millivolt thermostat interface

[dan king]

Hi Everyone,
I am working on a home project where I am building a wireless thermostat for my pellet stove. I have a 18F2445 connected to a 434Mhz transmitter that reads the status of a commercial thermostat. The results of the status check are then transmitted in a wireless packet to the receiver connected to another 18f2445.

My question is a little off topic at this point as I have everything working except the interface to my stove. The manual states the thermostat must be a millivolt type which leads to my question, what is a millivolt thermostat. Everything I have read states they are thermopyle arrangements used to power other components with the generated voltage but the commercial thermostat, which states it is millivolt compatible, uses a contact closure.

Does anyone have experience in this area to let me know what type of signals my PIC has to generate for proper operation? or is contact closure adequate for the interface.

I don't want to damage my stove controller so any help is appreciated.

Thanks,

Dan

[FvM]

Is it a passive thermostat (without a powersupply)? I assume, that it uses a low ohmic (thick wire) thermocouple. The current can be rather high then, e. g. 100 mA and more. A bit difficult to produce it by electronic means.

[dan king]

The commercial thermostat that is connected to the transmit setup is powered by 2 AA batteries. I had to design a charge pump to get 5 volts for the PIC. The commercial unit has a relay connected to the heat terminals so I'm assuming that I might get away with contact closure on the receive side and that maybe the contacts are just pulling a port high or low on the stove controller. I just wish the stove manufacturer provided more info but I guess they want you to buy their unit for a lot of $$$.

Thanks,

Dan

[mickent]

If your stove turns on and off with a contact closure then the type of thermostat is irrelevant. You just have to use a relay at the stove that follows your remote thermostat. Maybe two tones; high for on, low for off. Google "millivolt thermostat" for more info on how they work.
_________________
Mick

[dan king]

Hi,

I already googled the millivolt tstats but don't understand why a modern control would specify this since the external tstat is the part that needs the temperature measurement and not the stove control. The only thing the stove control needs is a command to provide heat or not. That's why I'm not sure what's going on.

I don't know if contact closure will work or not. Do I need to provide a millivolt signal with the PIC or just close a contact. I'm ASSUMING that contact closure will work but I'm not an HVAC person. My design is complete except for this last detail. It's even installed already with a small reed relay but I'm afraid to hook it up for fear of damaging the control.

I've posted this same question on a pellet stove site but I'm hoping to find someone on this board with technical experience to shed some light on the subject.

Thanks

[FvM]

Hello,

if the controller is battery powered, the sensor input is probably for a standard thermocouple (NiCr/Ni or similar). The latter has a signal of about 4 mV/100°C and is allowed to have some internal resistance. It can be simulated easily by a PIC, e. g. through a PWM DAC, scaling down the signal with a voltage divider to the mV range.

I share however the view, that it would be much easier to control the HV circuit by a relay directly.

Regards,
Frank

[jaymiller]

With respect to the 'why millivolt stat' question. It could be a safety aspect .A millivolt system is selfpowered. The thermocouple generates the voltage that feeds the stat that that controls either the gas valve(in a gas fireplace) or a servo valve(in a wood burner).During a power failure your stove or fireplace will still safely operate.

As to replacing the stat, as long as your replacement will handle both the voltage and current of the system you will be okay.Just be sure to read the 'specs' of the millivolt equipment in your system.They are NOT all the same....

Jay

[dan king]

Thanks for all of the info but I think I'm not expaining what I'm trying to do correctly.

The stove I'm working on does not have it's own thermostat, it will run manually open loop, unless you connect an external thermostat to two terminals on the stove. The stove doesn't measure temperature itself. The stove manual states that the external thermostat needs to be millivolt compatible - why I don't know since the stove doesn't measure temperature.

I bought a honeywell thermostat that states it is millivolt compatible but when I used my DMM to measure resistance, the terminals are connected together with an internal relay during calls for heat.

I understand the theory behind the older systems that actually measure temperature and control a low voltage valve without external power. My stove will not operate when power has failed, everything just burns out and then stops. I don't think the stove "wants" a thermocouple connected to it since there aren't any temperature controls on the stove itself, it is looking for an external thermostat to tell it what to do.

I have simulated this system with a couple PICs and a 434Mhz wireless connection and a relay to connect to the stove external thermostat terminals that shunts those terminals.

Why would the stove manufacturer specify thermostat compatible thermostats when it appears to be looking for contact closure or am I wrong and this isn't going to work?

I will attempt what Frank suggested if I need to provide millivolts to the stove but I don't think this is the case. If someone can shed some light on this I would appreciate it.

Thanks in advance,

Dan

[Ttelmah]

The problem here is nomenclature.

Historically, millivolt, has it's true 'technical' meaning, of 1/1000th volt.

Then in furnaces (in the UK, this implies really high temperature operation, rather than 'boilers', for water heating systems), you have mV thermocouple systems used to actually measure the temperature. These use thermocouples.

Then in home boilers, you have the 'millivolt' system, where the thermostat input, only carries a low voltage, without either mains, or higher voltage DC connections.

All the term implies in this context, is that the voltage the thermostat has to control, is a low voltage signal, with low current flowing as well. As such you can switch this with things like reed relays, which would not handle the normal higher power for other systems. However the system can equally well be switched by many (most) of the standard high voltage relays as well, provided they are self powered. It is used as a term, to distinguish from 24v, or mains switched systems.

Best Wishes

[dan king]

Hi,

Thanks for the additional info. I finally got a response from the manufacturer and it caused more confusion than it helped when I asked if I can use contact closure to call for heat:

[SherpaDoug]

Until I read your quote from the manufacturer I though I had your answer....

Generally in relays millivolt contacts are different from power contacts in that they are designed for small voltages. In a 120V relay system the contacts can develop a thin oxide layer without harm. When the power contacts close the 120V burns through the oxide layer completing the connection.

In a millivolt circuit even a slight oxide layer may keep the circuit open, so the chemistry and geometry of the contacts is different to prevent it. Often power contacts use flat nickel or silver for high power handling. Millivolt contacts tend to use narrow ridges of gold or mercury wetting to avoid any oxide layer.

If you were switching TTL signals you would want millivolt contacts instead of power contacts.

Unfortunately this does not seem to be what your manufacturer is talking about. In 20+ years of instrumentation design I have never heard of 250, 500, or 750 millivolt contacts.
_________________
The search for better is endless. Instead simply find very good and get the job done.

[crystal_lattice]

Interesting yet very confusing!!!

I've worked on various furnaces and commercial ovens and most of them employ a thermocouple, usually a type K, but i have come across type J and type R.

Either way, they produce a small voltage (mV range) with a change in temperature.

As a test you could use a external thermometer to check the temp when unit is cool, then also check the voltage on the two wires. Then heat the unit up to a safe working temp for the external thermometer and repeat the tests.

If your readings differ then have a look on the internet for a thermocouple chart. I have found some in the past but will have to look for them. They have the voltages listed for the whole temperature range (in about 0.5 Deg increments)

You might have problems finding the correct type thermocouple spec, one way of confirming the type is to check the wire colours going to the termination on the oven (check that they are not covered by high temp insulation/heatshrink). Again use a chart found from the net. Problem is various countries use different colour schemes.

I personally think the unit has built-in thermocouple and the control unit can then be bought as an add-on, this ensures the thermocouple is installed to factory spec /location and also makes it easier for the client to do an upgrade.

As a general rule i don't always believe what the manual or sales person says, the factory can sometimes not be trusted either as the person you speak to might not be as knowledgeable as you would hope.

[FvM]

Hello,

I'm also surprized to hear of a 250, 500 or 750 mV thermostat. They are never used in industrial measurement and control. I also have a HVAC electronics textbook from the US and it doesn't even mention them, although it discusses verbosely all types of thermostats.

I have learned now that they are thermopiles (basically a bundle of a few ten thermocoupes in an electrical series connection) and used to control valves and relays without a supply voltage. I know this technique with single thick wire thermocouples, as I mentioned above. I found an information from a catalog distributor, that has some basic electrical specifications:
http://www.pexsupply.com/product_dtl.asp?pID=1743&brand=Honeywell&cID=284
As you can calculate from the specification, the thermopile can supply a short circuit current of several hundred mA. If it's loaded with a low ohmic valve or relay, an equivalent circuit replacing the thermopile must produce the same amount of current, respectively have a few ohms output resistance. This is very unpleasant for a battery operated uP circuit, even if using a switching regulator to supply e. g. 800 mV low voltage supply.

Thus my idea, to supply an equivalent voltage doesn't work. But a reed relay should be able control the mV signal in a reliable way. As another option, a small low voltage MOSFET (it may have an Rds,on down to 100 mohm or below) could be used as a switch, eliminating any static current consumption.

Regards,
Frank

[crystal_lattice]

Couldn't find the PDF's i had on the thermocouples but found these links:

http://www.omega.com/techref/thermcolorcodes.html
http://www.peaksensors.co.uk/thermocouplecolourcodes.html
http://instrumentation-central.com/pages/thermocouple_reference_table.htm

[FvM]

As pointed out, the mv thermostats are not using regular thermocouples, none of the types listed in your link.