I once saw a unit with a single bimetallic probe, that merely held a "relay" closed. As I recall, the relay had an actuator button on it, and the user put the probe in the fluid and pressed the actuator on the relay. If the fluid dropped, or the probe lifted, or anything went wrong, the N/C contact closed, which was monitored. In this instance an indicator light and a "piezzo" buzzer sounded. The probe had a cupric alloy on one face, and zinc or aluminum on the other face, with an insulating material between, which the metallic faces were bonded to. I believe the probe was a few inches long.

Of course, this is a galvanic circuit, and the electrochemical mechanism is known as a "cell". It would take 2 or more connected in series to obtain what is known as a "battery"...... wouldn't it!

 

I keep thinking this forum is going to send me notices when I get a new reply, but it doesn't seem to.

@MarkF and tranzz: Thanks for the reply. I think the sensor I'm going with is going to work well. It is just a stainless rod that sets up a galvanic cell with the aluminum pot, similar to what tranzz was talking about. I can't think of a cheaper, less breakable solution than a stainless rod, so I'm going with that until it proves not to work for some reason. I'm working on a bracket that will hang on the side of the pot, and when I have that I'll FINALLY be ready to put together a circuit, which I will need much help with.

 

Maybe a silicone bracket? Easy to clean, takes the heat, people are familiar with it in the kitchen. Or maybe the plastic they make cutting boards out of.

 

Maybe a silicone bracket? Easy to clean, takes the heat, people are familiar with it in the kitchen. Or maybe the plastic they make cutting boards out of.

I could be wrong, but silicone is usually a more rubbery material like rubber oven mitts. I'm also thinking a simple metal clip that would slip onto the side of the pot. Since each pot as a different radius I need something narrow that will slip over the edge like a clip so it doesn't matter if it is a large or small pot. If you've ever seen a candy thermometer that's the kind of thing. Even if the whole thing is metal, I can connect one wire to the clip (aluminum anode) and then make a separate metal piece to hold the stainless (cathode). The two pieces could be separated with some non conductive material and bolted or bonded together. At least for a prototype, I'm thinking to make something like that that I can bend myself into a crude shape that might work.

 

For a prototype you could use a spring clothes peg.

 

Hi everybody! Well it's been 3 months but I'm back. I hope I haven't worn out my welcome with these long absences. The holidays can be a busy time for me, plus you know how life gets in the way of the fun stuff like this!

I managed to put together a prototype bracket for holding the stainless rod. Maybe I'll post a pic. Anyway it's only a temporary thing. I took a piece of stainless sheet about the size of a tongue depressor. I bent it into a "C" shape and drilled holes equal to the rod diameter about 1/2 inch from each end. So by gently squeezing the open end of the "C" together, I can insert the rod through the two holes. Tension in the stainless "bracket" holds the rod in place. It's kind of a shallow "C" so that it has to be squeezed together. I mounted to this to another piece of stainless using a nylon bolt and some electrical tape and hung it over the lip of the pot. I did find that the tape tended to get soft and wet, seemingly losing some of its ability to insulate electrically.

In general, I can say that mv readings go up drastically when the water is heated. Typical operating mv ranges are typically 1000 to 1200 mv, and that is with only a small portion of the rod submerged. I think I'm getting a better reading by using a bracket instead of trying to hold the rod in my fingers up against a multi meter. I can still foresee readings below 400 mv, but only when the liquid is cold. With a short amount of heating, I believe all values will be well over 400 mv.

Here's the first thing that I'd like to understand better. KeepIt, if I understand correctly, is a proponent of the LT6700, which requires a 400 mv reference. On the other hand, some of you other guys say using a 200 or 100 or 50 mv reference is just as easy. Am I seeing a difference of opinion, or do I have that wrong? If I could save some cost by using only one component like the LT6700, that might be more important than being able to go down to 50 mv as with very cold water. It seems like that LT6700 is a pretty good thing and might be worth living with 400 mv.

The other observation, when boiling down chicken broth, is that the frothy bubbles do not impact the reading as much as I was expecting. I'd say as long as there were bubbles on the probe, it gave a pretty good reading. As soon as the liquid dropped to maybe 1/16" below the probe, I started seeing drop outs on the reading, and then back up with a splash of bubbles. This went on for one or two minutes before the reading remained at 0 mv for any length of time.

So I guess in starting to look at the circuit itself I'm wondering what the verdict is on the heart of the matter, the comparator.

 

Here's the first thing that I'd like to understand better. KeepIt, if I understand correctly, is a proponent of the LT6700, which requires a 400 mv reference. On the other hand, some of you other guys say using a 200 or 100 or 50 mv reference is just as easy. Am I seeing a difference of opinion, or do I have that wrong? If I could save some cost by using only one component like the LT6700, that might be more important than being able to go down to 50 mv as with very cold water. It seems like that LT6700 is a pretty good thing and mig

There's some competing issues like micropower, low voltage operation and possibly input impeadance. To compare against something, you need a reference. Hysteresis is essentially a deadband around the setpoint. An older reference/OP amp, I'm familiar with is the LM10 which has a 200 mV reference.

Internally, the reference in the LT6700 is 400 mV. So, if you wanted 40 mV to switch at, you just multiply the input by 10.

What was your polarity observation? Depending on the version of the LT6700, I think you can also make it <400 mV or > 400 mV.

Let's also make the assumption that no one has an induction stove.

Define what you can tolerate as a power source.

Welcome back!

 

There's some competing issues like micropower, low voltage operation and possibly input impeadance. To compare against something, you need a reference. Hysteresis is essentially a deadband around the setpoint. An older reference/OP amp, I'm familiar with is the LM10 which has a 200 mV reference.

Internally, the reference in the LT6700 is 400 mV. So, if you wanted 40 mV to switch at, you just multiply the input by 10.

What was your polarity observation? Depending on the version of the LT6700, I think you can also make it <400 mV or > 400 mV.

Let's also make the assumption that no one has an induction stove.

Define what you can tolerate as a power source.

Welcome back!

Thank you, Sir! I learned a bit about impedance tonight, Z. Seems like it won't be an issue for me as I'm not looking to be that exact, plus the LT6700 has it pretty well defined.

It is safe to say that nearly all users of this probe will be using an aluminum pot. If I'm thinking of this right, I believe we have a galvanic cell in which the aluminum is the anode, providing the small current to the stainless probe. I think the polarity would always be the same, but if it is just a matter of buying a different comparator of the same size, complexity and cost, then are there any downsides to have it measure <>?

I think I see what you are saying about switching at 40 mv. I could multiply the 1200 mv signal times 10 so that when it dips down to 400 mv it is, in reality, only 40 mv and nearly a dry probe. Hmm, something to think about. (Does it take more battery power to amplify a signal like that rather than just leave it alone?) Thataway you could for sure put the probe in cold broth, turn on the heat and forget about it instead of possibly having to wait for it to heat up first, and then drop in the probe.

Power source? I wanted the whole shebang to fit into something a bit smaller than a cigarette box, like a remote digital oven thermometer. I figured the majority of the battery would be for sounding the alarm, but how much juice will the circuit itself need when operating?

 

I figured the majority of the battery would be for sounding the alarm, but how much juice will the circuit itself need when operating?

Correct, sounding the alarm will use more power than anything else. Do you have room for a standard 9V battery? I think that's a good choice here. There are large button cells that may be good options too. I'd want to look for 6V, possibly two 3V cells in series. Garage door remotes use small 6V batteries that might be a good choice. Getting down to 3V or less starts making the circuitry more difficult.

 

Correct, sounding the alarm will use more power than anything else. Do you have room for a standard 9V battery? I think that's a good choice here. There are large button cells that may be good options too. I'd want to look for 6V, possibly two 3V cells in series. Garage door remotes use small 6V batteries that might be a good choice. Getting down to 3V or less starts making the circuitry more difficult.

I see no need to try and make this something James Bond would wear on the tip of his pen. Too small is something that can find its way into the trash along with the carrot and eggplant skins. My initial thought is that 9V wouldn't be too large, as I'm learning that the circuit will likely be pretty small. I think if a little larger battery gets you longer battery life, then that is more important.

 

Power requirements do go up. I just looked at this MAX4194-MAX4197 https://datasheets.maximintegrated.com/en/ds/MAX4194-MAX4197.pdf Instrumentation amp. It does come in a variable gain version or a x10 gain version.

The datasheet says 93 uA at 2.7 V with a maximum of 8V, but it doesn't include any supporting stuff.

Another thing to watch out for is reversing the battery. Things get a lot more difficult when the supply voltage gets low.

To jump ahead, take a look at the HT12 series modules at http://www.holtek.com.tw

I really don't want to jump ahead too far.

 

mounted to this to another piece of stainless using a nylon bolt and some electrical tape and hung it over the lip of the pot. I did find that the tape tended to get soft and wet, seemingly losing some of its ability to insulate electrically.

You could use a fiber shoulder washer http://www.keyelco.com/category.cfm/Multi-Purpose-Hardware/Washers-Flat-Shoulder-Nylon-Fibre/id/448 to do the insulating. Fibre probably has a higher temperature rating than Nylon. Nylon is available at the local hardware store.

These http://www.digikey.com/product-detail/en/aavid-thermalloy/4880SG/4880SG-ND/1625671 or TO-220 are used to electrically insulate, but thermally couple a to-220 transistor which is basically a piece of flat metal with a hole. The step washer is still the major component,
So, a screw, ring terminal under screw, washer, insulating step collar, a piece of metal#1, a flat insulating washer, metal#2, lock-washer, nut would give you electrical access to metal #2 from the ring terminal. Other variants would work.

I'm surprised you didn't use a clothes pin?

Electrical tape should be used like a crayon. The colors are used to "mark" the intended wire color. e.g. It is used to mark a white wire as black when it's actually switching the back wire when a ceiling fixture has power and the switch doesn't. The white wire on the switch needs to be coded black. New code wants the switch to have a neutral. There's a good reason to put a neutral in the ceiling and the wall fixture.

 

there are lots of eBay water level switches about that might do the job very cheap. For the probe you could consider a different approach. If you had a stainless steel rod inside a stainless steel pipe with some for of electrical insulation between. The rod would be the source ac voltage and the pipe the return. If the pipe was inserted nearly to the bottom of the pot and the rod set at the low level required inside the pipe. As long as there are vent holes at the top of the pipe, the fluid level should equalise inside the pipe with the level in the pot. This was you would avoid any splatter from boiling liquid.

Just an idea.

 

You have to remember that we are talking about chicken broth which has meat and vegetables in it. So, it's not just a liquid.

His sensor depends on different materials with the broth as the electrolyte essentially making a battery. It's not the stand way of making a water sensor, where an AC voltage is used to prevent plating of the electrodes. It's supposed to help prevent the heating of an empty pot and capable of being used by people with a low and high IQ.

 

You have to remember that we are talking about chicken broth which has meat and vegetables in it. So, it's not just a liquid.

His sensor depends on different materials with the broth as the electrolyte essentially making a battery. It's not the stand way of making a water sensor, where an AC voltage is used to prevent plating of the electrodes. It's supposed to help prevent the heating of an empty pot and capable of being used by people with a low and high IQ.

Maybe I should call it the KISS level sensor. lol.

Actually what generally happens with something like gravy is that you slow boil raw chicken scraps and veggies like tomato and celery and whatnot. After a day of gentle low boiling, the meat, bones and veggies have given up their collagen and flavors. Let this chill overnight and skim off the solidified fat on top. Strain everything with a fine sieve and dump it back into a large pot. So now that we have extracted the flavor, the job is to reduce the volume and concentrate. So really the stuff we are boiling is pretty clean. The final reduction can be used as a gravy, sauce, soup stock, etc. That's the basic idea of what I want to do. Of course we've discussed the other requirements as to power source, durability, etc.

 

I actually have another idea, but I need to know how many pots will be in operation at any one time.

These http://becintegrated.com/linear-dxsr-1504?site=google_base come in 4 and 8 channel versions. The sell sheet is here: http://www.nortekcontrol.com/pdf/literature/DXSR-1504_DXSR-1508_Lit.pdf

This http://www.linear-access-controls.com/su4chredx.html link has some manual links.

I hacked into an earlier version of these wrist/pendant xmitters http://www.nortekcontrol.com/product_detail.php?productId=1510 and made an expensive wireless doorbell, but I have other uses for the other channels. i know the version I had, the batteries were not replaceable. but were CR2032 batteries with tabs. Another generation of the transmitters added a parallel battery for longer life. The transmitters did have a LED when the button was pressed. I essentially removed the battery and put leads on the LED and out it into a wireless doorbell case.

The DXS series of transmitters are used in Linear's (now Nortek) supervised transmitters for their PERS (Personal Emergency Systems) that I installed at home. They use the same series of transmitters for security as well. The receivers go through a learning mode/process when you pair them to the receiver. With one test activation per month, the batteries last for multiple years.

The transmitters are capable to do battery monitoring and transmitter out of range reporting to the receiver.

Getting relay outputs that I needed takes an extra step. I added an alarm silence button for the supervisory tasks.
Note that the receiver can accept multiple kinds of outputs such as Momentary, toggle and Latch

Not putting a whole lot of energy into looking at the system, you might be able to use the toggle mode with an auxiliary button at the transmitter and tie all of the sensors to a single buzzer.

You might be able to make use of say a button on the sensor to toggle the alarm for that sensor with a switch that forces it off or use a pendant or wall mounted transmitter to reset the receiver.

The wireless side of things becomes the expensive part.

 

I actually have another idea, but I need to know how many pots will be in operation at any one time.

The wireless side of things becomes the expensive part.

If somebody has 3 pots going at one time, then I can sell them 3! Don't kill our action, man.

Wireless does sound expensive, though. I think for most people one is enough, though I could realistically imagine two being needed. Maybe the KISS sensor version 2.0 can be wireless, otherwise KISS version 1.0 won't be so SS.

 

So how do I go about putting together a circuit? Is there a resource/idiot's guide where I can find the building blocks I'll need once I know what I have to buy? I imagine I'll have something I can solder together and take apart/rearrange as necessary. Once it is working properly, then I can worry about making one that looks nice.

For now all I know is I'll need probably a 9v battery, an on/off switch, a buzzer of some kind, a comparator (LT6700?).

 

You'll want to get yourself some protoboards. Here's an example. Note that there are several sizes such as 4x6, 5x7, 6x8 and so on. They'll be cheapest if you can wait for them to come from Hong Kong and a bit more if you want them faster from a U.S. supplier. As far as I know the green boards are all the same quality, so just shop on what best fits your needs.

Do you have a soldering tool? I have one of these and like it a lot. I was drawn to it in part because the replacement tips are cheap. You can spend a lot more but for my low usage I'm plenty happy with the cheap unit. I lived with even cheaper options you can pick up at the hardware store, but I don't recommend those. Get some solder to go with it. Like this. (The soldering station and solder are both available elsewhere, and you might not need that much solder but that's the kind you want, 60/40 tin:lead.)

Hmmm.... This is for kitchen use and should probably be lead free. I've never used lead-free solder in my projects so you should get someone else's advice on this.

For hookup wire, you'll want 22 gauge solid copper, insulated wire. You can buy it, but you can find it around. Good wire strippers are a must. I have one a lot like this but got one similar to this for my daughter, who says it's amazing. Don't mess with the cheapest ones like this. Been there.

 

The soldering station is something else in terms of price. I agree with the strippers. T-strippers also work.

If you have never soldered before, tin/lead might be good for a prototype. 60/40 was the standard solder until Surface Mount showed up. Then 63/37 works much better for SMT parts. Why? Because it melts and solidifies at the same point.

You might end up using Sn96. Sn96 is relatively hard to work with. The tin/bismith alloys are nice low-melting point alloys. I would not use them in a final product/

A solderless breadboard such as https://www.amazon.com/BB400-Solderless-Plug-BreadBoard-tie-points/dp/B0040Z1ERO MIGHT be worthwhile.

In any event, surface mount will continue being the norm. For breadboarding, www.proto-advantage.com does well in making SMT to DIP adapters and will solder a digikey part # on their board for a small fee.