wireless thermocouple circuit

[mekitron]

Hello all,

I want to design a wireless thermocouple circuit which will meausre internal human body temperature and transmits the readings to an external reciever. I want some ideas concerning this design and if someone already worked or working on something like this then please share your engineering.

Thank you.

 

[betwixt]

Is there a particular reason why you choose a thermocouple for measurement? They can be tricky to interface and require additional circuitry to amplify and compensate their measurement. They also produce an analog output which makes it more difficult to transmit.

I build water temperature monitoring equipment covering a similar temperature range (-10C to +50C) using 1-wire sensors and simple radio protocol, it's accurate to 0.125 degrees and has a range of about 50 metres.

Brian.

 

[mekitron]

Thermocouple is just a wire like temperature senesor. I selected it because of size. To meausre "human internal body" temperature, I need very small system and I think it is only thermocouple which can solve my problem. How small was your 1-wire sensor system? If possible explain your system to me and suggest me what I'd do. I understand the complexties of thermocouple's anologue output but right now there is no choice to select anyother sensor (I think).

 

[betwixt]

A thermocouple or a tiny bead thermistor are the smallest temperature sensors. The thermocouple will only produce a few mV which will be small compared to other probes but that could be an advantage for safety reasons. A thermistor in a bridge or potential divider configuration will possibly give volts of output so less, if any, amplification might be needed. I suppose the real answer is it depends on where the temperature is being measured, if it's skin temperature use a larger but more friendly probe, if it is deeply invasive in the body, use a thermocouple.

The probes I build are using T0-92 devices so they are too large to be inserted inside someone but there are smaller SMD devices which might be small enough. Mine are wired to a 3mm x 3mm microcontroller which converts the temperature reading into a serial data stream and the whole thing is encased in clear resin 'torpedo' about 10mm long by 5mm diameter.

Brian.

 

[mekitron]

Brian,

Thank you very much for your guidelines.
I did search on different devices. I want my complete circuitry (with in) 9mm to 11mm of diameter. May you suggest me some smaller SMD devices at that level? I want to design the circuit as much simple as possible. I am not strong in circuit designing so can't design complex circuits. I do not know what will be the design of transmitter and receiver to find the temperature. Should I follow fm transmitter like designs?

Thank you,
Meki

 

[betwixt]

I would advise you t use FM transmitters rather than AM, especially in critical applications like body temperature measurment. AM is more prone to interference which could result in random or inacurate readings. FM is a little more complicated because you have to encode your transmitted data and decode it in the receiver but the readings are likely to be more reliable.

It is quite possible to design in the size you want but first please explain exctly where these will be when in use, are they measuring internal or skin temperature? There may be problems with the battery and signal propagation if it is for internal use.

Brian.

 

[mekitron]

Brian,

You are right at this point that there'll be energy problems. I'd like to confirm that it's about internal human body temperature measurement. The device will go in stomach, will meausre the temperature. Batteries will be discharged and device will come out as everything come out. I am sorry if you feel it funny but this is really what I wanted to do. Can you give me some wide directions to get started?

Regards,
Meki

 

[betwixt]

I'm not an expert in medical electronics but I see no reason why thermocouple (complicated) or thermistor(simple) can't be used as long as they are coated in an insulating layer or encased in a thin plastic sleeve. The transmitter has to be outside the body anyway so it's signal can be picked up. Body tissue is a good signal screen! For simplicity, I would meaure the temperature as a voltage, amplified if necessary, then send it to the transmitter with 'Manchester' encoding. The receiver can then recover the voltage reading and use math to convert it back to temperature. Doing the conversion in the receiver lets you calibrate it in software to get an exact result without having to adjust anything at the 'body' end.

Brian.

 

[Pha5e]

I read of a pill system a while back which the patient swallows. It takes measurements inside the body including the temperature and then sends them via a built-in transmitter to a receiver outside of the body. The pill containing the electronics is recovered when the bowels are emptied and can be used again?

 

[mekitron]

@Brian,

Can you explain "I would measure the temperature as a voltage, amplified if necessary, then send it to the transmitter"? How you'll "send" the measure voltages to an external transmitter? Brian, I think you just assumed the "send" property of device. Just imagine, it reached in stomach and if it is only encased in an insulating material how it'll "send" the measured voltages outside the human body? If anyway it sent then why need a "transmitter to receive" and receive it again using a receiver? If use a thermistor then how it can be simple?
-Meki

- - - Updated - - -

@Pha5e:

I think it cannot be used again. It'll be great if you share the link of that pill. Thank you.

-Meki

 

[betwixt]

What I mean is there are no devices that give an analog reading like a glass thermometer would and even if such a thing existed it would be impossible to read while inside a body. All sensors return a representation of the temperature as a voltage or directly in digital form. You cannot easily transmit a steady analog voltage by wireless link but it's relatively simple to transmit numbers in binary (= digital) format. So if you use a thermocouple or thermistor it won't return a temperature but a voltage, in the case of a thermocouple it will need considerable amplification and some kind of ice-point compensation circuit. Once you have a voltage that is of usable range, you have to convert it to digital format using an ADC before it can be transmitted.

By 'sending' I mean the wireless link. You have to transfer the measurement to a receiver where the data can be processed. You might be able to do it using near field techniques in which case you might be able to power the sensor externally, otherwise you have to include a battery inside it or place the probe inside the body by wire it to a transmitter carried outside the body.

The reason I think a thermistor may be more appropriate is to do with the way ADCs work. Typically they use 10-bit resolution which means each measurable step is ((2^10) -1)/Vr Volts where Vr is a reference voltage setting the full-scale of the measurment. I assume you are only expecting to measure temperatures between say 30C to 40C so ideally you want the numeric range to cover those extremes with highest accuracy. A typical 5V ADC would have a resolution of about 4.88mV per count which equates to several degrees change in voltage from a thermocouple but with suitable series resitor could be a fraction of a degree from a thermistor. In other words, without a dedicated thermocouple interface circuit, you may not be able to measure body temperature to within less than maybe 3 or 4 degrees!

I cannot help you if you want to put the whole thing inside a swallowable 'pill', I have no experience in that field and I'm cautious of the damage wrong advice can do. You really need to speak to a company with experience of such complicated and potentially dangerous devices.

Brian.

 

[Pha5e]

Here is details of a swallowable pill which resists damage from stomach acids:

https://voices.yahoo.com/intelligent-pill-dispensing-internal-medication-2202248.html

 

[mekitron]

Thank you guys, I am still working on this temperature system. I'll update this thread soon. Stay tuned on EDA with me. I'll need help. I'll share my further ideas. I am not going to give this kind of pill anyone, however the size of temperature sensor will be almost like Pha5e's referred pill.

-Meki

 

[mekitron]

This project is alive and going on. Stay tuned with the thread.

 

[kam1787]

instead of the stomach, have you considered other ......points, such as renal measurement or underarm.

 

[mekitron]

No, not at all. It is time taking project to design such a complex bio-medical device. Let me conquer stomach first then travel to next destinations. Do you have any suggestions?

 

[Alan Garcia Sanchez]

I have a question, as it links the wireless signal to an Arduino microcontroller, or a raspberry pi, I just want to measure the temperature of the surface of the skin of a cow, either putting an earring in the ear of the cow or wearing a stethoscope and place the sensor and send data wirelessly, only temperature, which type of sensor is best suited?, lm35 or thermocouple?, and communication protocol? xbee or RF

 

[KlausST]

Hi,

I didn't read the folowing point before in this thead...
A thermocouple is a relative measurement method. It measures the difference of the junction temperature and the so called "cold junction". To measure absolute temperatures you need an additional temperature sensor and a very low offset amplifier. Therfore the size of thermocouple circuit always is bigger than other temperature sensors.
For me the benefit of a thermocouple is mainly the temperature range...

Temperature measurement with earring.
Because the low flow of blood in the ears i think the temperature of the ears is influenced by air temperature, humidity, airflow and direct sunlight.

Klaus

 

[Alan Garcia Sanchez]

In this case taking into account the variables that can affect the measurement due to external factors such as humidity, air flow and direct sunlight, would have to measure invasively and thus have an accurate reading of the animal, could isolate, using a collar to protect the sensor and measure directly, or another feasible alternative should be used.

 

[mekitron]

Quite hard................. I left the project for someone else.