation is still possible with option #1 whereas it is not with option #2.

I would prefer Option #1.

EDIT: I was composing when you posted the above post. So multiple "displays" would be available and presented on the 2-line character display.

I'd still like option #1 but would be pleased with either option 1 or 2.

 

[London Lad]

I love this!

 

[get-lit]

With this setup, I could actually still go with a "graphic" OLED character display with "shield" for the section below the thumb slider, in which case we could still display animated meters if needed. The sole issue with using a "graphic" OLED display for the entire display, not just the section below the slider, is that they aren't available large enough with shields. There's plenty available small enough for the display area below the thumb slider.

Also my thinking is this, why apply so many discreet pixels to areas of the display that don't need to change. It's a waste of resources.

EDIT - There's also a major benefit to having the primary controls remain static, you never lose awareness of the controls, and they're still available for use while navigating menus.

 

[ez78]

I think that glass touch panel setup has the potential to look solid and high quality without the weird effects that LCD panels shows when viewed from the side. And I actually like the simple looks too. Of course an OLED would be nice too if suitable model materializes somewhere.

 

[get-lit]

Won't work.. when cycling the menu, the thumb will cover it.

I've come to realize that overall comfort is more important than just the thumb's comfort on the display. By having a large screen with lots of room to extend the thumb forward, the grip has to be moved back off the center of gravity. I'm finding I prefer to have the grip forward without tons of thumb space on the controls. Therefore a small display just might be best. The largest OLED currently available can be done well with some tweaks.

In the display below, the buttons on the left are FLASH and LOW, and the buttons on the right contained within beam spread slider are POWER and SETTINGS. Those two buttons must be held down for a period of time to activate, otherwise those touches function as a beam spread adjustment.

I could have an info screen pop-up covering the display after a short period of inactivity, the info screen containing all of the meters and current beam spread degrees, run time, etc. When touched, the controls could appear.

The settings menu would then interact as touch screen.

1.32" x 1.06" OLED Display...

EDIT - In my last post I mentioned the benefit of static controls. There is also a benefit to swapping pages of the display between the controls page, the info page, and the settings pages.. there is less light coming at you when only one is displayed at a time, so each could be independently brighter and thus easier to read, than if all were lit at once.

Like or no?