Jason Levin, CEO and founder of UpToke, talks about Spyre, a vaporizer that packs a (technical) punch. He discusses how they developed a custom PCB, precision pressure sensor, and more to create a versatile and intuitive electronic cigar.
UpToke’s story starts in Florida in 2009. I was living there at the time, and witnessed e-cigarettes (battery-powered vaporizers) become popular that year. There was a quick influx of vendors, and I couldn’t help but notice that consumers who bought e-cigarettes had a low level of confidence in the quality of the brands that sold them. All of the models were imported from Asia, and they had the same basic function and construction. The only differentiator was the branded wrapper, and it was easy to sell cheap knockoffs.
Meanwhile, the cannabis industry in California was gaining steam, and when I moved there in 2010, I realized that cannabis and e-cigarettes would soon overlap in a big way.
My team and I set out to create a vaporizer that would improve on e-cigarette offerings. We wanted to build a hearty, versatile, and high-quality branded device that would support many materials, from solid plant leaves like tobacco and cannabis and chamomile to chemicals like caffeine, melatonin, and herbal oil blends. We also wanted it to mimic the intuitive experience of smoking a cigarette or cigar, but spare our users the unclean waste materials that often come with smoking.
We ended up with a smart device loaded with sensors, an intelligent power delivery system, and the fastest heater on the market (by a large margin). This article shares the development of our prototype process and the technical challenges we faced during our two and a half years prototyping. We are proud to introduce our commercially distributed model, the Spyre.
Meet The Spyre Vaporizer
To start with, what is a vaporizer? The vaporization process gently heats materials to release active chemicals, but never hits combustion temperatures. Active compounds boil into the air and can be inhaled without smoke, ash, tar, or carcinogens.
The Spyre is our current product. We compare the form factor to a cigar. It is heftier than the e-cigarette or the slightly larger vaporizer pen, and aims to replicate the experience of smoking a cigar—intuitive interaction, easy to pass on to a friend when smoking socially, and entirely portable—but with a cleaner experience than a traditional burning cigar. We used medical grade materials and custom-built technology for the highest-quality experience possible.
We went through several product iterations. Our R&D kicked off with a 3D-printed proof of concept. Then, we designed version 1.0 of our product, internally named Churchill (like a Churchill cigar). This helped us raise angel financing, but we knew we needed to improve on it—Churchill was large and the industrial design wasn’t high caliber.
We started working on version 1.1, the Spyre, by bringing in industrial designers to help with the form factor and user interface. Meanwhile, we changed the airflow, circuit board orientations, and other internal electronics, like the pressure sensor and battery protection (more on this below). Each decision we made was a delicate balance of aesthetics and functionality. Overall, the upgrade from version 1.0 to version 1.1 brought the product up to commercial standards: becoming safer, cleaner, with user friendly differentiators, and easier to produce.
We’re currently working with manufacturers like Sierra Circuits and a major distributor to bring the Spyre into mass production. We continue to iterate on the boards and the sophistication of our sensor system, and are proud to innovate in electrical engineering to create an intuitive and delightful experience for our users.
The Most Advanced Vaporizer On The Market
The UpToke Spyre packs a lot of technology into a small space. Here are some of the features that set it apart from competitors.
CPU-regulated Intelligent Power Delivery System
When you’re not actively drawing breath from a vaporizer, it shouldn’t heat, because this would waste both material and the vaporizer’s battery. So we built the Spyre with automatic power adjustment features that rely on sensors to detect your inhalation speed and measure the existing temperature in the heating chamber. The CPU uses data from these sensors to dynamically regulate energy delivery, scaling the power up and down as needed. This provides all-day battery life and enables rapid heating.
Rapid Heat Response With Thermal Air Path
The Spyre uses a proprietary bulb heater which consists of a tungsten filament encased in quartz glass. The quartz buffers the heat to create a gentle, indirect heat source for a clean and flavorful vapor.
What’s most remarkable is its heat-up speed. Most vaporizers on the market have a preheat cycle, and the best ones take about 30 to 40 seconds to heat up for an inhale. This long preheating detracts from the intuitive experience of smoking a real cigar or cigarette, which you can just hand to a friend without explaining how to operate it or waiting for it to begin working.
Our vaporizer’s heater surface hits operating temperatures in just two and a half seconds. It reaches its ideal operating temperature of 374° F (190° C) in the span of a single breath. It then maintains this temperature, and our CPU and firmware make dynamic adjustments to temperature based on user inhalation. We prioritized rapid reheat time in development and worked hard to develop a custom PCM that would deliver power at a high amperage to make this possible (more on our custom PCM below).
While most vaporizers use mini-ovens that heat from the outside in and cause uneven heat distribution, the Spyre heats material from the inside out. Our model is more efficient because the heat radiates away from the heater and then is picked up by air that the user pulls in from the bottom of the device. The air then travels through a clever thermal path which grabs escaping heat and loops it back through the bottom of the grinding chamber, recirculating and recycling it.
The Spyre is the only vaporizer that has a built-in grinder with stainless steel teeth. Not only does this save time and reduce mess for the user, but it’s also a heat-saving tactic. Without a grinder, the user would have to open the heating chamber to stir the material for even heat distribution—which lets most of their heat escape.
The grinder is a customer favorite, but it was challenging to incorporate such a complex mechanical motion into the small space in a way that didn’t cause problems with the electrical connection. It’s difficult to run a reliable electrical connection to a rotating segment that is also detachable, because the user detaches the section to load material and then must reattach it properly to reestablish a key electrical contact. Instead, we opted to use capacitive sensors to wirelessly detect the user’s touch.
Hidden Micro-USB Battery Charging
The tip of the vaporizer slides out to reveal a micro USB port behind the LED ring to plug in a battery charging cord. The sliding port also has a hidden magnet that has a pleasant audible click when a user opens and closes it. It also has an important function: it activates and deactivates a hidden reed switch, which works as the device’s master on/off switch, allowing the device to have no visible buttons or switches, just like a real cigar.
A user can charge the Spyre with a standard micro-USB cord or the USB charging collar that comes with the device. The wall adapter accepts 90-264 Volts AC, so users can travel with it internationally without a voltage adapter. The device takes about 2.75 hours to fully charge, and a full battery supports up to 250 puffs.
Printed Circuit Boards
The Spyre features five printed circuit boards:
1. Sensor board: The sensor board holds the “brains” of the device. The CPU detects inputs from different sensors around the device, like the thermistor temperature sensor and precision pressure sensor. It then regulates power dynamically based on chamber temperature and inhalation speed.
2. Heater board: This board connects to the heater.
3. Charging board: The charging board holds the micro USB port for charging and LEDs for the user interface, like the RGB LED ring at the bottom of the device.
4. PCM (protection circuit module): The PCM protects the lithium-ion battery pack from damage or explosion through overcharging and overdischarging.
5. Power board: The power board conditions the power from the battery pack before it reaches the heater.
We use a variety of materials, including a high-temperature polycarbonate, PEEK (polyether ether ketone), quartz, ceramics, aluminum, and steel. The PCM uses a conductive epoxy fill because it supports such a high current, delivering up to 15 amps in 15 seconds, which is a huge number for a single lithium-ion battery cell.
Most of our boards are multi-layered, rigid FR-4 boards of two different thicknesses, but we also have flex circuits. Designing them to be mechanically resilient was a challenge. Most applications use flex to accommodate interesting internal orientations, bend around corners, and get into tight places. But when you start moving them, you run into mechanical issues.
For example, we used two flex boards to connect the charging board (with the USB charger) to the sensor board and PCM. The USB charger is accessed through a sliding mechanism on the tip of the device, which clicks open and closed. The flex boards are in a very small space, so after about 200 clicks of the tip, we’d start to see one part of the circuit wear through as the copper experienced strain, hardening and became brittle. We adjusted our mechanical design and the orientation of the flex boards to give them enough mechanical clearance and space, so they wouldn’t crimp and break.
Custom PCB For High Power Delivery And Battery Pack Safety
We are most proud of Spyre’s rapid reheating time, made possible through power delivery and heat efficiency. We developed a custom protection circuit module (PCM) to keep our battery safe as it delivered high-amperage power.
Most vaporizes use a battery pack bought off-the-shelf from a manufacturer. These off-the-shelf batteries wouldn’t suffice for the power delivery we were hoping for. So in our first model, we put together a proof of concept that used two battery cells but no PCM to provide the power we needed.
This iteration was never intended to be mass-produced for commercial distribution. The PCM keeps the battery safe by regulating discharge and essentially prevents it from blowing up. Our device was safe under normal use, but we were aware that many people in the e-cigarette and vaporizer community were “modders”—people who tinker with electronics and cut open or modify the battery to augment the performance without proper battery protection. (These activities can result in serious injury or death, and we strongly advise against them.)
It was important to us that we were up to commercial standards to keep everyone safe as we grew our company and began distribution. For our next product iteration, we wanted to find a custom PCM solution that offered both safety and high amperage power delivery for a fast heat-up.
After searching high and low, we got lucky: a customer of ours happened to be a hardware engineer who specialized in lithium ion battery packs. He helped us design a custom PCM with a very low impedance power path. When you have a high power application like a vaporizer, any impedance in the circuit will convert some of the power you’re delivering into waste heat instead of getting it to the heater. That may seem ideal, because you’re trying to produce heat anyway, but we wanted the heater itself to rise in temperature, not the circuit board chamber. To achieve this, we used very low rds(on) MOSFETs and carefully considered the routing paths to minimize lengths.
Our engineer also connected us with a vendor that could manufacture our battery pack design. We went from two battery cells to one, which helped cut down on space so our next product version could be smaller.
Precision Pressure Sensor
We updated our pressure sensor so it’s easier to manufacture and gives the user a cleaner experience.
Our first pressure sensor was an expensive industrial sensor which had been pre-calibrated and pre-loaded with code. This made it easy to use, because uncalibrated pressure sensors require users to write their own firmware to interpret the voltage response from the sensor. However, this also made it expensive. It was also very large (about the size of a quarter) because it was intended for industrial uses, not commercial applications.
In our original Churchill model, before the upgrade, our PCBs were stacked like floors in a four- or five-story building between big perpendicular girders that support the whole structure. They looked like little levels that went down the length of the tube.
One of the boards was about 20% wider than the others, allowing it to lay flush against a lip on the wall of the tube and create a seal.
As a user inhaled, air would enter from the bottom of the vaporizer and rise up through the device. It hit the seal from the larger PCB, which created a chamber that registered changes in air pressure. A pipe from the sensor board detected both ambient pressure and pressure change and fed that data to the CPU.
Even though it functioned properly, there were still some issues with this arrangement. It would have been difficult to mass-produce, because if a board were off by even a millimeter and didn’t fit into the enclosure perfectly, the sensor wouldn’t detect a pressure change. Assembly was successful when completed by hand, but we intended to produce many units and this wouldn’t have been scalable.
Another issue was that the air flowed over and through uncovered PCBs, and we didn’t want users to worry about cleanliness and outgassing.
To fix this, in our current model, we sealed off one contained chamber with the battery pack and PCBs. It has double air seals that separate it from the grinding chamber and the airflow, which runs up the side of the device to the mouthpiece. We also changed the way the PCBs were oriented inside the tube, because the current vertical stacking took up a lot of space. Every millimeter counts when you’re miniaturizing a device.
We continue to evaluate the sensors we’re using. In a future model, we may add sensors that detect hand contact, lip contact, and more.