Embedded project needs devices which can work continuously at +125 degree C

[baileychic]

Embedded project needs devices which can work continuously at +125 degree C

I am planning to use these devices other that other passive devices like capacitors, resistors.

1N4007 Vishay make
PIC18LF46K22 - Microchip
LM1086-3.3
LM1086-ADJ
ADG726
MAX31855KASA
Leds
P65SMT101 switch
7 Segment display Common Cathode Red (2 digits , make not yet decided)
ESP8266-12E
8x K type thermocouples probes with 1.75 mtr max probe length
Xbee or LoRa

The device needs to send 8x thermocouple data once a while (timer settings based on BCD switches, one for minutes and another for seconds) to a PC for logging through LoRa.

I want to know if above devices can work continuously at +125 degree C.

The K type thermocouple will be used to measure temperatures in the range 0 to 500 degree C.

A peltier is also used to charge a Li-Ion 3.7V battery and the battery powers the circuit.

If one or more of the above devices cannot work at +125 degree C continuously then which other devices can I use ?

This is an urgent requirement.

Find the attached circuit which I have designed.

 

[Easyrider83]

Commercial: 0 °C to 85 °C.
Industrial: −40 °C to 100 °C.
Automotive: −40 °C to 125 °C.
Extended: −40 °C to 125 °C.

 

[FvM]

I think, you should start to work out the power supply concept. Operating Li-Ion at 125 °C is just a fiction.

 

[baileychic]

@ EasyRider and FvM

Thanks.

So, I can use Automotive/Extended series devices without any issues ?

@ FvM

Client still have to finalise the battery.

 

[betwixt]

Good luck!
I think you will have other issues, for example applying the cold junction compensation after the multiplexer. For best accuracy you need to use the MAX device at each thermocouple and multiplex the measurements afterwards. Apart from the likelihood of not compensating correctly, using a MUX to carry tiny (mV) signals is prone to problems.

Why the reset circuit on the ESP8266? If it really needs resetting it makes more sense to drive it from the PIC. I too have concerns over the power supply, not only because of the battery temperature but the use of a Peltier as the prime power source. it might work but you would have to do some serious testing first, remember that Peltier devices produce a voltage when there is temperature difference across them, if both sides are the same temperature they do nothing.

Brian.

 

[baileychic]

@ betwixt

Okay, What kind of problems with Thermocouple interfacing circuit ?

LiSoCl battery is used.

 

[baileychic]

Is the circuit okay now ?

- - - Updated - - -

Couldn't find any RF modules like XBee or LoRa which can operate continuously at +125 degree C. Please mention few. Also mention which ceramic antenna I can use. I prefer Antenova antennas. Antenna also should be operable at +125 degree C and so I prefer ceramic type.

 

[betwixt]

Okay, What kind of problems with Thermocouple interfacing circuit ?

The new design looks far better.
There are two main problems in the original design:
1. The MAX31855 incorporates it's own temperature sensor which is used to offset the thermocouple output voltage, that is how cold junction compensation works. The junction it compensates is the location the thermocouple wires (or special connector) meets the IC. You have to remember that the thermocouple effect happens when two dissimilar metals come into contact so there are two junction places in the circuit, the one at the thermocouple tip and the one where the thermocouple connects to the IC. The compensation removes the latter so only the tip voltage is considered. If you wire the thermocouple to a mux, you generate a voltage at the mux input which the MAX31855 is physically removed from and so cannot compensate for.
2. A thermocouple produces a very small voltage, it needs considerable amplification before it can be read by an ADC. The mux will inevitably introduce a small voltage offset at it's output. even if it is only a few uV it will upset the measurement after amplification.

Your new design should handle the cold junction problem and by moving the multiplexing into the digital domain it removes the mux offset problem

Brian.

 

[dick_freebird]

First, you should distinguish "operate" from "operate forever"
(reliability) and "operate exactly to spec" (performance).

If there is offered a "HiRel" equivalent (-55C to 125C on the
datasheet) of the same part, then die level reliability is
adequate and you only have to worry about package
integrity. An "industrial" hermetic part would be a good
bet for high humidity, but absent moisture ingress concerns
even a plastic package can last for some while.

Now parts like regulators / references may be "cherry
picked" in production to net the extended-temp-range
population for the higher-value parts, meaning that the
endpoint accuracy of the supply may be a bit out of
spec on "commercial" bottom-grade parts. But if your
"1%" Vreg becomes "2%" at high temp, and you are still
feeding parts that demand 5% tolerance, what's the
harm?

Peltier modules **** for efficiency and need really good
cooling to throw any current.

Tadiran makes some extreme-environment lithium primary
cells and rechargeables, I see them in military electronics
type magazines all the time. You might consider whether
you'd be better served by replacing the Peltier assembly
with an equal volume of lithium primary cells, and holding
the supply current to a minimum by some combination of
standard and "heroic" measures.

 

[andre_luis]

Are you serious? Do you think that it is reasonable to put here a project part list expecting to thave an answer that could be obtained by a simple research in parametric search tools at supplier's webpages ? One thing is one look for passive or single devices able to operate at this temperature; another thing is to use complex devices that were not primarily designed to operate in extended temperatures. Furthermore, as usual you did not present the complete scenario, such as whether this temperature is environmental or if it is steady inside a case (e.g without internal forced ventilation, to circulate the heat), and in this case, instead of looking for devices capable of working in that range, the suited action would be to put the focus in the heat exchange between inside the case and the external environment, or in other words, decrease the temperature at which all these components are subjected. In this context it would make sense to think of 'peltier' as a means of extracting the heat from within and throwing it out.

- - - Updated - - -

By the way, with respect to the singular "125°C" requisite, your project has issues with regard to the design of the power supply stage, which being not of any switched topology, will be burning all the excedent energy between the 12v and 3v3 in the form of pure heat.

 

[baileychic]

@ andre_teprom

The +125 degree C is environment temperature where the device is placed. The 12V is not actual 12V. I just used it as reference for simulation when I was using a different LDO in simulation.

Here is the datasheet of the Peltier. The Peltier hot side temperature will not exceed 400 degree C and Maximum Load Output Voltage will not exceed 5V. I am only dropping max 5V to 4.2V for battery charging.

The cold side of Peltier is cooled with a large heatsink. The Peltier and its heatsink is mounted away from device. The device will be enclosed in dust and water proof enclosure.

 

[betwixt]

The cold side of Peltier is cooled with a large heatsink. The Peltier and its heatsink is mounted away from device.

More details needed:
The power is produced when there is a temperature difference across the device, a heatsink in the same environment will adopt ambient temperature so there will be no difference and no power. If the Peltier is mounted away from the heat source, where does the heat it needs come from? If indeed it is mounted in a cooler place, why not move the electronics there too?

Brian.

 

[andre_luis]

The +125 degree C is environment temperature where the device is placed

The short advice to your daydream is, forget about this; don't waste your and our time thinking of putting your system to work within an oven. You will be lucky if it simply don't work; it is likely to burn, or even to take fire; keep in mind that you are adding to the ficticious environment temperature (125oC), the self heating of the components (e.g junction temperature on silicon devices), in addition to the accumulated heat confined within the case, which will cumulately at the end subject them to an actual temperature much higher than any standard temperature range, whatever it is, industrial, extended, military, etc...unless you're willing to deal with a life expectancy of your board of few seconds.

 

[baileychic]

Which RF modules like XBee or LoRa I can use for sending thermocouple data to a PC (Computer) at a distance of not more than 50 meters ? The RF module should also works at +125 degree C. If RF modules doesn't work at that temperature then can I use ESP8266-EX ?

If yes, Would the coding for ESP8266 will be same as general purpose ESP8266 ?

We will be using headers between thermocouple probe and MX31850. Will this affect the thermocouple readings in any way ?

The RF module or WiFi module should be able to use patch or ceramic internal antenna. We cannot provide external antenna on the enclosure.

 

[andre_luis]

If yes, Would the coding for ESP8266 will be same as general purpose ESP8266 ?

You probably had the curiosity to search for information, besides the microcontroller itself, of other components of the module, such as the flash memory mounted there, and if you look at the datasheet you will realize that it is specified to a limit temperature less than 100 degrees celcius.

AGAIN: One thing is you subject a system where some devices are designed to operate at 125oC as its internal temperature, other thing is consider that these devices has their own self heating, as for example, the ESP8266 core, according to its datasheet can operate at 125oC, but even at a normal temperature environment e.g 25 oC, you can feel with your fingers that it operate hot, what means that at external 125oC temperature this part would reach much more higher temperature.

 

[betwixt]

...and yes, there is a voltage generated at each metal-metal junction, that's why the cold compensation should take place as close to the thermocouple connection as possible. The connection itself creates a voltage which can be greater than the voltage at the thermocouple tip so you have to be very careful how it is wired. Normal connectors are unsuitable.

An ESP8266 is a programmable MCU so any module using one will work but some do not bring all pins to the outside so make sure if you use a pre-built module it gives you the access it needs. Again though, I have to ask, if you have a cold region to make the Peltier work, why not move the rest of the electronics there?

Brian.

 

[baileychic]

Yes, Peltier has a cold region. it has a large heatsink ad a cooling fan powered from Petier output.

 

[dick_freebird]

I think you will be disappointed in your plans for a
Peltier (TEC) module powered system. When I tried the
very same thing, with much higher deltaT, the module
was unable to power its own brushless fan - or rather,
gradually lost its ability to do so, as the heat flux that
was needed to get the fan going, soon overran the fan's
ability to extract that heat, raising the "cold" face and
losing the deltaT.

You had better find a better module and a better fan
(and better than the 6x6x1" finned aluminum heat sink
as well, might oughtta) and I still hold little hope for it.
Probably you want to aim for 50% - 100% higher than
fan nominal voltage, and not let the fan pull more than
maybe 20% of the module's advertised current while
pulling mucho airflow. Might want to think outside the
"box" (fan), maybe an impeller style would be more
efficient?

 

[betwixt]

Baileychic, can I ask what this device is measuring? I can see further problems with wireless link if it is inside a metal enclosure.

Obviously, I do not have access to your design requirements but from what you have described, there may be far better ways to achieve the end result that do not involve borderline technical issues.

Brian.

 

[baileychic]

@betwixt

It is for measuring furnace temperature.