Sure. i've been waiting to post until i had some concrete progress, but yeah, i've been neglecting the weekly status update.
i've been relentlessly testing the most recent cube, Unit #2. i can always diddle with the temperature control software, but it really is working exactly the way i want. There's a little overshoot at startup (about 20F) -- with SETPOINT at 385F, it will hit 410F with a fresh battery pack, in about 30 seconds. Then it drifts down to about 10F below SETPOINT and over the next 60 seconds drifts back up to SETPOINT. Then the green LED comes on. At this point, the bud is nicely toasted and the vapor will flow. The second and third hits are bigger and richer than the first, usually (depends on the bud and trichome density - the first hit can be just as nice).
i'm still thinking about what is going on here. Certainly, the interior is getting heat saturated, and there is a lot to heat up -- glass oven, glass vial, bud. And i think the waxy cover of the trichomes is getting melted. i'm almost sure (not a physicist here) there is a heat of vaporization associated with this melting, so heat will be absorbed by this process.
So, although it is taking 90 seconds to the first toke, i really doubt i can speed this up in this design geometry, and it doesn't appear necessary.
Now, after 6 to 12 hits, as the trichomes are drained and the vapor begins to thin out, the temperature drifts up (as reported by the digital thermometer). Toward the end of the 7.5 minute session, the display shows 390F, instead of the nominal 385F. All in all, a slightly increasing temperature toward the end of the session is kind of idea for maximum extraction.
However, all during the session, the green LED is slowly blinking on and off. This is the ideal operation. The green LED is ON, when the thermocouple reads a temperature at or above the SETPOINT temperature. So, as far as the PIC knows, the temperature is as stable as it can be, given the jitter in the readings (approximately 1C).
Keep in mind, the digital thermometer (DT) has an entirely different time constant from the type-K thermocouple used by the PIC to control the temperature -- the DT lags the actual temperature by 10s of seconds. So the DT is really showing a long-term moving average of the temperature - which is just as well, considering the jitter in the temperature reading the PIC is getting from the MAX6675.
So, i find the green LED to be a reliable indicator of when it is time to start toking. And it is pretty clear when the vapor is too thin to continue -- about 7 minutes for a 1/2 full vial.
And, although a slow toke evokes the densest vapor, the volume and speed of toke do not drop the temperature more than 0.5F, which then quickly recovers. Temperature control is a solved problem. Even so, i'll probably play with the PID coefficients a bit -- lots more to learn here.
So there is one last design problem -- the "battery sled". i played around with some 18650 batteries for a few weeks (usually while sipping Merlot). The smaller diameter offers some design opportunities, but the reduced capacity keeps this option on the back burner. Here is a drawing ... i made a prototype but i'm just not happy with the 18650 batteries. my initial impressions are correct -- wait for the next generation battery to get to a smaller form factor.
All the 26650 batteries i use have a soldered Deans (female) plug and are coated in plastic using Dip-It. Safety over beauty. If you have a MFLB and ever have a battery get shorted out in your pocket, these batteries would blast a hole through your leg. The power of these cells has kept me freaked out for a while.
The problem has been there are NO battery holders available in the fucking world for dual 26650 batteries. i could get a MakerBot and fab my own holders, but then i've got a plastic component which pollutes the design. So making a small wood box, just the size of the batteries has been my main problem - how to hold the sides together without adding too much thickness. But i've finally figured out an approach using 1/8" wood (cherry, oak, mohagony, etc.) with cutouts and tabs that should be very sweet. i made a foam prototype to have something to handle and keep next to the keyboard on my desk as extra motivation. Now i just need to cut, groove and drill the wood to try it out. And the tabs are keyed so the pieces can't be assembled incorrectly, shorting out the batteries.
So, no more soldering on the batteries, elimination of one Deans plug per battery (at least 4 batteries per Bud Toaster kit), small size, nice looking wood, and with a place to rest the vial and toke tube. The 36" power cable will wrap around the outside of the sled -- neat and tidy.
When the battery sled is finalized, i can finally get the leather case figured out. my plan is to make a few cases by hand to work out size and layout and then outsource case production to china. It will be a nice leather zipper clam-shell case, approximately 6" long, by 3" wide, by 2" tall. Holding the Bud Toaster, the battery sled, with 2 cells, glass stem, several vials, and an Altoids container of herb.
i'm sure there will be a bunch of production issues to resolve -- mainly through doing and adjusting, and none of them will be show stoppers -- merely engineering problems and i do engineering.
My regular job has been more demanding going into 2nd Quarter, but i'm reserving Rapture Day for the Bud Toaster -- and some brownies.
i've been relentlessly testing the most recent cube, Unit #2. i can always diddle with the temperature control software, but it really is working exactly the way i want. There's a little overshoot at startup (about 20F) -- with SETPOINT at 385F, it will hit 410F with a fresh battery pack, in about 30 seconds. Then it drifts down to about 10F below SETPOINT and over the next 60 seconds drifts back up to SETPOINT. Then the green LED comes on. At this point, the bud is nicely toasted and the vapor will flow. The second and third hits are bigger and richer than the first, usually (depends on the bud and trichome density - the first hit can be just as nice).
i'm still thinking about what is going on here. Certainly, the interior is getting heat saturated, and there is a lot to heat up -- glass oven, glass vial, bud. And i think the waxy cover of the trichomes is getting melted. i'm almost sure (not a physicist here) there is a heat of vaporization associated with this melting, so heat will be absorbed by this process.
So, although it is taking 90 seconds to the first toke, i really doubt i can speed this up in this design geometry, and it doesn't appear necessary.
Now, after 6 to 12 hits, as the trichomes are drained and the vapor begins to thin out, the temperature drifts up (as reported by the digital thermometer). Toward the end of the 7.5 minute session, the display shows 390F, instead of the nominal 385F. All in all, a slightly increasing temperature toward the end of the session is kind of idea for maximum extraction.
However, all during the session, the green LED is slowly blinking on and off. This is the ideal operation. The green LED is ON, when the thermocouple reads a temperature at or above the SETPOINT temperature. So, as far as the PIC knows, the temperature is as stable as it can be, given the jitter in the readings (approximately 1C).
Keep in mind, the digital thermometer (DT) has an entirely different time constant from the type-K thermocouple used by the PIC to control the temperature -- the DT lags the actual temperature by 10s of seconds. So the DT is really showing a long-term moving average of the temperature - which is just as well, considering the jitter in the temperature reading the PIC is getting from the MAX6675.
So, i find the green LED to be a reliable indicator of when it is time to start toking. And it is pretty clear when the vapor is too thin to continue -- about 7 minutes for a 1/2 full vial.
And, although a slow toke evokes the densest vapor, the volume and speed of toke do not drop the temperature more than 0.5F, which then quickly recovers. Temperature control is a solved problem. Even so, i'll probably play with the PID coefficients a bit -- lots more to learn here.
So there is one last design problem -- the "battery sled". i played around with some 18650 batteries for a few weeks (usually while sipping Merlot). The smaller diameter offers some design opportunities, but the reduced capacity keeps this option on the back burner. Here is a drawing ... i made a prototype but i'm just not happy with the 18650 batteries. my initial impressions are correct -- wait for the next generation battery to get to a smaller form factor.
All the 26650 batteries i use have a soldered Deans (female) plug and are coated in plastic using Dip-It. Safety over beauty. If you have a MFLB and ever have a battery get shorted out in your pocket, these batteries would blast a hole through your leg. The power of these cells has kept me freaked out for a while.
The problem has been there are NO battery holders available in the fucking world for dual 26650 batteries. i could get a MakerBot and fab my own holders, but then i've got a plastic component which pollutes the design. So making a small wood box, just the size of the batteries has been my main problem - how to hold the sides together without adding too much thickness. But i've finally figured out an approach using 1/8" wood (cherry, oak, mohagony, etc.) with cutouts and tabs that should be very sweet. i made a foam prototype to have something to handle and keep next to the keyboard on my desk as extra motivation. Now i just need to cut, groove and drill the wood to try it out. And the tabs are keyed so the pieces can't be assembled incorrectly, shorting out the batteries.
So, no more soldering on the batteries, elimination of one Deans plug per battery (at least 4 batteries per Bud Toaster kit), small size, nice looking wood, and with a place to rest the vial and toke tube. The 36" power cable will wrap around the outside of the sled -- neat and tidy.
When the battery sled is finalized, i can finally get the leather case figured out. my plan is to make a few cases by hand to work out size and layout and then outsource case production to china. It will be a nice leather zipper clam-shell case, approximately 6" long, by 3" wide, by 2" tall. Holding the Bud Toaster, the battery sled, with 2 cells, glass stem, several vials, and an Altoids container of herb.
i'm sure there will be a bunch of production issues to resolve -- mainly through doing and adjusting, and none of them will be show stoppers -- merely engineering problems and i do engineering.
My regular job has been more demanding going into 2nd Quarter, but i'm reserving Rapture Day for the Bud Toaster -- and some brownies.
