DIY DIY EC Meter
- Thread starter donkeykong
- Start date
Cheers for your nice comments guys. I think it definitely helps some people think outside the box, for me, im convinced. always do my best work after a few jays 
Sorry been a few days, been knee deep in my little project though.
The reflow oven worked out real nice. Had a lot of support from the guys at ESTechnical and got it up and running, and baked the first temperature sensor onto a PCB I had ordered for proto-vantage. The little adaptor converts the teeny QFN16 to a DIP16 kind of chip size, but wider.
I have a BusPirate , very useful tool for hacking about with interfaces like I2C, SPI and working out an IC before trying it in Arduino. The temp sensor checked in OK and readings were OK but something worried me. The temp sensor has a thermal pad underneath, with this adaptor that had ended bonded to the board, which connected it to a large heatsink area stretching the top and bottom of the adaptor..... so its going to measure PCB temp very well, but not air temp.
After that I got the thinking cap on and made my own little adaptor board. Its about a quarter the size of the Proto-Vantage one, has no spare copper either side of the minimal pcb area. There is even a tiny hole right under the pcb footprint so the chip is least connected to the board as possible!
Sounds difficult but the design of the pcb was simple, and I used a really quick, fast and cheap method to etch the board. Ill have to expand, since you might find it useful ;
the pcb artwork was done with designspark, a free tool. Then it was printed on laser to a transfer sheet and ironed onto a polished copper clad pcb. I've built about 3 or 4 etching tanks over the years, its not so straightforward to made a good job of it and it requires a lot of etchant. I had read an article recently and was itching to give it a whirl (esp since my tank is out of action and really needs to be thrown out and rebuilt).
I followed this method to etch the pcb, worked a treat even with the fine pitches I needed.
Now I have a good sensor to check things against. So I built up a little program in C# that drives the buspirate and talks to the temp IC's and gets the readings and calibration data, calculates the temperatures. I also connects to the arduino and gets the readings from the ambient temp and air temp thermistor to compare against.
Been trying to figure how to calibrate temperature best...
Right now I placed the sensors close to each other in a 3 walled cardboard box with some packing, used a heater to raise the temp to about 40 deg, and taped the box back up. Now as it cools Im logging the data from the sensors, which is working quite nice. The new 24 bit sensor (red line) is amazing and the thermistor (green) is following nicely. The TMP100 IC is not so good (grey) and the blue line is the pcb temp sensing IC, always following but with a lower temp.
Want to see if the thermistor will hold over a range of 15 - 35 deg c, that will cover as much as we'll need and I think it will give 0.1deg or better accuracy by the looks of it.
TMP100 IC could be compensated with a correction.
I'll post a few more pics later so you can see what we are talking about.
Sorry been a few days, been knee deep in my little project though.
The reflow oven worked out real nice. Had a lot of support from the guys at ESTechnical and got it up and running, and baked the first temperature sensor onto a PCB I had ordered for proto-vantage. The little adaptor converts the teeny QFN16 to a DIP16 kind of chip size, but wider.
I have a BusPirate , very useful tool for hacking about with interfaces like I2C, SPI and working out an IC before trying it in Arduino. The temp sensor checked in OK and readings were OK but something worried me. The temp sensor has a thermal pad underneath, with this adaptor that had ended bonded to the board, which connected it to a large heatsink area stretching the top and bottom of the adaptor..... so its going to measure PCB temp very well, but not air temp.
After that I got the thinking cap on and made my own little adaptor board. Its about a quarter the size of the Proto-Vantage one, has no spare copper either side of the minimal pcb area. There is even a tiny hole right under the pcb footprint so the chip is least connected to the board as possible!
Sounds difficult but the design of the pcb was simple, and I used a really quick, fast and cheap method to etch the board. Ill have to expand, since you might find it useful ;
the pcb artwork was done with designspark, a free tool. Then it was printed on laser to a transfer sheet and ironed onto a polished copper clad pcb. I've built about 3 or 4 etching tanks over the years, its not so straightforward to made a good job of it and it requires a lot of etchant. I had read an article recently and was itching to give it a whirl (esp since my tank is out of action and really needs to be thrown out and rebuilt).
I followed this method to etch the pcb, worked a treat even with the fine pitches I needed.
Now I have a good sensor to check things against. So I built up a little program in C# that drives the buspirate and talks to the temp IC's and gets the readings and calibration data, calculates the temperatures. I also connects to the arduino and gets the readings from the ambient temp and air temp thermistor to compare against.
Been trying to figure how to calibrate temperature best...
Right now I placed the sensors close to each other in a 3 walled cardboard box with some packing, used a heater to raise the temp to about 40 deg, and taped the box back up. Now as it cools Im logging the data from the sensors, which is working quite nice. The new 24 bit sensor (red line) is amazing and the thermistor (green) is following nicely. The TMP100 IC is not so good (grey) and the blue line is the pcb temp sensing IC, always following but with a lower temp.
Want to see if the thermistor will hold over a range of 15 - 35 deg c, that will cover as much as we'll need and I think it will give 0.1deg or better accuracy by the looks of it.
TMP100 IC could be compensated with a correction.
I'll post a few more pics later so you can see what we are talking about.
good results from overnight, the NTC thermistor and the accurate Air Temp IC (TSYS01) are holding together very well across the range I need plus a bit more, 15deg C -> 35 deg C.
The lower two traces are the PCB temp configured TSYS01 , and the TMP100 IC. I think the TMP100 must be doing a good job of PCB temp rather than air temp also, as it matches so well. Probably I will check again how the chip is mounted and find to well thermally connected to its adaptor board.
So, what did I learn, with these temperature IC's the mounting is very important. If you are trying to measure air temp then we want least copper on the pcb, least coupling with the pcb temp.
My custom adaptor board does the job well on this side.
Also maybe I didn't explain what the cardbox box lark is all about. To check the calibration, I place all the probes close together. In free air like a room there are minor local variations in temperature, even depending on closeness to table top etc. To make it the sensors are in their own little environment where the air temp can be more even, in a largish cardboard box. Second thing that does is shield them from fast temperature changes and so they cool slowly, depending on the ambient around the box. This is quite important as if they are heated or cooled too rapidly, the sensors react in faster/slower times and there is a larger delta between them - you cant check calibration like that. The temp needs to drop slower than the slowest probe, so the box helps achieve that. Heating it up inside to 40-ish worked well and so I closed the box and logged it falling towards ambient room temp. Then I placed the box outside where night temp would slowly cool it further down to 15 deg c.
Took some thinking about how to do this kind of calibration without crazy equipment, but very pleased with the results. that precision thermistor is very good, at least over the range we need its linear enough and keeps up with the TSYS01, £13 chip IC with 24-Bit DAC and 5 order calibration parameters.
Have to say that TYSS01 is amazing though. The resolution off the 24bit-ADC is extremely smooth and super small variations in temp are visible. You can place your hand 9 inches from the chip and watch the heat radiating off your hand begin to warm the chip by hundreths of a degree..... The readings it gives are consistent to 0.01 degree resolution.... yikes!
This really has made me think about temperature in a different way. Still now at least I satisfied myself about what the damned temperature really is!
Now Ill be able to fix the ambient reading and make it correct , and also I have a method and some SW tools now to calibrate the temp sensors of the EC probes next. Thankfully calibrating the EC is lot quicker and easier than calibrating for temperature
The lower two traces are the PCB temp configured TSYS01 , and the TMP100 IC. I think the TMP100 must be doing a good job of PCB temp rather than air temp also, as it matches so well. Probably I will check again how the chip is mounted and find to well thermally connected to its adaptor board.
So, what did I learn, with these temperature IC's the mounting is very important. If you are trying to measure air temp then we want least copper on the pcb, least coupling with the pcb temp.
My custom adaptor board does the job well on this side.
Also maybe I didn't explain what the cardbox box lark is all about. To check the calibration, I place all the probes close together. In free air like a room there are minor local variations in temperature, even depending on closeness to table top etc. To make it the sensors are in their own little environment where the air temp can be more even, in a largish cardboard box. Second thing that does is shield them from fast temperature changes and so they cool slowly, depending on the ambient around the box. This is quite important as if they are heated or cooled too rapidly, the sensors react in faster/slower times and there is a larger delta between them - you cant check calibration like that. The temp needs to drop slower than the slowest probe, so the box helps achieve that. Heating it up inside to 40-ish worked well and so I closed the box and logged it falling towards ambient room temp. Then I placed the box outside where night temp would slowly cool it further down to 15 deg c.
Took some thinking about how to do this kind of calibration without crazy equipment, but very pleased with the results. that precision thermistor is very good, at least over the range we need its linear enough and keeps up with the TSYS01, £13 chip IC with 24-Bit DAC and 5 order calibration parameters.
Have to say that TYSS01 is amazing though. The resolution off the 24bit-ADC is extremely smooth and super small variations in temp are visible. You can place your hand 9 inches from the chip and watch the heat radiating off your hand begin to warm the chip by hundreths of a degree..... The readings it gives are consistent to 0.01 degree resolution.... yikes!
This really has made me think about temperature in a different way. Still now at least I satisfied myself about what the damned temperature really is!
Now Ill be able to fix the ambient reading and make it correct , and also I have a method and some SW tools now to calibrate the temp sensors of the EC probes next. Thankfully calibrating the EC is lot quicker and easier than calibrating for temperature
L
lowbudge
Guest
max 
for not stopping!
for not stopping!
Donkey Kong stop throwing barrels at me!
Haha just kidding
:. You are one of my personal heroes here on AFN for attempting something like this.
You don't need to do this, but you do this by heart with much passion! Much respect :High 5:.
:tiphat:
Haha just kidding
:. You are one of my personal heroes here on AFN for attempting something like this.You don't need to do this, but you do this by heart with much passion! Much respect :High 5:.
:tiphat:
Happy days happy days
Made a order for
Arduino leopard R3 / temp/humidity sensor / relays / breadboard / lcd display a bounch of cables e.t.c.
I will wait with ph probe / shield.. need to figure out how to code this thing before going all the way $$..
So nice to have some new "toys" to play with.
Looking forward to this "automatic ph adjusting" i will make
Happy working
Peace
Made a order for
Arduino leopard R3 / temp/humidity sensor / relays / breadboard / lcd display a bounch of cables e.t.c.
I will wait with ph probe / shield.. need to figure out how to code this thing before going all the way $$..
So nice to have some new "toys" to play with.
Looking forward to this "automatic ph adjusting" i will make
Happy working
Peace
cheers guys!
Im always up to one crazy project or another, its just nice that others are enjoying it also, glad I decided to blog it like this.
Been a busy week, out of the country on work all week so not much could be done.
Still have not moved off the temperature topic yet! After the last update I came across a couple more interesting and accurate sensors that I wanted to check. I found another nice IC good deal cheaper than the TSYS01 , its a through hole type package rather than QFN16 so much easier to mount and use, less hassle to measure what we want, air temp.
IST HYT271 Hygrochip (0.2deg c accuracy plus humidity sensor)
TSIC 506F 0.1deg C accuracy temp sensor
I really wanted this to come in within 0.1deg c of the TSYS01 so to be sure they are aligned.. but alas. Something so simple, so difficult to get right!
Well, I really like the part, it uses a one-wire interface, there is arduino library already for the "ZACWire" interface, and it will support good cable runs, so this would work nicely to run from the box into the tent and outside the tent measuring ambient.
I thought nice option would be maybe two for the ambient, for loft type installs. One above the "tent" , and one same high but the other side of the space, so to directly measure the thermal spot which might show on loft, garage or outbiulding with a bird in the sky. my one LED light wont make any impact at all, but others might find useful option! dum... dum dum, hello my name is dr greenthumb
Going to get a few more of those TSIC 506F, to compare if I can mount them and get readings within range of each other. Id like to be able to make a recommendation out of this in the end, so far this one is my fav I think. Can be buried in a tube etc for sensing liquid temp also.
Im always up to one crazy project or another, its just nice that others are enjoying it also, glad I decided to blog it like this.
Been a busy week, out of the country on work all week so not much could be done.
Still have not moved off the temperature topic yet! After the last update I came across a couple more interesting and accurate sensors that I wanted to check. I found another nice IC good deal cheaper than the TSYS01 , its a through hole type package rather than QFN16 so much easier to mount and use, less hassle to measure what we want, air temp.
IST HYT271 Hygrochip (0.2deg c accuracy plus humidity sensor)
TSIC 506F 0.1deg C accuracy temp sensor
I really wanted this to come in within 0.1deg c of the TSYS01 so to be sure they are aligned.. but alas. Something so simple, so difficult to get right!
Well, I really like the part, it uses a one-wire interface, there is arduino library already for the "ZACWire" interface, and it will support good cable runs, so this would work nicely to run from the box into the tent and outside the tent measuring ambient.
I thought nice option would be maybe two for the ambient, for loft type installs. One above the "tent" , and one same high but the other side of the space, so to directly measure the thermal spot which might show on loft, garage or outbiulding with a bird in the sky. my one LED light wont make any impact at all, but others might find useful option! dum... dum dum, hello my name is dr greenthumb
Going to get a few more of those TSIC 506F, to compare if I can mount them and get readings within range of each other. Id like to be able to make a recommendation out of this in the end, so far this one is my fav I think. Can be buried in a tube etc for sensing liquid temp also.
Nice!! You got a bunch of fun stuff on the way, good party pack to get busy with
Will be real interesting to see how that works out with adjusting pH!!
Will be real interesting to see how that works out with adjusting pH!!
Similar threads
