Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
The DS18B20 is very likeable solution and it has it all. Those Max people do not miss very much, do they.. Somwhere on another thread I do remember going through pages of sensors and the 1pin solution is very tempting, but originally I have dismissed it because it does require the extra i/face to convert the serial digital information an all the extras to address the chip. I would be quite inclined to look at this solution in the environment of "many" sensors - it is a very elegant option and in line with today's trend to go PIC (they're quite affordable too).
In this project, there is just one temp sensor and I do feel that LM35 would give me the resolution that I want maintaining the ease of interfacing to the rest of (btw - analog) circuitry.
The next obstacle now seems to be the mechanical construction of the sensor itself. The LM35 has an excellent response time when immersed in bath but quite long in the air. As the sampled environment is a "running past" vapour/steam, I need to enchance the propagation of the temperature to the sensor.
Some kind of a "heat-sink" on top of the TO-46 case will improve the propagation but it can not be very bulky because this in turn will introduce inertia. Little bit of a thin copper sheet will do the job, perhaps a "snap-on" heatsink.
The same goes for the pipe the steam goes through. I want to measure the temp of the steam, not the pipe - so the sensor must be thermally isolated from the pipe. This is not a big problem, some teflon here - will do the job.
By the time I am done with this sensor, it will probably "cost" more than a COTS sensor I did like on digikey but I don't mind to do some experimenting here and this would only be a "mechanical" challenge. It will in turn give me some more insight to the "inertia" issues.
There are many COTS sensors that I would like and could use but their prices are just - to scary. I know, you would say that I only get what I pay for and this is truth, but I do trust - some experimenting here could take me a long way.
A spiral of tightly wound Copper wire around the T-46 casing of the sensor spreading out in a "cone - like" fashion would not be very obstructive to the flow of the steam and certainly helpfull to propagate the heat to the sensor.
I would ofcourse require some means to attach that heatsink after the sensor is installed and this the case indeed.
In those terms, the "sensing" applies more than the "measurement" Eric.
This is however hard to distinguish because the accurate response from the sensor will be compared against the presettable value, giving therefore as good indication of the measurement, as if it was displayed.
I do trust that an accurate calibration of the 100 deg C of the sensor's response versus a set point of a 10 turn pot with the "null error" will promulgate to other "preset" points of the pot within the range required, and - as the temperature will "follow" the pot setting, the lack of an "error" will indicate that a particular temperature has been reached as per the pot's scale. However indirectly, the sensing becomes thus an expression of the measurement.
As I did mention before, I am far more concerned now with the inertia of the "sensor" against the inertia of the "plant". As there is not much I can do to improve on the plant's performance - I do need to ensure the best response from the sensor in order to maintain the integrity of the project.
When you are required to deal with a process having a considerable time delay (heating), you would not like to contribute to make that delay increased by the means of a sensor that is just simply to slow in response.
As for the thermocouples, the answer is no. The temperature range of the process is to small and so is the response from thermocouples versus the price.
The CA3140 has been around for a long time to replace the 741 (if my memory serves me right). Well, interfacing of LM35 is something I do need to think about, it has been no more than 24 hours since I have made the decission to go this way. I will run some simulations and will certainly let you know of the outcome Eric.
I will cut out the PIC, the AO will do me. I am bit oldfashioned here. Since my arrival on this forum I have been told on at least one occasion to get real with those PICs. I can see the advantage and simplicity in programming, just not ready to attack it yet. I feel the need for some "PIC for dummies" tutorial before I can endavour on a project that can implement them, it has the required apperance of simplicity and oddiness at the same time..
PICs are a very modern trend to "attack" just about any problem there is. What the PLL and DSP have achieved in communication, those PICs seem to be doing in all others and - they are also cheap and no argument here.
Yet, having a longstanding involvement with industry in the maintenance sector, I had not seen much of the evidence of PICs making it through. I could easily assume that working with the old technology is the reason.
But it is not the case.
The single most positively appealling to me "revolution", was the introduction of "reflective memory" as the means to make several systems speaking a quite different language (Unix, PCDOS, Windows and others), to communicate together and exchange information without the requirement to employ sophisticated "translators" to achieve this task.
Multimillion companies do provide sollutions/improvements worth millions of $ still using "old" methods. They are simpe for the purpose of this message; if it is complicated - it is done in sofware, if it is not - a hardware sollutions (digital/analog). No PICs. System of AO/AIs and DO/DIs does provide a very simple answer to the question why PICs have not found its way through. But this moment will probably come and I would like to be here when it happens.
This is quite outside this thread, sorry.
The temperature range is enclosed within 60 and 100 deg C Eric. I am pretty sure you have figured this out already. A vapour is a steam, it all depends on the temperature of the boilling point.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
but I don't mind to do some experimenting here and this would only be a "mechanical" challenge. It will in turn give me some more insight to the "inertia" issues.
