Paul, I think you may have mis-read my email. I was referring to the "polymer" blob of material on the device, not any polyimide material that may be present. Your response would be dead-on if I was referring to any polyimide though! At work, polyimide is a daily word (although mostly referred to as Kapton).
As for taking it apart, it doesn't take much to pop the cap off as my work had one happen a few years ago when the board was dropped onto the edge of a bench. So, if you do decide to open it up, here is a reference picture of one that only went through pick-n-place and reflow then was busted open.
As for using hot melt to "seal" bare metal, you would be surprised at how much it is used in PCBA manufacturing by certain companies. At work, we have one particular customer that uses enough for us to justify the purchase of an industrial hot melt gun. I was a bit surprised when I was handed a purchase req for signoff approval for a "$480 Glue Gun" for that one......
Conformal coating can be built up with multiple layers, enough to bury those sharp edges. I prefer silicone resin, for its flexibility. Moisture can penetrate the conformal coating, but if the surface is squeaky clean, the moisture won't find a conductive path between the circuit conductors. Solder mask on a circuit board is important too. An unmasked board is much harder to protect. Epoxy is an option too, but it is rigid, so thermal expansion and contraction can break components or lift them right off a circuit board. It is better to use a flexible coating, or a flexible coating between the components and the epoxy. In the old days, the preferred coating was wax with a fungicide.
What is an SI2007?
The SHT2x is not too hard to deal with. It is leadless, but only 6 pins and a fairly wide pitch. I do have solder paste, stencil and a reflow oven to solder it uniformly.
Thanks for chip view. I can clearly see the capacitor although not the connection to it. There looks like epoxy under the chip with some gold metallization underneath. I don't see any wire connections to the chip. Perhaps there is a new way in use that I don't know.
I remember Kapton and when it first came out for flexible circuits. We tried using polyimide for an interlayer dielectric but didn't work out too well. Interestingly we used thermally cured photoresist, which is still in use for that product today, 20 years later.
I didn't realize that hot melt was used that much. I have some remote to-92 sensors that I finally put a blob of hot melt on and works excellently.
I think we're getting off topic for the thread and perhaps continue it off line.
Paul
Conformal coating can be built up with multiple layers, enough to bury those sharp edges. I prefer silicone resin, for its flexibility. Moisture can penetrate the conformal coating, but if the surface is squeaky clean, the moisture won't find a conductive path between the circuit conductors. Solder mask on a circuit board is important too. An unmasked board is much harder to protect. Epoxy is an option too, but it is rigid, so thermal expansion and contraction can break components or lift them right off a circuit board. It is better to use a flexible coating, or a flexible coating between the components and the epoxy. In the old days, the preferred coating was wax with a fungicide.
What is an SI2007?
The SHT2x is not too hard to deal with. It is leadless, but only 6 pins and a fairly wide pitch. I do have solder paste, stencil and a reflow oven to solder it uniformly.
I still lose partly written responses. I had been in this sort of business for quite a while (age 76) but I sure don't remember wax coatings. I agree with your comments re:coatings. My only concern with silicone coatings is that silicone rubber is permeable to water but if its a lacquer type that would work fine, too.
I misquoted the chip number. It's a Silicon Labs 7005 humidity sensor. I bought from China through Tindie and it's mounted on a break out board too with associated components
I didn't realize that the conversion code for the SHT-21 was different from the SHT-11. I thought they were pin compatible and code compatible too. There's a little more work for me. As you have seen I plan on using the I2C protocol, bought parts, etc. I have on order several break out modules with the SHT-21 from China. I have no idea when they will arrive; it's been about two weeks so far. Also ordered are a few SI2007 sensors. I did buy a bidirectional level shifter and will use just a r/2r voltage divider as the current draw is so small. It seems wasteful to me to use a 3.3 V regulator with a load resistor to draw enough current t get the regulator to work.
Just to compare in red and blue, the formulae for the SHT2x:
°C = -46.85 + 175.72 * rawT / (2^resT) resT is 14 bits default %RH = -6 + 125 * rawH / (2^resH) resH is 12 bits default
res = resolution, tradeoff with speed, set in the user register.
with formulae for SHT1x: °C = -40.1 + 0.01 * rawT on 14 bit setting and 5V supply %RH1 = -2.0468 + 0.0367 * rawH - 1.5955E-6 * rawH^2 on 12 bit setting
then temperature compensation %RH = (°C - 25) * (.01 + 0.00008 * rawH) + RH1
The SHT1x too has both high and low resolution modes.
The Si2005 looks like an interesting RH sensor. I'll be interested to hear how is stacks up. I'm not clear on what you mean by, "I did buy a bidirectional level shifter and will use just a r/2r voltage divider as the current draw is so small. It seems wasteful to me to use a 3.3 V regulator with a load resistor to draw enough current t get the regulator to work."
Just to compare in red and blue, the formulae for the SHT2x:
°C = -46.85 + 175.72 * rawT / (2^resT) resT is 14 bits default %RH = -6 + 125 * rawH / (2^resH) resH is 12 bits default
res = resolution, tradeoff with speed, set in the user register.
with formulae for SHT1x: °C = -40.1 + 0.01 * rawT on 14 bit setting and 5V supply %RH1 = -2.0468 + 0.0367 * rawH - 1.5955E-6 * rawH^2 on 12 bit setting
then temperature compensation %RH = (°C - 25) * (.01 + 0.00008 * rawH) + RH1
The SHT1x too has both high and low resolution modes.
The Si2005 looks like an interesting RH sensor. I'll be interested to hear how is stacks up. I'm not clear on what you mean by, "I did buy a bidirectional level shifter and will use just a r/2r voltage divider as the current draw is so small. It seems wasteful to me to use a 3.3 V regulator with a load resistor to draw enough current t get the regulator to work."
Thanks for the info on the conversion differences.
What I plan to do with the SHT-21 (3.3v) is just to use a r/2r divider to power it. If I draw less 1 ma say, and set the resistance total to draw 10 ma, then the 3.3 supply won't drop below 3v. I think I explained about the level shifter. Correct me if I'm not correct. The I/o pins on the stamp have pullups to 5v which wouldn't work, would they? PJ Allen mentioned that to me earlier.
Paul
I just received the SHT-21 break out boards and posted the photos I think. The board is about 10 by 13 mm.
That board already has all of the i2c level shifters and a 3.3V regulator installed. So, wow! No worries for you; the interfacing advice is moot. Arduino and Stamp, compatible 5V tolerant i/o. I went to search for SHT21 on ebay, and the first hit is what appears to be your board. Please read the description there below the item.
That board already has all of the i2c level shifters and a 3.3V regulator installed. So, wow! No worries for you; the interfacing advice is moot. Arduino and Stamp, compatible 5V tolerant i/o. I went to search for SHT21 on ebay, and the first hit is what appears to be your board. Please read the description there below the item.
Yes, I see that now. I don't think that's who I bought it from but it is identical. It was perhaps the other sensor board that needed the level shifter. I am the original space cadet and get confused a lot. I didn't realize that they could make 3.3 v regulators that small. If I fire it up at 5 v and it blows will you buy me another one? ;-) It'll take me a few days to get this thing running and I'll report then.
How are you running your board? Level shifters, regulators, etc.?
I've run SHT21 only with the Propeller, so it is 3.3V all the way through. Actually I've not used the SHT2x for much. For a variety of reasons, I've stuck with the SHT1x and SHT7x. I like the form factor of the SHT7x particularly, for its rapid response.
If you fire your board up at 5V with correct polarity and it blows, I'll do something to make it up to you to get the RH up and running!
There are some results to show on programming my SHT-21 breakout board. The weekend was full; life interferes.
Although several vendors advertise the same board, some say 3.3V and some 5V. No wonder my confusion. It is pretty sure a 5V part. I.e., it didn't blow up
.
Except for the one I powered backwards. Minus one sensor
.
The second was an interesting experience. Wired correctly it sent back curious results. Debug output appeared with very low numbers and a strange ASCII character. I soon realized I was only reading the first byte of word data. I dropped the ASCII and used hex/binary. Learning how to format debug output data. Later data was correc after reading two bytes. More info to follow.
At the moment temperature data is not right and is all 1's
.
Programming is a hack, in the old sense of the word. Just a few I2CIN and I2COUT commands. There's no point in reproducing it. All the command information is in the Sensirion data sheet. The first command is a soft reset, thena RH trigger to generate a RH response in the device and then a generic read to get the data. All the commands have a one second delay. The delay can be reduced I'm sure. The commands are "no hold masater" as the one second delay obviates it in this case (I think.)
As far as interpreting the data, the data conversion requires a calculation of a fraction less than one which of course doesn't work. I suspect Tracy Allen's work on calculations has the answer but I don't care as this is a <controller> not a computer (calculating machine.) As I mentioned in previous postings, I am using Python to run the Stamp, do calculations, display the data, store in a database and and post to my website. I strongly reccomend Python. It's much simpler than C on an Arduino, runs on Windows, Mac, and Linux and has a huge helpful community
.
For now I will be working on the temperature issue and installing the sensor in my greenhouse for a week or two. I plan on using hot melt glue and shrink tubing for corrosion protection. I also bought some Gortex and plan on glueing a small piece over the sensor opening. There's no provision for a protective cap. There's also a description of an assembly for protecting a submersed sensor somewhere in the Google cloud that will be instructive.
Good update, Paul. As always, the simple things seem to take the longest to figure out. One thing I noticed about the eBay auction page is that it never actually says it is a Sensirion sensor. It uses the same part number of SHT21, but does it have markings on top of the body of the chip per Sensirion's datasheet? I ask because if these are knockoffs or gray market Sensirion parts, you may have odd issues that are not related to your code.
Tracy, I still have about a dozen SHT21's and even some filter caps. Any chance I can buy some PCBs from you so I can use them?
Hi Andrew,
The temperature reading is now correct. Left out a command or two. Double checked by reading at RT and finger heating and breathing into he sensor.
The device looks like a real SHT21. It reads, SHT21 and underneath D1BMS, laser engraved. It would be virtually impossible to reproduce the chips especially with the same equations as the thickness, processing and quality of the capacitor dielectric would impossible to reproduce.
Off to the next step in a few days. Off to take grand son to music lesson this afternoon and who knows about tomorrow.
Good news, glad you got it working. After I put the comment about a knockoff, I did think that it would be a bit more effort than it's worth, but gray market devices would be viable depending on where Sensirion has the devices manufactured.
We dealt with a component two years ago that fell into that category. Our customer supplied some parts for a build that they purchased off eBay or alibaba or some other non-standard channel. We built the boards, but when they went to test them, every unit failed out of spec. Through their troubleshooting, they found that every one of the suspect devices did functionally work, but well outside the specified limits. When they went through the routes with the manufacturer, it turned out that the particular lot codes were supposed to be discarded because the entire lot would not pass the the manufacturing tests. Apparently, someone at the factory decided to sell them as good parts instead of disposing them and they made it to the states. Your comments reminded me of that scenario so that is why I asked.
Have fun and keep us posted. Once you get this project done, how about setting up a Raspberry Pi with it's camera for time lapse photography to capture blooming: http://vimeo.com/66968162
Tracy, sounds great. I will PM you with an address.
ps. I removed my place of business from your post since I don't mix business with pleasure, so to speak and my work has policies regarding "online representation of the company" and I don't want to have to add the typical "opinions of mine are not necessarily those of company x" blah, blah, blah, to each of my posts.
Good news, glad you got it working. After I put the comment about a knockoff, I did think that it would be a bit more effort than it's worth, but gray market devices would be viable depending on where Sensirion has the devices manufactured.
We dealt with a component two years ago that fell into that category. Our customer supplied some parts for a build that they purchased off eBay or alibaba or some other non-standard channel. We built the boards, but when they went to test them, every unit failed out of spec. Through their troubleshooting, they found that every one of the suspect devices did functionally work, but well outside the specified limits. When they went through the routes with the manufacturer, it turned out that the particular lot codes were supposed to be discarded because the entire lot would not pass the the manufacturing tests. Apparently, someone at the factory decided to sell them as good parts instead of disposing them and they made it to the states. Your comments reminded me of that scenario so that is why I asked.
Have fun and keep us posted. Once you get this project done, how about setting up a Raspberry Pi with it's camera for time lapse photography to capture blooming: http://vimeo.com/66968162[/QUOTE
About bad components, a large motherboard manufacturer used poor quality capacitors and gave Dell headaches for a long time. It can be a real nuisance to say the least. When I was working we were sputtering magnetic alloys and got a target from a well known supplier and after not being able to make it work for a long time, we found out it was a terribly wrong composition. Etc.etc.
I'll post here occasionally as things progress. There are the other sensors I bought to test, too.
The videos are interesting but enough to get engaged in. However if you want to set it up here let me know. I do have some stereo photos of orchids that I think are neat, too.
Paul
Comments
As for taking it apart, it doesn't take much to pop the cap off as my work had one happen a few years ago when the board was dropped onto the edge of a bench. So, if you do decide to open it up, here is a reference picture of one that only went through pick-n-place and reflow then was busted open.
As for using hot melt to "seal" bare metal, you would be surprised at how much it is used in PCBA manufacturing by certain companies. At work, we have one particular customer that uses enough for us to justify the purchase of an industrial hot melt gun. I was a bit surprised when I was handed a purchase req for signoff approval for a "$480 Glue Gun" for that one......
What is an SI2007?
The SHT2x is not too hard to deal with. It is leadless, but only 6 pins and a fairly wide pitch. I do have solder paste, stencil and a reflow oven to solder it uniformly.
I remember Kapton and when it first came out for flexible circuits. We tried using polyimide for an interlayer dielectric but didn't work out too well. Interestingly we used thermally cured photoresist, which is still in use for that product today, 20 years later.
I didn't realize that hot melt was used that much. I have some remote to-92 sensors that I finally put a blob of hot melt on and works excellently.
I think we're getting off topic for the thread and perhaps continue it off line.
Paul
I still lose partly written responses. I had been in this sort of business for quite a while (age 76) but I sure don't remember wax coatings. I agree with your comments re:coatings. My only concern with silicone coatings is that silicone rubber is permeable to water but if its a lacquer type that would work fine, too.
I misquoted the chip number. It's a Silicon Labs 7005 humidity sensor. I bought from China through Tindie and it's mounted on a break out board too with associated components
Regards
Paul
Just to compare in red and blue, the formulae for the SHT2x:
°C = -46.85 + 175.72 * rawT / (2^resT) resT is 14 bits default
%RH = -6 + 125 * rawH / (2^resH) resH is 12 bits default
res = resolution, tradeoff with speed, set in the user register.
with formulae for SHT1x:
°C = -40.1 + 0.01 * rawT on 14 bit setting and 5V supply
%RH1 = -2.0468 + 0.0367 * rawH - 1.5955E-6 * rawH^2 on 12 bit setting
then temperature compensation
%RH = (°C - 25) * (.01 + 0.00008 * rawH) + RH1
The SHT1x too has both high and low resolution modes.
The Si2005 looks like an interesting RH sensor. I'll be interested to hear how is stacks up. I'm not clear on what you mean by, "I did buy a bidirectional level shifter and will use just a r/2r voltage divider as the current draw is so small. It seems wasteful to me to use a 3.3 V regulator with a load resistor to draw enough current t get the regulator to work."
Thanks for the info on the conversion differences.
What I plan to do with the SHT-21 (3.3v) is just to use a r/2r divider to power it. If I draw less 1 ma say, and set the resistance total to draw 10 ma, then the 3.3 supply won't drop below 3v. I think I explained about the level shifter. Correct me if I'm not correct. The I/o pins on the stamp have pullups to 5v which wouldn't work, would they? PJ Allen mentioned that to me earlier.
Paul
I just received the SHT-21 break out boards and posted the photos I think. The board is about 10 by 13 mm.
Yes, I see that now. I don't think that's who I bought it from but it is identical. It was perhaps the other sensor board that needed the level shifter. I am the original space cadet and get confused a lot. I didn't realize that they could make 3.3 v regulators that small. If I fire it up at 5 v and it blows will you buy me another one? ;-) It'll take me a few days to get this thing running and I'll report then.
How are you running your board? Level shifters, regulators, etc.?
Paul
I've run SHT21 only with the Propeller, so it is 3.3V all the way through. Actually I've not used the SHT2x for much. For a variety of reasons, I've stuck with the SHT1x and SHT7x. I like the form factor of the SHT7x particularly, for its rapid response.
If you fire your board up at 5V with correct polarity and it blows, I'll do something to make it up to you to get the RH up and running!
Although several vendors advertise the same board, some say 3.3V and some 5V. No wonder my confusion. It is pretty sure a 5V part. I.e., it didn't blow up
.
Except for the one I powered backwards. Minus one sensor
.
The second was an interesting experience. Wired correctly it sent back curious results. Debug output appeared with very low numbers and a strange ASCII character. I soon realized I was only reading the first byte of word data. I dropped the ASCII and used hex/binary. Learning how to format debug output data. Later data was correc after reading two bytes. More info to follow.
At the moment temperature data is not right and is all 1's
.
Programming is a hack, in the old sense of the word. Just a few I2CIN and I2COUT commands. There's no point in reproducing it. All the command information is in the Sensirion data sheet. The first command is a soft reset, thena RH trigger to generate a RH response in the device and then a generic read to get the data. All the commands have a one second delay. The delay can be reduced I'm sure. The commands are "no hold masater" as the one second delay obviates it in this case (I think.)
As far as interpreting the data, the data conversion requires a calculation of a fraction less than one which of course doesn't work. I suspect Tracy Allen's work on calculations has the answer but I don't care as this is a <controller> not a computer (calculating machine.) As I mentioned in previous postings, I am using Python to run the Stamp, do calculations, display the data, store in a database and and post to my website. I strongly reccomend Python. It's much simpler than C on an Arduino, runs on Windows, Mac, and Linux and has a huge helpful community
.
For now I will be working on the temperature issue and installing the sensor in my greenhouse for a week or two. I plan on using hot melt glue and shrink tubing for corrosion protection. I also bought some Gortex and plan on glueing a small piece over the sensor opening. There's no provision for a protective cap. There's also a description of an assembly for protecting a submersed sensor somewhere in the Google cloud that will be instructive.
Paul
Tracy, I still have about a dozen SHT21's and even some filter caps. Any chance I can buy some PCBs from you so I can use them?
The temperature reading is now correct. Left out a command or two. Double checked by reading at RT and finger heating and breathing into he sensor.
The device looks like a real SHT21. It reads, SHT21 and underneath D1BMS, laser engraved. It would be virtually impossible to reproduce the chips especially with the same equations as the thickness, processing and quality of the capacitor dielectric would impossible to reproduce.
Off to the next step in a few days. Off to take grand son to music lesson this afternoon and who knows about tomorrow.
Paul
We dealt with a component two years ago that fell into that category. Our customer supplied some parts for a build that they purchased off eBay or alibaba or some other non-standard channel. We built the boards, but when they went to test them, every unit failed out of spec. Through their troubleshooting, they found that every one of the suspect devices did functionally work, but well outside the specified limits. When they went through the routes with the manufacturer, it turned out that the particular lot codes were supposed to be discarded because the entire lot would not pass the the manufacturing tests. Apparently, someone at the factory decided to sell them as good parts instead of disposing them and they made it to the states. Your comments reminded me of that scenario so that is why I asked.
Have fun and keep us posted. Once you get this project done, how about setting up a Raspberry Pi with it's camera for time lapse photography to capture blooming: http://vimeo.com/66968162
Andrew, I'll send you some boards (free). Three strips of 5 boards each on 0.031" material. Can I mail them to you at <<your work>>?
ps. I removed my place of business from your post since I don't mix business with pleasure, so to speak and my work has policies regarding "online representation of the company" and I don't want to have to add the typical "opinions of mine are not necessarily those of company x" blah, blah, blah, to each of my posts.