Freezer Temperature Monitor

[ljh]

a newbie here - want to limit the run time of a standby generator utilizing freezer temperature. Have a Microchip TC622EAT sensor that works great on the breadboard. Can set the critical temperature with resistance. It flip/flops the 12 vdc from high output to a low output when critical trigger temperature is reached. However, I need to operate a 12 vdc relay that can handle 10 amps. The one I have draws 75 ma and the sensor max is 600 microamps. So, I screwed up the sensor output - only puts out 4 vdc which obviously won't trigger the relay coil. (I have a couple more sensors to play with)
The relay is NO, closed utilizing high out from sensor until temperature reaches trigger point and the 12 vdc flips.
The idea is if there is a prolonged power outage and I am out of town, the generator would run 24/7 and run the propane tank dry. If it only ran when the freezer needed it, it would conserve massive amounts of fuel.
Any and all suggestions appreciated. Thanks

 

[dknguyen]

So basically, you have a temperature sensor that needs to drive a relay right? Except the sensor output is too low? Try using a MOSFET transistor to use the weak sensor signal (connected to MOSFET gate) to amplify the signal for the relay coil terminals. NMOS are the more common, cheaper, and efficient kind of MOSFET, so use those. The only thing though is to keep the MOSFET the closest device to ground (it's a technical thing with the way the voltages work to switch them, and makes it so only a lower voltage is required to switch them on and off).

Here are the connections.

Sensor Output -> MOSFET GATE
+12V->+Relay Coil Input
-Relay Coil Input->NMOS Drain->Ground

Just make sure to pick a MOSFET that can handle your current, +12V across source-drain, and can be switched on and off at the gate with your sensor 12V output.

The MOSFET can be switched by your weak sensor output to control a larger current, which then feeds your relay coils, which in turn controls an even larger current. So in this case, it's pretty much like another relay (with some differences, like DC only).

EDIT: You said it goes low when the critical temperature is reached? This would change things a bit, like using a PMOS transistor maybe and then you would have to stick it nearest to +12V instead of ground (like with the NMOS). With PMOS, Source to +12V.

 

[philba]

that would certainly work but you could also use a cheap NPN transistor (2N2222, 2N3904, ...). It would be in the same position as the n channel mosfet with a resistor between the base and the sensor output. 2.7K would work.

if your relay has NO and NC contacts, you simply pick the ones that work the way you want. I think you want the relay energized when demanding power.

as to the TC622EAT. that thing will never drive a relay - it's max current is listed as 600 uA. That's 100 times less than your relay will draw which would likely toast the device. that is probably what happened to your first one. The device has both an active low and active high output so pick the one that makes sense.

 

[dknguyen]

Please post all replies on the forum rather than emailing.

I noticed you managed to catch the mention of a triac possibly replacing the relay and being driven directly by the sensor- forget about that. It's a bit more complicated, and a triac of the size you need needs too much gate current to be controlled by your sensor, so you would need an amplification transistor anyways. Just stick to the sensor driving a transistor, and the transistor driving the relay.

Find a transistor that can handle 75mA@12V and has a gate trigger voltage of 4V so that it can handle the relay coil voltage/current and can be trigged by your sensor.

 

[philba]

the TC622EAT has a Voh (V output high) of at least .8*Vdd. If all you are getting is 4V output with a Vdd of 12, it's damaged and may go all the way at anytime. I wouldn't use it for that circuit unless you don't care if the freezer thaws out.

I take back the 2N2222/2N3904 suggestion. Not enough gain based on the output current of the TC622EAT. better to use a darlington.

If you are using a mosfet, you don't need a sensitive gate one since the output high should be at least 9.6 (.8*12).

I didn't see the triac thing but agree with dk, fuggetaboutit. the relay will be ultra reliable and provides a measure of isolation.

 

[dknguyen]

I'm curious though, how often does your power fail and how valuable is the stuff in your freezer? Because in Florida, it would seem that power is very reliable and having a gas generator be on standby waiting for the power go out seems very wasteful- the relative reduction in power consumption of a gas generator on standby is not nearly as good as when a microcontroller can go on standby...

Philba also brought up a good point about the triac- if you want isolation between the electronics and the generator AC line, you will need to have more circuitry to isolate the triac which will make it more complicated. The relay inherently provides this isolation. Plus relays are better at conducting than triacs anyways (triacs are better at switching really fast). With relays, switching wears them down, while conducting does very little damage. With triacs, conducting wears them down, and switching does very little damage- the relay is a better choice since I don't think you will have power outages that recover every few seconds (or milliseconds).

 

[ljh]

Update

Thanks, Maybe I can answer some questions - last year the hurricanes created power outages lasting up to ten days or more. This is an automatic standby generator that operates automatically when there is a power outage through a transfer switch. If I had a refrigerator/freezer and separate freezer almost full, the loss could be substantial.
Let's assume I leave due to a hurricane for approximately a week.
In the case of a prolonged outage, the generator would run 24/7 - it uses a fair amount of propane and could exhaust the supply in my absence. That would allow all contents to spoil.
I should have said "dummie" rather than newbie because I don't know how to do the necessary circuit utilizing the components mentioned.
I went surfing and found the attached - would this work using the output from the sensor feeding the "open Collector I/O pin" and the high current load being the relay that draws 75 ma?
I don't know what the actual current is from the sensor (not having a micro ammeter) but if what I read that the transistor can amplify the current up to 200 X, seems it would work.
Thanks for your help so far - think we are getting closer.

 

[ljh]

More update - I also have a 555 timer chip and socket - could this be utilized or am I better off with the Darlington? If so, what number - 2N6426, BC618??
Thanks again

 

[philba]

here's 2 circuits I worked up, one using a mosfet, one using a darlington. I've not tested either but they should be ok.

the base resistor for the darlington should be enough to limit the current to 500 uA. I'd start with 20K or so. I'm not sure what you do with the relay output, though. the pot is to adjust your temperature setpoint. You will probably have to do this out of circuit. see the datasheet for details.

finally, how are you expecting to power this circuit? battery?

edit: answering your questions. the 2N6426 or BC618 would work ok. I don't know what you'd use the 555 for in this case.

 

[ljh]

Am using the generator battery as a 12 vdc source - it is constantly being charged so should work out ok.
Just picked up the components and will be playing with the circuits shortly.
Thanks a lot for the drawings - am going to try both circuits - will let you know how I do.
Thanks again - really appreciate your input

 

[philba]

note that the darlington will have a several volt drop so your relay will see something on the order of 10V. This is probably ok but check the relay datasheet for it's "must operate" or "pull-in" voltage.

 

[ljh]

With the MOSFET circuit, I want to verify I am reading the schematic correctly for the IRF520. S goes to GND, G goes to sensor output and D goes to one side of relay coil.

 

[ljh]

With the Darlington circuit, I connected 3(E) to GND, 1 (B) to the sensor output with 20K resistance in series and 2 (C) to one side of the relay coil but I am not getting any voltage out of Pin #2. Not sure if I did something wrong or if the 20K is too much resistance taking the sensor output too low?

 

[philba]

first, please make sure that the pin numbers for the darlington you are using are the same as the schematic, if not, make the changes. [I knew I should have found a darlington symbol that didn't have pin numbers.] then, once you have it wired as shown (ignoring the pin numbers), between C and E, you should see close to the full voltage when the device is off (sensor output low) and around 2ish V when the sensour output is high. Assuming that C is connected to +12V through the relay coil. Check the base voltage, also. it should be around 9-10V when the sensor output is high. For testing purposes, you could live with the damaged sensor output of 4V, by the way.

what darlington did you get? by the way?

 

[ljh]

I was able to get the TIP120 Darlington but could only get a IRF510 MOSFET rather than the IRF520 you showed.
For the Darlington, I got 1=B,2=C and 3=E off the data sheet.
Check the prior message about the MOSFET connections.
Thanks again

 

[ljh]

Maybe there is a polarity problem - I am running +12vdc to VDD on the sensor with GND being -12vdc. The sensor out is +12vdc. One side of the relay coil is +12vdc and the other is GND.
You mention that C should be connected to +12vdc through the relay coil - I have +12 vdc coming from the sensor. Maybe I should draw what I have and scan to a PDF?
Incidentally, the TC622 sensor that I thought was damaged, now appears to be working fine - +12 vdc when warm through High out and +12 vdc when cold through Low Out.
I am connecting +12 vdc to VDD on the Sensor and then going through 91K Ohms to tset - should give me about -4 C trigger point between High and Low output.
Am getting +12vdc out from the sensor - High out.
I can insert either the Darlington or the MOSFET into the mix - with the MOSFET, I get +12vdc to the relay but not enough current to activate it. With the Darlington, I am not getting any voltage measured between C and ground.
Is this hopeless? Maybe I should just order some Salmon from Pike Place Mkt and forget about this. But then again, that might be what is in the freezer!!

 

[dknguyen]

IRF510 is fine. The only difference between IRF520 is that the 510 handles 5A between source-drain and 520 can handle 10A. Your relay coil current is nowhere near that much, only 75mA, so it's all good.

You say you can't get enough current? Did you make the proper connections?

Sensor output goes to ground (I am assuming the sensor output is >4V below critical temperature and <2V above critical temperature, is this correct?)

+12V ->+ relay coil
-relay coil->MOFSET drain
MOSFET source->GND

Can you connect an ammeter in series with the relay coil (or anything else along that line) to measure the current that is flowing through the MOSFET source-drain and relay coils when you apply 12V to the gate of the MOSFET? What is the current being measured?

Just something I am wondering...don't fridges "shut off" when it's cool enough and only turn on when it gets too warm? Like when your fridge humms after you have left the door open for too long and then stops humming after the door has been closed for long enough...

 

[Hero999]

I wouldn't waste such a large MOSFET on powering such a small relay, I'd use something smaller like a 2N7000.

Don't forget to put a diode in reverse parallel with the relay and as 10A is quite a high current I'd also strongly recommend adding a reisitor the same value as the coil in series with the diode.

 

[philba]

It's best to debug one circuit at a time. darlington - make sure you have the parts wired up the way the schematic shows. Measure with the the common lead of your dmm connected to ground. Then, with the sensor output not connected and r3 connected to gnd, starting at the top, measure the following:

• +12 rail - should be, well, +12.
• top of relay coil. should be +12
• bottom of relay coil. should be +12 or slighly less
• top (collector) of tip120. should be +12 or slightly less
• bottom (emitter) of tip120. should be 0

Now, connect r3 to +12V. you should see the relay activate. the above reading will be the same except bottom of realy and collector should read around 2V.

If that works, lets look at the sensor. read the output voltage when its triggered - it should be around 10V (> .8*12). I've got friends here so I'll be offline for a while. good luck.

 

[Tarsil]

a newbie here - want to limit the run time of a standby generator utilizing freezer temperature. Have a Microchip TC622EAT sensor that works great on the breadboard. Can set the critical temperature with resistance. It flip/flops the 12 vdc from high output to a low output when critical trigger temperature is reached. However, I need to operate a 12 vdc relay that can handle 10 amps. The one I have draws 75 ma and the sensor max is 600 microamps. So, I screwed up the sensor output - only puts out 4 vdc which obviously won't trigger the relay coil. (I have a couple more sensors to play with)
The relay is NO, closed utilizing high out from sensor until temperature reaches trigger point and the 12 vdc flips.
The idea is if there is a prolonged power outage and I am out of town, the generator would run 24/7 and run the propane tank dry. If it only ran when the freezer needed it, it would conserve massive amounts of fuel.
Any and all suggestions appreciated. Thanks

U could consider a simple BC547C with a 8.2K resistor driving a Triac> The TIC263M is like 25A/600V and it costs less then 5$. U can use a triac rated at 16A/400V too (this one requires a biger heatsink but u can buy one for 2$). The HS for TIC263M doesn't have to be v large (Like one from an old PC supply)