The Nightsword project

[get-lit]

Testing with this enclosure needed to be done before the light housing mold because I needed to optimize the size and shape of the exhaust ducting in order to minimize the air flow required to cool the anode seal. The lower the required air flow, the smaller the large intake ducting needs to be while retaining effective air/water/dust separation. That's why I had to do this before the final mold for the light housing is done.

Also, it was very quiet while testing even at full power. The final light housing will be just as sealed as this enclosure, and with much more rigid walls, with the exhaust and intake ducting blocking much of the fan noise, this will turn out to be fairly quiet after all, practically inaudible from a short distance.

EDIT: The rough drywall surface inside this test enclosure might be helping attenuate the noise. On the electronics housing of the final light housing, I'll apply thin layer of sound absorption material which would be much more effective. It still sounds business up close even though it's quiet, because you can still hear the high RPM of the fan and the whoosh of the exiting air.

 

[Knifefeak]

Awesome light, glad to hear you eyes got better get-lit. Look forward to seeing more of your progress.

 

[get-lit]

Thanks knifefeak.

I've completed temp testing with the test enclosure and found that I will have to go with 33% more air flow than anticipated in order to accommodate the movement of the anode seal away from the optimal cooling point when defocusing the light, so it won't be as quiet as I had described, but still not loud. Now onto finalizing the intake ducting size to accommodate air/water/mist/dust separation with this new air flow.

EDIT: Fairly excited to find a real cheap and easy way to make the many parts I need for the intake and exhaust, a slip roll machine.

 

[get-lit]

This might be somewhat boring to others here, but it's been very interesting learning about the interaction of air, water, and mist with air speed. There's a threshold of maximum air speed at which air begins to draw water through the air/water separation ducting, not good. There's also a threshold of minimum air speed required for fine mist in the air to impinge and attach (entrainment) to the ducting to become extracted from the incoming air. By maintaining an air speed within both these thresholds, and optimizing certain ducting surfaces for air/water separation and optimizing other certain ducting surfaces for air/mist separation, the intake will effectively serve both functions in a fairly simple and easy to make assembly, thanks in part to the slip roll machine. I'm expecting IP66 from any spray angle, and from any operating orientation.

EDIT: I've been plugging away at ironing out these issues before doing the horizontal beam shots and lux measurements because I anticipate those tests making me too excited to remain very patient through these issues in order to resolve them best. It's like forcing yourself to eat the green beans before the steak, otherwise you know you won't finish the green beans. Sure everything will be encapsulated and water, mist, and dust won't be detrimental, but eliminating the possibility of them entering the light would allow the use of high reflectance aluminum coated optics for 20-30% more output and full lamp durability in all weather conditions.

 

[BVH]

I know exactly what you mean about the excitement of doing the horizontal shots. Here I've had the 800 Watt Trakkabeam and the 1,600 Watt NightSun for how many months now and I haven't done any meaningful beamshots. But I'm excited when thinking about doing them. The two big issues for me are:

1. Will my 4.5KW (4.0KW continuous) 120/240 generator provide enough power to power the Realtec 100 Amp DC rectifier with the starting load of the NightSun.
2. The noise and brilliant light beam factors being generated in a close by, new neighborhood.

The max amp draw of the Realtech is 21 @ 240 Volts input. That's for the full 29 VDC/100 Amp output. When running, the NS is pulling 62 Amps @ 28 VDC according to the analog ammeter on the Realtech so lets say about 65% of capacity. So 21 x's 65% = 13.65 Amps @ 240V to provide the light with 62 Amps @28 VDC. The generator can sustain 16.6 Amps @ 240V. So the realtec should easily handle the running Amps. The max starting load for the generator is 18.75 Amps @ 240V. I am guessing that the starting current of the light is going to impose a 240V load higher than that so I'm hesitant in even trying it. However, when plugged into 240V house current, the light starts just fine so the starting current must impose a 240V load of less than 21 Amps. So it's close. That's the dilemma.

I'm also concerned about neighbors calling the police. While it might be perfectly legal to shine the light out over my test range (miles of rolling hills ((private land)) with no dwellings or risk of lighting up vehicle or aircraft traffic) - it also might NOT be legal. I don't know.

 

[get-lit]

Wow, the Nighsun power supply is over 92% efficient, assuming the Nightsun is running the lamp at 1600 Watts. That's really good.

As far as I've been able to find as of yet, there are no laws within the US regarding searchlights, other than FAA guidelines for searchlight operators to notify nearby airports of times and location of operation. There may be some municipalities that have their own local laws, possibly NYC. It would be nice to know where they're locally regulated. Being mindful is the only way to keep them legal in your area and across the board. What's nice about the Nightsword is complete lack of spill light and corona, so no errant light to be bothersome.

 

[BVH]

Not only no significant spill but I can run it from 120VAC using my new Yahama 2KW inverter Genset which is nice a quiet and not going to break my back. Whole setup should not even be noticed by anyone near my range.

Went and checked those power readings and this is what I said:

"According to two ammeter sources, power draw is between 60 Amps (meter on the Lorain PS) and 64 Amps (Estech clamp-on meter on 400 Amp scale). The cooling fan consumes 1.4 Amps in that total. So let's be conservative and say the light minus fan is pulling 58.5 Amps @ 27.5 Volts so total power consumption of the lamp is around 1608 Watts which is right on the 1600 Watt money."

The caveats are that I was reading voltage at the junction box output, not the lamp and same with the Amps, not at the lamp. According to the manual, once the start and boost circuits are done, it's "system" Voltage minus the Diode loss which is present at the lamp. I now have a new 1,10, 100 Amp milliamp resolution clamp-on meter which will give me a more accurate reading. When I get the new lamp, I will make some temp connections and measure Voltage and current at the lamp to nail it down.

 

[get-lit]

I doubt the SX16 power supply could be 99.5% efficient, I'm thinking the lamp is operating at less than 1600W.

 

[get-lit]

The ducting baffles for the intake would amount to a net weight of 1 pound. That's about the lightest I can go. I hope that's not considered too much of a weight penalty for IP66. It will be at the rear of the light helping offset the weight of the lens and reflector.

 

[BVH]

I don't remember if you given an estimated weight of the light at its' latest design?

 

[get-lit]

Net weight for likely IP66 for dust/water/mist is 10.88 lbs.
Net weight for likely IP64 for dust/water is 10.15 lbs.

So about 7% more weight for the added ingress and mist protection.
I'm for this because there's a lot of mist here at times, and also I want it to be sea worthy.

EDIT: 9.88 lbs with IP42, basically just against light rain under a limited range of orientation.

 

[BVH]

In the overall scheme of things, the extra .75 pounds is not a big deal as I see it especially given the benefits of mist reduction/removal and better balance. I live on the left coast, 4000 feet from the waters' edge and on many if not most "clear" nights, if I light up a short arc, there's easily visible water droplets/mist present even though you can't naturally see it or feel it nor detect high humidity. In the end, it will weigh just about what a Costco HID, and probably a Thor 15mcp weighs - about 11 lbs.

 

[get-lit]

While saving up for funds for the slip roll machine, I'm working on programming the controller and came across another solution to simplify the control buttons. I had planned in using four force-sensing resistors embedded into the top of the handle, for durable, space saving, waterproof control buttons...

These get expensive and each need a voltage divider circuit and each need three wires. I came across a circular positional feedback sensor, like the round sensor on the Apple iPod. This would simplify the construction, reduce the number of components to just one control sensor and one voltage divider circuit. I would them program it to apply the top position as a focus button, bottom as defocus, left at strobe, right at dim, and various sliding movements for other functions, such as initiate the function to set the focus position, or a function to assign a new turn-on sequence, like a password reset function, etc.

I could also go with a single linear positional feedback sensor, and use linear sliding touch to move the focus, and then taps at each end for the other functions...

I'm leaning toward the linear sensor.

EDIT: Correction... iPod uses a capacitive sensor, whereas these are resistive sensors. A capacitive sensor only works when touched by something that carries electricity... fingers, not gloves. Instead, these resistive sensors rely on a slight pressure, so they can be used with gloves.

 

[get-lit]

Getting past the headache/learning curve with controller programming and it's getting fun now. The force-sensing resistors I have are working incredibly well and I'm looking forward to working with the linear sensor to replace them.

The display will have various output modes:

-Wattage (fun for startup)
-Lamp Hours
-System Hours
-Current Run Time
-Adjust Focus End-Stop Position
-Current Beam Spread
-Set Power On Sequence (Password)
-% Fan Power (helpful for knowing when approaching headwind pressure limit)
-Serial Number

Startup Screen to say: NIGHTSWORD - CPF EDITION

Different colors can be assigned to the various modes, for instance red for cooling airflow alert, in the event of fan failure, excessive headwind pressure, or air filter clog.

TTL input/ouput could be later added after assembly to fully remote control the light.
The controller chip is about the size of a quarter and can be removed for mailing to have any future updates loaded.

EDIT: The display is 2.6" x .63". I was originally planning to locate the display on the side of the light, but I'm trying to think of the best place where it can be visible when holding the light.

I'm thinking of using the linear sensor in the following manner during operation:

Slide = Adjust Beam Spread
Press @ Top = Strobe
Press @ Bottom = Low Power
Press @ Center = Mode Cycle

Once in "Adjust Focus End-Stop Position" mode, Slide action would then apply to setting the end stop position, with the focus motor moving very slow in relation to slide movement during adjustment.

Once in "Set Power On Sequence" mode, a determined number of presses and slides are recorded and must be re-entered to confirm as new password. I'm thinking six actions would be sufficient (an action being a press or a slide). Maybe allow the entry of any number of actions, from one to ten actions, so the user can determine their own level of safety, from no safety to fairly difficult.

 

[BVH]

This fabulous controls system is going to be icing on the cake! This will be the light of a lifetime!

 

[get-lit]

That's the whole idea! I'm going all out on the programming front because I program for a living and this is the first time I've done programming for a hands-on project rather than network communications and databases which remain entirely digital. Turning logic into actual physical results is very exciting for me. I've outlined the programming phases to be completed and can't wait to tackle each one day by day.

 

[get-lit]

After considering display locations, I think the best location is directly above the linear "thumb control" sensor. The display I had in mind would not fit in this location, so I'm going with a 1.8" 160x128 pixel high-contrast color TFT display. I could just do a touch screen instead of the linear sensor, but the touch screen would be too wide to rest the thumb to the side when not in use.

EDIT: Will be going with an OLED display for higher range of temps, contrast, true black, efficiency, etc. Much better viewing angle is important since it would be located at a 45-degree angle to the handle, toward the top of the optics housing.

Among the currently available OLEDs, the most suitable is white rather than color because it's available with it's own buffer and preloaded fonts, reducing the controller's RAM and processing utilization 100 fold. Also just two I2C communication pins are needed instead of five SPI pins.

A Youtube video of the display in someone else's project (poor video focus)...


Here's a color OLED in someone else's project, but currently not available with it's own buffer and fonts, and needs 5 SPI pins...


I might be able to possibly wing the color OLED, but it would consume most of the RAM and com pins, leaving no headroom for furture expandability for things like remote control etc.



EDIT: I've found a touchscreen OLED that is small enough to fit under the thumb, and narrow enough to rest the thumb to the side in portrait orientation (<1" Wide). It's not in use in other DIY projects so there would be more leg work involved. I have to see if it can be programmed for portrait orientation without too much headache, but this is the direction I'd like to go...

 

[TEEJ]

Very cool!


The screen going dark between the cold, normal and hot display is probably something I'd want some sort of transition between instead of a blank screen though.


The analog with moving bar and a digital readout format is probably better over all...such as the exhaust temp graphic.

 

[get-lit]

I agree with those suggestions. Keep in mind those were videos of displays in someone else's projects, shown to illustrate the displays, not necessarily representative of how the displays would function on this project.

I inquired with the company with the touchscreen but haven't heard back. The touchscreen function may take up too many com pins on the controller board, and I may have to resort to the separate linear sensor button which only uses one com pin.

Either way would be fine. The separate linear sensor button has the advantage of always seeing the screen while the thumb is over the button. The touchscreen has the advantage of the areas to be pressed being separately illuminated, whereas there's no way to visually identify the exact separate segments of the linear sensor in a precise manner. However, I could add three small SMD LEDs under the linear sensor, each centered under a button press function, in a tri-illuminated manner. Something would have to be done to make the buttons easy to identify at night, since it's the only time the light would be used anyhow.

Are there any preferences for the single touchscreen under the thumb, or the tri-illuminated linear sensor button under the thumb with separate screen above?

EDIT: Here's drawings of the two control concepts:

Now I kind of prefer the separate control sensor. With the buttons separated from the display, various data can be displayed on the screen rather than the buttons.

 

[BVH]

I didn't want to appear pushy so didn't say anything before but since you're asking - Touchscreen, Touchscreen and Touchscreen.