The Nightsword project

[get-lit]

BVH's recent Megaray acquisition got me thinking about ellipsoidal-based lights and their incredible flood capability, and I wanted to see what it would take to achieve the same beam power as the current long FL parabolic design. Over the years I've seen comments and articles claiming that the parabolic has the most powerful beam for a given size optical system, and I could understand how due to it's efficient use of space. To be sure, I modeled a retro-reflected ellipsoidal-based optical system with equivalent beam power as the retro-reflected long-FL parabolic, and found this to be true. Although the ellipsoidal is somewhat narrower (9" vs 11"), it's overall size is much larger with about 3-4x the length, so I'm confident in proceeding with the current design with more limited flood use...

[FRITZHID]

nice work!

[jmpaul320]

i want one

edit: i would just like to also add that i appreciate all the work you are putting into this... wish i knew half of what you did! lol

[Solscud007]

Sorry for not reading everything here, But why not have the housing 3D printed? There are materials that are strong and non-metallic.

[get-lit]

Thanks! This will really help with a few small internal parts and bezel etc. Probably stick with layered composites for the housing for the most lightweight strength.

[get-lit]

Tremendous success today. Had a mandrel made by Phoenix Electroformed for the retro-reflector and received two retro-reflectors for testing. On the first test, the coating burned up in about a minute because I was in a hurry and I passed the airflow perpendicular to the lamp and it was not entering the retro-reflector effectively. For the second test with the other retro-reflector, I passed the air inline with the lamp, forcing it to enter directly into the retro-reflector just as the final housing would, and it ran for an hour just fine. Lux from the lamp increased by 31.6%. I expect that if the retro-reflector were to be coated with an additional enhanced coating, that figure would just about a 38% increase in lamp lux. This is ever so slightly under my predicted outcome but a tremendous success nonetheless. After all losses, this will place a net lumen output just over 40k within the beam, at 70 MCP within the entire diameter without any corona whatsoever.

I'm very much impressed with the quality of the retro-reflector made by Phoenix Electroformed. I provided a precise drawing and got it just as requested. I'm also very pleased with how well my design worked out. It's small, very light-weight single-piece without an intricate mounting method, no blocking of light to the primary reflector, re-useable, and very simple to align, and it solves the problem of stray light damaging your eyes. It's nice when things go just right. I'm extremely happy with this.

[ma_sha1]

I wonder if they will make a replica of Firefox III HID flashlight reflector (2.5") in SMO, how much is the cost range?
I wonder if I can convince Lip to organize a group buy, he should be able to get the reflector drawing from the mfg.

[get-lit]

It would not be a problem for Phoenix Electroformed. That's a bit larger than my retro-reflector, but mine had some extra things going on and some machining, so I couldn't estimate a cost for the mandrel. You may be able to send them a unit and have them get the dimensions. There's a good chance it's a common size that Phoenix already has.

[BVH]

Excellent report, Get! Glad to see it is going well!

[BVH]

Hi Get-Lit. Anything new?

[get-lit]

The design is complete but I have to get caught up with finances and other responsibilities for a while before I can handle the costs for production.

[BVH]

10-4!

[BVH]

Hi Get-Lit. Just checking in for any updates?

In your post #121, one shot shows the Anode located rearward relative to the front of the light. The other shows it forward. My understanding from the guys at A.R.C. is that the Anode should be favoring the "up" position during use of the light so as not to have the flame blow back against the thin Cathode. I understand why the airborn lights I am familiar with have the Anode in the back - because they are designed to be used pointing down - where the Anode is "up. Will your light have the Anode in front so that as the light is pointed between horizontal and upward, the flame will be directed towards it?

[get-lit]

Hi BVH. Having to wait for tax returns to proceed. Lamps for handheld use should really be chosen based on universal orientation, and the lamp I'm using has universal orientation.

[shuen]

Nice and so cool !

[get-lit]

While working on funds, I've made some slight improvements to the housing design according to recent feedback here and what I've wanted to try and implement before anyhow.

Also in the interim I'm needing to come up with a fairly complex solution for a cooling fan controller. One of the problems I've experienced is the lamp can be over cooled, because it's so much more efficient than a short-arc Xenon lamp. This makes it about impossible to keep the lamp at optimal operating temp while using a constant fan speed in different climates as you could with short-arc Xenon lamps. So for use in different climates, a fan speed controller with a thermistor probe is required. So far that's the easy part. There's schematics for simple little DIY controllers to do this, however there's several additional requirements that I'm not sure how to accomplish, and they all have to work in conjunction with each other.

When powering down, the lamp envelope heats up fast, and much hotter than when the lamp is on. I'm not sure why this happens, but there's a ton of heat energy released when the lamp powers off. Maybe it's the noble gasses releasing all their acquired heat energy as they drop back down to solid state from gas state. It does take minutes to acquire that energy, so it makes sense that there's a lot of stored energy to be released. It's a general precaution with short arc lamps and most other HID type lamps to run the fan for a minute after powering down.

So the second consideration for a fan controller is that it must run for a minute after powering down, which means not only do I need a timing circuit, but that the fan controller needs to be powered separate from the lamp on/off switch.

But if the fan is powered separate from the lamp... after the timer has shut the fan off, how would it know that the lamp is turned on again. So the fan controller needs a function to handle a separate "power on" input signal.

Maybe I could instead just use two separate controllers, both wired to the fan. The one that controls fan speed via temp probe could be powered by the lamp power switch. The other, powered separately, would monitor when the lamp is powered down and blast the fan at full power for a minute. This is what I'm looking for some direction with at the moment. Hmm, maybe just a large triggered capacitor that stores enough energy to power the fan for a minute, about 1/4 WH. Just a thought, but still looking.

[langham]

Just put the fan's switch in series with the light so that it has to be on in order for the light to be turned on and use a temperature sensing circuit as before mentioned that will allow the light to be automatically cooled. How about an led indicator that will tell you when the light no longer requires cooling? When it lights up simply turn off the main battery disconnect and save power. Anyone who gets one of these is not going to be an amateur. I would recommend the ellipsoidal design myself as it seems to be the best way to have a nice transition to spot and flood, but it will cause a sacrifice in lumen output, which seems like the opposite of what you are going for.

[get-lit]

I know manual fan control is how it's been done for decades in theaters but it just doesn't appeal to me. Today's newer stuff has automatic fan controllers. Even my boat has an automatic bilge blower with a five minute delayed shut off. Just a bit of extra work in the beginning is a fare trade for not having to deal with it anymore.

I'd given the optic method of choice extensive consideration over the years and I have no doubt I'll be sticking with the parabolic reflector because it packs the most beam punch for a given overall size. I describe in my first post on this page the reason for my choice over the ellipsoidal method for this project. I plan to go with either the ellipsoidal or short-FL parabolic for a more practical light with better flood pattern on a future project once this is completed.

[HKJ]

So the second consideration for a fan controller is that it must run for a minute after powering down, which means not only do I need a timing circuit, but that the fan controller needs to be powered separate from the lamp on/off switch.

But if the fan is powered separate from the lamp... after the timer has shut the fan off, how would it know that the lamp is turned on again. So the fan controller needs a function to handle a separate "power on" input signal.

Use a diode from after the lamp power switch to the fan controller and then a transistor to directly supply the fan controller from the power. When the fan controller get power via the diode it will turn on the transistor, then it will have power until it turns off the transistor, even if the main switch is turned off.
The transistor could be controlled from the lamp temperature or from a timer.

[get-lit]

Thank you HKJ, this is very helpful. Let me see if I understand correctly by re-phrasing by how I understand this...

When the lamp is powered on, the signal from the lamp's power switch is also connected through a diode to a transistor that is also independently connected to the power supply, which then powers the fan controller even after the lamp power is turned off, and can be signaled to turn off via separate signal such as lamp temperature or a timer?

If this is correct, my question would then be... How would the shut off signal (temp or timer) then signal the transistor to turn off the fan controller "only after" the lamp is powered off and not at some point while the lamp is still on?

Would the transistor turn off the fan controller only when "both" the lamp signal is off and the shut off signal (temp or timer) is triggered?

How would a timer initiate only after the lamp power is turned off?

[langham]

I think that would work for a temperature circuit, but if you wanted a timer then it would have to be initiated separately when you turn the power off.

[HKJ]

When the lamp is powered on, the signal from the lamp's power switch is also connected through a diode to a transistor that is also independently connected to the power supply, which then powers the fan controller even after the lamp power is turned off, and can be signaled to turn off via separate signal such as lamp temperature or a timer?

That is basically correct.
A diode is a one way street, i.e. it allows power to flow from the lamp circuit to the temperature controller, but not the other way, this is important when we wish to power the temperature controller with the lamp off. The transistor is an electronic switch that is turned on as long as it gets a signal.
Letting the temperature control the transistor (From a separate minimum temperature limit), it can supply a signal, until the lamp is cool enough, then it removes the signal, the transistor will turn off and the power will be removed from the temperature controller.

If this is correct, my question would then be... How would the shut off signal (temp or timer) then signal the transistor to turn off the fan controller "only after" the lamp is powered off and not at some point while the lamp is still on?

As long as the power is on, it does not matter if the transistor is on or off, because the temperature controller get power via the diode.

Would the transistor turn off the fan controller only when "both" the lamp signal is off and the shut off signal (temp or timer) is triggered?

Again, the temperature controller is powered from the diode, when the lamp is on, only when the lamp is off does it need the transistor to the keep it power on.

How would a timer initiate only after the lamp power is turned off?

Using the power to the lamp as a trigger, a timer could easily be made with an extra diode an a capacitor.

There is a few practical details, like voltage, current, time period and does the temperature controller voltage need to turn off at once, or will it work if the voltage fades slowly, when these questions are answered the circuit can be made for less than $5.

[get-lit]

Greatly appreciated! Everything is 12v, including the lamp switch. There's several temperature based fan controller circuits I came across, here's one I'm looking at now...
http://electronics-diy.com/electroni...ic.php?id=1036

There would need to be a way to initially adjust how much power to apply to the fan relative to temperature.

Due to the lamp's high ignition voltage, the temperature probe must be located away from the lamp, in the air flow past the lamp. Therefore, the fan must always be running in order to enure that heat from the lamp always reaches the temperature probe.

Turning off the fan after the lamp is powered down would probably be better based on a temperature baseline rather than timing.

[HKJ]

Something like this will work (It is supposed to work together with the thermostat you linked to):



The drawing is untested and might need a few additions, like a capacitor accross the supply and a resistor from pin 1 to 3 on the IC (This will give it hysteresis, i.e. it does not turn both on and off at the same temperature).
The potmeter is used to adjust the temperature.

[get-lit]

Thanks again HKJ. As a programmer I see these components as programming operators. I'd like to stay focused on the other components of the light and it would help if I could pay someone to build these fan controllers as simple little modules I could just wire in. I could wing $500 for a batch to make it worth someone's while. I don't know if you'd be up to this or know of someone who would.

[HKJ]

Thanks again HKJ. As a programmer I see these components as programming operators. I'd like to stay focused on the other components of the light and it would help if I could pay someone to build these fan controllers as simple little modules I could just wire in. I could wing $500 for a batch to make it worth someone's while. I don't know if you'd be up to this or know of someone who would.

No and no.
As a programmer you could also use a small microcontroller for it.
Connect the microcontroller directly to the battery, use a ADC input to measure the temperature, use a digital input for lamp on/off and use a digital output for controlling the fan with pwm for speed control. If you want the possibility to adjust the temperature, use one more ADC input for a trimpot.

[nealitc]

Sounds like a perfect Arduino project, you could also use it to control other functions, indicate low or high input voltage, monitor current, indicate overtemp...

[get-lit]

Would it be much more difficult to make such a controller turn off the lamp in the event of over temp, which could be caused by a fan failure or filter clog? That would prevent a catastrophic event.

[HKJ]

Would it be much more difficult to make such a controller turn off the lamp in the event of over temp, which could be caused by a fan failure or filter clog? That would prevent a catastrophic event.

Not much, the power switch must be replaced/supplemented with a electronic switch (this is just a power mos transistor).

[get-lit]

Perfect!

The Arduino looks interesting. Arduino Mini is tiny but takes only 9v. The Arduino Micro is larger but may be overkill.