Hello all! As the title says, I am designing a method of controlling tens of thousands (up to several millions) of heating elements that does not use traditional wireless communications protocols nor does it use a microcontroller of any kind. I am trying to avoid adding a wireless receiver to each circuit (i.e. each heating element) to keep costs down (same logic for avoiding the microcontroller). The individual circuits will be identical. They will be installed on devices in an open field. Each one will be connected to a small solar panel, no more than 5-10W. The solar panel will act as both the control input and power source. The circuits may actually control up to 3 heating elements per circuit, but for simplicity, let's consider that they only control 1 for now.

So how will I control each of these? In the center of this field, I will install a tower of sufficient height to maintain line of sight with each device. On the tower, I will install one or more high-powered lasers. I've already designed this for a variety of different wavelengths. To act as the controller, the laser will be very briefly aimed at an individual solar panel. This spike relative to the normal solar irradiance acts as the control input signal.

Ok, back to the circuit. Each circuit will consist of a solar panel connected to both a battery and a solid-state relay as the switch. The battery is connected to the heating element through the solid-state relay switch then back to the battery. I considered using a capacitor instead of a battery, due to their ability to discharge more rapidly without damage, but they are more expensive. The element should be capable of being heated once every 20 seconds (up to about 45C). The solar panel keeps the battery charged. Once the laser is aimed at the panel, it activates the relay which completes the heating element circuit. Does this sound reasonable? I'm a chemical engineer with no circuits background, so you'll have to excuse any errors I've made.

The other thing to consider is that the laser is only very briefly aimed at the panel, maybe a tenth of a second or a twentieth of a second. Therefore, I expect the relay might close too soon, and the battery will not be able to heat the element as the circuit was not closed long enough. Is there a cheap way to keep the circuit closed long enough to heat the element but not damage the battery?

Any help is greatly appreciated.

 

Let's see, up to1M devices each individually controlled by a .1sec pulse. 100,000 seconds (27 hours) to visit them all with a single laser. Or do you have 1M (or other large number of) lasers as well? Unless you have 1 per device, there is also time involved in re-aiming them.

This whole thing sounds rather ill-conceived to me.

What are you trying to accomplish? Maybe there is a better way...

 

The other thing to consider is that the laser is only very briefly aimed at the panel, maybe a tenth of a second or a twentieth of a second. Therefore, I expect the relay might close too soon, and the battery will not be able to heat the element as the circuit was not closed long enough. Is there a cheap way to keep the circuit closed long enough to heat the element but not damage the battery?

You need something called a latch. One brief input triggers the switching, then the circuit holds that state until something else triggers it to stop. I'm guessing in your case that might be a timer set to run a few seconds. A latching relay is a thing, but you don't want it to stay on forever once triggered, so you need a way to return the relay to "off".

 

if you went with radio, you could broadcast, group, or activate separately. At a million pieces, you should be able to get a custom ic containing radio, processor, battery charger, and switching element in one piece for a reasonable price, and increased flexibility. Embed it into the solar panel for a broader application range. How much does a tower, laser, and associated design cost? I suspect the panels will be facing the sun, meaning looking away from field center for a portion.

 

Let's see, up to1M devices each individually controlled by a .1sec pulse. 100,000 seconds (27 hours) to visit them all with a single laser. Or do you have 1M (or other large number of) lasers as well? Unless you have 1 per device, there is also time involved in re-aiming them.

This whole thing sounds rather ill-conceived to me.

What are you trying to accomplish? Maybe there is a better way...

Thanks for the response. Of course, a single tower or a single laser on a single tower will only be able to control up to a fixed number of these devices. Then I'll have to add more lasers or more towers. I'll give a better description with pictures in about a week or so--after I've had time to 3d print and assemble the prototype.

 

You need something called a latch. One brief input triggers the switching, then the circuit holds that state until something else triggers it to stop. I'm guessing in your case that might be a timer set to run a few seconds. A latching relay is a thing, but you don't want it to stay on forever once triggered, so you need a way to return the relay to "off".

Thank you. I read about latching relays and had the concern that you mention needing a second control signal. I would rather it be time-delay mechanism or, alternatively, a mechanism that activates when the element achieves the goal temperature.

 

if you went with radio, you could broadcast, group, or activate separately. At a million pieces, you should be able to get a custom ic containing radio, processor, battery charger, and switching element in one piece for a reasonable price, and increased flexibility. Embed it into the solar panel for a broader application range. How much does a tower, laser, and associated design cost? I suspect the panels will be facing the sun, meaning looking away from field center for a portion.

The tower has to be installed even with a radio. The laser and associated components will be $15,000 at most and more likely less than $10,000. Any idea on what the per unit cost would be for a system like you described? Consider that I'll be producing or procuring millions of these per year in the most optimal case, driving down the cost of production. Of course, production will start around 100,000, but I expect to ramp up quickly.

Edit: I've already designed a laser transmitter based on two types of lasers, a 100 W CO2 laser and a 95 W diode laser. They both have pros and cons. Ultimately, I may look at fiber lasers due to their ability to have a tighter beam, but they're so expensive for just making a prototype.

 

I've already designed a laser transmitter based on two types of lasers, a 100 W CO2 laser and a 95 W diode laser. They both have pros and cons.

I've seen this episode!

 

Edit: I've already designed a laser transmitter based on two types of lasers, a 100 W CO2 laser and a 95 W diode laser. They both have pros and cons. Ultimately, I may look at fiber lasers due to their ability to have a tighter beam, but they're so expensive for just making a prototype.

Why would the laser need that kind of power?

 

Why would the laser need that kind of power?

To make up for the distance between each target and the laser (divergence causes significant reduction in W/area--I can focus it only so much), to sufficiently distinguish itself from normal solar irradiance, and because I thought it was cool. It sounds inherently unsafe, but I have some safety features baked in that prevent it from triggering if there is an obstacle between the laser and the target.

 

To make up for the distance between each target and the laser (divergence causes significant reduction in W/area--I can focus it only so much), to sufficiently distinguish itself from normal solar irradiance, and because I thought it was cool. It sounds inherently unsafe, but I have some safety features baked in that prevent it from triggering if there is an obstacle between the laser and the target.

Also, another reason is that I can intentionally spread the beam to trigger multiple elements at the same time. Think spreading the laser horizontally or vertically as it pans over an area.

 

Is your laser signal modulated, like a TV remote? You can get a lot more distance with a lot less power that way. Think of it as sending information rather than power.

 

Is your laser signal modulated, like a TV remote? You can get a lot more distance with a lot less power that way. Think of it as sending information rather than power.

This has been in the back of my mind, but I have not seriously considered it yet. If the PVs act as sensors in this case, the range will need to be above the maximum solar irradiance if I plan on controlling them during the day. Also, wouldn't this require a microcontroller for each element? And I'm not sure I understand if there would be a benefit by doing this. The laser would still only transmit a single amount of power "signal" to a single panel or set of panels.

 

but I have some safety features baked in...

That's exactly what McKay said!

 

Is your laser signal modulated, like a TV remote? You can get a lot more distance with a lot less power that way. Think of it as sending information rather than power.

Exactly, it never even occurred to me that he meant to use the solar panels as the detector and illuminate them with more radiation than than sunlight. I am beginning to think my original characterization as ill-conceived was too kind.

 

Welcome to AAC.

It is probably important to consider that using light to control things is a traditional wireless method. Because of this, many of the problems of light-based control have solutions. Using the PV cell as the receiver is probably not going to end well for a variety of reasons, including slow response times and dynamic range that might not let the panel "see" the laser in full sun no matter how powerful you make it.

It is also misguided to avoid the use of microcontrollers because at the scale you are discussing they will cost, literally cents, provide all sorts of functionality that would otherwise require a considerable part count, and offer the opportunity to make updates and upgrades to your devices to cover the inevitable unexpected problems with your design.

Radio is also a likely win over a dangerously powerful laser on a large tower. The economy of scale you'd have available with so many of these devices deployed would make the radio chip you need very, very cheap. And, with technologies like self-forming mesh networks and agile protocols you wouldn't need any sort of large tower. You also get the bonus of the whole thing being powered by the PV cell without needing a very energy-expensive and dangerous laser.

Not to be presumptuous, and will all due respect, if the reason you are avoiding MCUs and radio is that you don't know how to use them, you aren't the right person to be designing and building this system. If you have a useful vision, it may be most sensible to work with one more engineers with experience in these technologies instead of asking "randos on the Internet" about how to implement your naïve solution.

If I am mistaken about this last part, and you have a good rationale for your (apparently) poor design, I apologize—but I don't see it. If you don't even know what a bistable multivibrator is, you probably need a lot of help. This really isn't a bad thing. Many times a person with an excellent idea doesn't have the domain specific knowledge to execute the details. There's no shame in being a person who organizes an innovative project and has others contribute.

[edited for typos, not content]

 

At the item volumes you are suggesting, radio communication to a microcontroller should be very cost-effective.

 

Thank you Ya’akov for refining what I simply called ill-conceived. I am also struck by the fact that one of the reasons he wants to use a high power laser is because “I thought it was cool.”. This sounds more like a teenage boy’s fantasy than an engineering project.

 

The element should be capable of being heated once every 20 seconds (up to about 45C).

Does that mean an element can also cool down in less than 20 seconds?
Is there any requirement as to when a particular element gets heated? For example, are the elements to be used in some sort of phased array?
Are you going to tell us what this vast number of heating elements is for? Keeping an Antarctic penguin colony warm perhaps?

 

Also, wouldn't this require a microcontroller for each element? And I'm not sure I understand if there would be a benefit by doing this.

No, the modulated signal can be detected with discrete components although a µC would allow more versatility (such as different commands, again just like a TV remote). There are scores of receiving/transmitting pairs available commercially that have already worked out many of the details (wavelength matching, etc.). I'm not sure if you could leverage a commercial product, though, because you already have the laser transmitter and the PV panel receiver. You might have to build your own detector circuit but certainly you want to follow well-established techniques. No need to reinvent the wheel.

The benefit is a vastly increased signal to noise ratio, the ability to pick out a weak signal at a greater distance. "Seeing" the laser while under full sun could be quite challenging otherwise. But if it's modulated at 50kHz, it's far easier to pick out of the background.