I developed a THT-PCB with multiple different types of loads like fans, waterpumps and peltier elements. The working voltage is 12 V for these loads. I control them with an ESP32 - so far so good, everything on the controlling side is working properly.
Now I'd like to add a "checking instance" to check if a load is really running - that's where my problem is.
The 4 pin fans have a tacho RPM sensor pin, so this shouldn't be a big deal.
For the pumps I thought of using either a Reed switch or an Hall effect sensor to check if the pump is running. Since I did not built and therefore not tested it yet, what do you think about this?
My biggest problem are the peltier elements. I thought of putting a voltage divider parallel to the elements and check if there's a voltage drop, but I'm not sure if this would work as expected.
Do you have some good ideas how to manage these checks?
Actually there are only two requirements:
Parts should be THT and therefore self-mountable on my PCB (so no external breakout-boards)
Oh man, why didn't I saw this so obvious solution
Good idea, this is already part of the PCB, so it's just a second usage for this parameter. Thank you!
Water pumps can get a little complex depending on what and how much you want to know. Voltage to the pump? Current to the pump? Does pump have actual flow? Depending on the pump type it can have voltage and show current but actually pumping anything. Depending on what the pump is moving there are flow sensors. Not concerned with flow rate but is there flow or not? A simple flow switch will give you a yes or no. All depends on what works for you.
Use a current shunt in series with the peltier devices and monitor the current instead of voltage. Pieltiers are current mode devices.
Used to joke that they had a negative temp coefficient so the the harder you drove 'em the colder they'd get & if you reduced the current to zero they'd burst into flame.
Hi,
Can I suggest two temperature sensors, one on each side of the Peltier and in contact with the surfaces.
The are in the specs of a peltier the maximum differential temperature you can have and the maximum highest operating temperature.
These ensure "efficient" and non destructive operation of the Peltier.
I think flow is enough, this shows at least that the pump is working. I could then add more details like flow rate to ensure it is running fast enough. Good idea!
@madmark2150: Do you mean something like an INA219? I already searched such ICs but I wasn't able to find anything suitable (= THT mount) for my application. Do you have a specific suggestion?
@TomGeorge: I'm already using temperature sensors, on on each side of the peltier (or rather: one on each side of the water tanks attached to the peltiers). I can't attach sensors directly to the peltiers, otherwise I would neither be able to cool the water nor to cool the peltier itself). But I like the hint about the maximum highest operating temperature - I will implement that. Thanks!
That's what I already do. The heat sink is a tank, too. So one water circuit is cooling the peltier (= warm water) and a second water circuit is transporting the cooled water.
So actually the most "problematic" part is measuring a current because I do not find a THT current sensor.
I also do not find an affordable G1/4 flow sensor, but I think this can be done another way.
You measure current with a shunt. What is your current range?
The big terminals are the current in/out, the small screws sense the voltage drop across the shunt. Measure THAT voltage.
Looking back at your first drawing with R19 & R20 in series shunting your Pelieier heater. With respect to ground R19 is going to drop about 3.33 Volts and R20 about 8.65 Volts so with 12 Volts applied your Pelitier_01_SIG voltage is about 8.65 volts. That what you want or did you want to reverse R19 and R20? On your U9 Relay you want a flayback diode across the relay coil. Cathode side to + relay coil terminal.
Your pump transistors have a maximum collector current of 1.5 amps so I will assume your pump max current is 1.0 Amp. With that low of a current I would just run with a simple current sensor module. A MAX471 3A Current Sensor Module can be had very inexpensive. There are plenty of similar modules to be had. While the MAX471 is no longer manufactured there are millions out there and they are a high side sensor. Using a shunt for so small a current I so not see as practical because with such a low DC current you will need to amplify the shunt output to get anything usable.
Again, all of this only applies if you want to take the current approach to know if a pump is running. If you know your expected flow rate I would just use a flow switch.
You are right, I changed R19 & R20. Now the voltage of PELTIER_01_SIG should be 3,3V. I also added a flyback diode across the relay coil (good hint, I forgot that).
Just to be sure I understand you correctly - this is what you meant, right:
I fully agree with this, therefore I thought of using Hall-effect sensors (or a Reed switch) to check if the pump is running. I have some of the MAX471 breakout boards at home, I will test this approach. I just saw those DIP chips are also (rarely) available as THT DIP, which would work perfectly for my PCB design. Do you have any other suggestions for such chips in DIP THT design?
Not so easy as I don't find any matching flow switch / sensor with G1/4". I don't want to add multiple adapters for different thread sizes as these increase the chance for leakages...
For the Peltier elements the maximum is about 2.5A.
Yes, that's what I had in mind for your divider and switching the two reaistoer.
Yes, one way to detect current is a magnetic reed switch. I have seen some crude home brew systems work. Just a matter of winding a coil sround the switch with one of your motor leads. Number of amp turns depends on the switch and wire gauge on the load. Take a common drinking straw and cut it to the length of the switch glass portion. Start wrapping turns till with pump running you get a good switch closure, just make sure you have enough turns and then some.
The heat sink is a tank, too. So one water circuit is cooling the peltier (= warm water) and a second water circuit is transporting the cooled water.
