DesignSpark Electrical Logolinkedin
Menu Search
Ask a Question

Custom Temperature Activated LED required build help required for a COMPLETE novice....

Good morning all.

Hopefully this group can help with this as what I'm looking for is a HUGE request of somebody out there to try and mentor me through what I want to achieve as follows, bearing in mind that I have NEVER EVER built a circuit before....

Throughout the summer months I do displays/talks about reptiles at public events. These animals for the duration of the event are kept in Really Useful Boxes (aka RUBs) and i want to make sure that after experiencing a few very hot days this year that the safety of my animals is enhanced with regard to them not overheating whilst in the RUBS (yes there are air holes etc but some days the general air flow is that low that i don't think that is enough)

So I have thought about building some very small circuits that have a thermometer probe that could be put down through one of the air holes in the lids and if the temperature goes above say 30-32.5C then a red LED will come on. There are a few things that it needs to be able to do/have as follows....

- the circuit/battery to fit in a cube say 35mm*50mm in plan and no higher than say 25mm tall (anything much bigger and it will just be too much of a distraction of what the animal is inside).

- Have an a switch for on/off operation (in the future I was thinking of gluing a magnet to the underside of the RUB lid next to the air hole and then using that to activate a reed switch to turn on the circuit as well as hold the circuit cube in place....)

- once the circuit is on, a green LED will need to be on so we know the circuit is active and that the battery isn't flat (this can either be a static ON light of a flashing light would flashing help with battery life?)

- This would need to be powered by something as small as a single CR2032 coin battery or similar.

- the thermistor bead probe to be able to be about 75-100mm long so it can dangle down through the lid adequately.

- whilst for now if we can somehow hard fix the LED turning red at say 30-32.5C, a future option would be to have some sort of adjustable setting as different animals have different comfortably maximum temperatures they prefer...

... I think that's the list of needs for now I really really do appreciate that finding somebody who can help with not only the design of this board but to coach me through component purchase, the actual build etc it is a MASSIVE request but with the help of the many online circuit design software out there I have found while trying to figure this out, i would like to think that somebody is up for the challenge.

I hope to hear from some of you soon.

Go safe.

Comments

October 18, 2019 14:58

Afternoon All....

Just wondering if anyone has been tempted by my post at all?

T

0 Votes

October 21, 2019 10:16

@Torcrayth I have, but I've been busy with some work deadlines. There are several approaches you can consider. If you try to do it yourself, as a complete novice, an analog approach probably requires the least overall design effort for this particular project. One question is just how accurate it needs to be. I'm guessing within a degree or degree and a half is sufficient. The answer affects the choice of temperature sensor. Some sensors are very easy to use, but not highly accurate in terms of interchangeability. They may be quite repeatable, though. Low power comparators with open collector or open drain outputs which can directly switch an LED are readily available. Using a voltage divider with two fixed resistors at one input of the comparator, and a voltage divider comprised of the thermistor and a potentiometer at the other input to the comparator. Place the thermistor in a calibration environment (enclosure the desired detection temperature (as measured by an accurate external thermometer)) and adjust the potentiometer until the LED comes on when placing the thermistor in the warm calibrated environment, and goes back off when removed. This is a minimalist approach - no human-readable display of actual temperature during normal operation, and takes a bit of time to set up the calibration environment when you want to change the alarm temperature. But minimal parts. The other aspects to be addressed with this approach though are power consumption and self-heating. I'd use the output of a timer IC (or dual timer IC) to power the very-low-power comparator and thermistor/divider, with the timer having a relatively small duty cycle (on for 0.01 seconds, off for the remainder of the second). You can use a capacitor with the LED to stretch the on-time of the LED to a tenth or quarter of a second, to make it more visible than a briefer pulse. The LED won't be on continuously - it will blink at the timer's interval rate. But I think that would be acceptable in this application. The low duty cycle of the timer reduces power consumption correspondingly, and also helps reduce self-heating of the sensing thermistor, since it would only be passing current during the much briefer interval. (Some other temperature sensing elements like platinum RTDs can be very accurate off-the-shelf, but have much lower resistance, requiring more care to avoid self-heating from the current passing through the sensor.) The green LED could also be powered by the timer output. The other approach I'd consider would be digital, using a device like the Texas Instruments TMP117. It has the temperature available in digital form via I2C output. It has a programmable alarm threshold, with a dedicated output pin for indicating the alarm condition. It is high accuracy, accurate to +/- 0.1 degree C over the -20 to +50C range, with NIST traceability and suitable for medical applications. If you operate it at a low (1/sec) sample rate, average power consumption is 22 microamps or less, depending on settings. It will operate over a supply range of 1.8 to 5.5 volts. Pair it with a small, low power microcontroller for configuration, and you have a lot of flexibility with minimal parts count. You could do recording of the environment over time, for example, to show the animal wasn't exposed to hazardous temps on the outing. I'd probably use an off-the-shelf microcontroller board like the Adafruit trinket or feather series which are available with the battery management features supplied for you. This approach could involve less hardware development and wiring than the analog approach, but with the complexity of programming. But the amount of programming is up to you and what features you want to include. A couple of push-buttons and display (only connected/enabled during setup to limit battery drain, if desired) could allow easy, precise setting of the alarm temperature.

October 22, 2019 12:33

@BradLevy The following is a bit long but I thought it was a useful way to reply to all the comments you made. BradLevy -- If you try to do it yourself, as a complete novice, an analog approach probably requires the least overall design effort for this particular project. Torcrayth -- Yep, Sounds good to me. BradLevy -- One question is just how accurate it needs to be. I'm guessing within a degree or degree and a half is sufficient. The answer affects the choice of temperature sensor. Torcrayth -- 1.5 degrees would be more than accurate enough as the idea is that this is just an extra backup to the procedures I have in place. BradLevy -- Some sensors are very easy to use, but not highly accurate in terms of interchangeability. They may be quite repeatable, though. Low power comparators with open collector or open drain outputs which can directly switch an LED are readily available. Using a voltage divider with two fixed resistors at one input of the comparator, and a voltage divider comprised of the thermistor and a potentiometer at the other input to the comparator. Torcrayth -- OK, I’m a little lost as to what you are on about here when it comes to collectors and drains but I’m happy to learn. BradLevy -- Place the thermistor in a calibration environment (enclosure the desired detection temperature (as measured by an accurate external thermometer)) and adjust the potentiometer until the LED comes on when placing the thermistor in the warm calibrated environment, and goes back off when removed. This is a minimalist approach - no human-readable display of actual temperature during normal operation, and takes a bit of time to set up the calibration environment when you want to change the alarm temperature. But minimal parts. Torcrayth -- I like the sound of this as a good way to simply achieve what I am trying to do and as a possible springboard for moving forward. Whilst the circuit will be used for various types of snake, the snake with the lowest high temperature level would be fine point to monitor the others against as well (if this works as well as I hope I may actually set different circuits at different temp settings for use with specific types of snake – but we’ll take this one step at a time …for now  ) BradLevy -- The other aspects to be addressed with this approach though are power consumption and self-heating. I'd use the output of a timer IC (or dual timer IC) to power the very-low-power comparator and thermistor/divider, with the timer having a relatively small duty cycle (on for 0.01 seconds, off for the remainder of the second). Torcrayth -- OK, I think I understand this in that effectively the circuit will only be powered on for 1/100th of a second, take a reading, turn on red LED if required? BradLevy -- You can use a capacitor with the LED to stretch the on-time of the LED to a tenth or quarter of a second, to make it more visible than a briefer pulse. The LED won't be on continuously - it will blink at the timer's interval rate. But I think that would be acceptable in this application. Torcrayth -- very acceptable as in many ways a blinking light will catch the eye more than a static one. BradLevy -- The low duty cycle of the timer reduces power consumption correspondingly, and also helps reduce self-heating of the sensing thermistor, since it would only be passing current during the much briefer interval. (Some other temperature sensing elements like platinum RTDs can be very accurate off-the-shelf, but have much lower resistance, requiring more care to avoid self-heating from the current passing through the sensor.) The green LED could also be powered by the timer output. Torcrayth -- So a green flashing LED to show all is good..Nice! BradLevy -- The other approach I'd consider would be digital, using a device like the Texas Instruments TMP117. It has the temperature available in digital form via I2C output. It has a programmable alarm threshold, with a dedicated output pin for indicating the alarm condition. It is high accuracy, accurate to +/- 0.1 degree C over the -20 to +50C range, with NIST traceability and suitable for medical applications. If you operate it at a low (1/sec) sample rate, average power consumption is 22 microamps or less, depending on settings. It will operate over a supply range of 1.8 to 5.5 volts. Pair it with a small, low power microcontroller for configuration, and you have a lot of flexibility with minimal parts count. Torcrayth -- OK, this is a touch deep for me at this stage but I think I know what you’re saying but with completion of the intial idea perhaps I could bring my understanding of this bit on more. BradLevy -- You could do recording of the environment over time, for example, to show the animal wasn't exposed to hazardous temps on the outing. Torcrayth -- I really like this idea and whilst currently there isn’t any legislation for us to record such things I do believe that it will occur at some point! BradLevy -- I'd probably use an off-the-shelf microcontroller board like the Adafruit trinket or feather series which are available with the battery management features supplied for you. This approach could involve less hardware development and wiring than the analog approach, but with the complexity of programming. But the amount of programming is up to you and what features you want to include. A couple of push-buttons and display (only connected/enabled during setup to limit battery drain, if desired) could allow easy, precise setting of the alarm temperature Torcrayth -- I’m certainly not afraid of trying to do a bit of programming at some point. All in all, thankyou BradLevy for taking the all that thinking about my query and getting back to me with it all. If you are up for helping me with this it would be much appreciated and I promise to not to become a pain and will always remember that there are lives outside of this here. Hope to hear from you soon. T

October 24, 2019 08:33

@BradLevy The following is a bit long but I thought it was a useful way to reply to all the comments you made. BradLevy -- If you try to do it yourself, as a complete novice, an analog approach probably requires the least overall design effort for this particular project. Torcrayth -- Yep, Sounds good to me. BradLevy -- One question is just how accurate it needs to be. I'm guessing within a degree or degree and a half is sufficient. The answer affects the choice of temperature sensor. Torcrayth -- 1.5 degrees would be more than accurate enough as the idea is that this is just an extra backup to the procedures I have in place. BradLevy -- Some sensors are very easy to use, but not highly accurate in terms of interchangeability. They may be quite repeatable, though. Low power comparators with open collector or open drain outputs which can directly switch an LED are readily available. Using a voltage divider with two fixed resistors at one input of the comparator, and a voltage divider comprised of the thermistor and a potentiometer at the other input to the comparator. Torcrayth -- OK, I’m a little lost as to what you are on about here when it comes to collectors and drains but I’m happy to learn. BradLevy -- Place the thermistor in a calibration environment (enclosure the desired detection temperature (as measured by an accurate external thermometer)) and adjust the potentiometer until the LED comes on when placing the thermistor in the warm calibrated environment, and goes back off when removed. This is a minimalist approach - no human-readable display of actual temperature during normal operation, and takes a bit of time to set up the calibration environment when you want to change the alarm temperature. But minimal parts. Torcrayth -- I like the sound of this as a good way to simply achieve what I am trying to do and as a possible springboard for moving forward. Whilst the circuit will be used for various types of snake, the snake with the lowest high temperature level would be fine point to monitor the others against as well (if this works as well as I hope I may actually set different circuits at different temp settings for use with specific types of snake – but we’ll take this one step at a time …for now  ) BradLevy -- The other aspects to be addressed with this approach though are power consumption and self-heating. I'd use the output of a timer IC (or dual timer IC) to power the very-low-power comparator and thermistor/divider, with the timer having a relatively small duty cycle (on for 0.01 seconds, off for the remainder of the second). Torcrayth -- OK, I think I understand this in that effectively the circuit will only be powered on for 1/100th of a second, take a reading, turn on red LED if required? BradLevy -- You can use a capacitor with the LED to stretch the on-time of the LED to a tenth or quarter of a second, to make it more visible than a briefer pulse. The LED won't be on continuously - it will blink at the timer's interval rate. But I think that would be acceptable in this application. Torcrayth -- very acceptable as in many ways a blinking light will catch the eye more than a static one. BradLevy -- The low duty cycle of the timer reduces power consumption correspondingly, and also helps reduce self-heating of the sensing thermistor, since it would only be passing current during the much briefer interval. (Some other temperature sensing elements like platinum RTDs can be very accurate off-the-shelf, but have much lower resistance, requiring more care to avoid self-heating from the current passing through the sensor.) The green LED could also be powered by the timer output. Torcrayth -- So a green flashing LED to show all is good..Nice! BradLevy -- The other approach I'd consider would be digital, using a device like the Texas Instruments TMP117. It has the temperature available in digital form via I2C output. It has a programmable alarm threshold, with a dedicated output pin for indicating the alarm condition. It is high accuracy, accurate to +/- 0.1 degree C over the -20 to +50C range, with NIST traceability and suitable for medical applications. If you operate it at a low (1/sec) sample rate, average power consumption is 22 microamps or less, depending on settings. It will operate over a supply range of 1.8 to 5.5 volts. Pair it with a small, low power microcontroller for configuration, and you have a lot of flexibility with minimal parts count. Torcrayth -- OK, this is a touch deep for me at this stage but I think I know what you’re saying but with completion of the intial idea perhaps I could bring my understanding of this bit on more. BradLevy -- You could do recording of the environment over time, for example, to show the animal wasn't exposed to hazardous temps on the outing. Torcrayth -- I really like this idea and whilst currently there isn’t any legislation for us to record such things I do believe that it will occur at some point! BradLevy -- I'd probably use an off-the-shelf microcontroller board like the Adafruit trinket or feather series which are available with the battery management features supplied for you. This approach could involve less hardware development and wiring than the analog approach, but with the complexity of programming. But the amount of programming is up to you and what features you want to include. A couple of push-buttons and display (only connected/enabled during setup to limit battery drain, if desired) could allow easy, precise setting of the alarm temperature Torcrayth -- I’m certainly not afraid of trying to do a bit of programming at some point. All in all, thankyou BradLevy for taking the all that thinking about my query and getting back to me with it all. If you are up for helping me with this it would be much appreciated and I promise to not to become a pain and will always remember that there are lives outside of this here. Hope to hear from you soon. T

October 24, 2019 08:34

@BradLevy First, a huge thanks for being interested in this enought to come back to it. I appreciate that there is a world outside of this query of mine and that must always come first of course. As your reply is so wonderfully long, I will take some time over my lunch break to try and answer it as concisely as i can in an effort to make this as painless as possible for you. Back soon.

October 15, 2019 07:35

That's quite a well specified requirement you have which is good. It is difficult what to suggest as you say a complete novice. The ideal solution would probably a microcontroller, but that needs coding and a programmer and the ability to debug the code. My suggestion is to look for a simple hardware solution using readily available chips. I had a search and some old but still available components as used in this link may provide a solution http://www.circuitstoday.com/led-bargraph-thermometer
Or search online for "temperature alarm" some freezer alarms or aquarium alarms cover -50C to 70C and may be suitable? Most are under £10.

0 Votes

October 15, 2019 07:33

@Boss A REPLY - YES!!!!!!!!! Firstly, thank you so much for taking the time to read my long post. Although i am a complete novice at building circuits i am technically minded (I've spent my entire working life producing and reading detailed construction drawings) and with guidance I think its something I could do and the more I've tried to find a way to build something the more I want to be able to do it... if there is any chance that you could help with a simple diagram of the components I'd need (and an explanation as to why I need them so I can learn) I'll do all I can to leave you in as much peace as I can.... Hopefully hear from you soon. T