Control of Central heating from PC - Thermostat Inputs - Relay outputs

is

A few ways you can do this Picot do printer port data loggers than can be read directly into Excel spread sheets --- I've tried this and it works. Howerver a better way of measuring temperature is using Dallals semiconductor DS18S20 temperature sensor, a selection of software for reading these is available off the net but you may find a combinatio of Autom8it and Quasar electroics 3145 temperature logger the easiest and cheapest to get working. Links

Check these out for some ideas...

M.

Personally, I would take the route of a custom application utilising a

1-wire bus for both control and temperature sensing. I figure if you are going to want precise control over the temperature at all, you may as well be able to control the whole system down to room level. For me anyway, the extra effort in rolling a custom application would be worth it. There are loads of 1-wire examples on the web to get you started.

regards Alastair

Another link:

I am using one of Phil's modules to read 3 DS18S20's (upstairs, downstairs, outdoors) and two analogue inputs (upstairs and downstairs potentiometers as room stats) to control a two zone heating system. I have been thinking of attaching DS18S20's to the flow and return, although the Dallas datasheet says they may not work well in parasitic mode at high temperatures due to increased current leakage inside the devices, and Phil's module does drive them in parasitic mode (multiple devices on a single cable pair).

Phil's module provides the data over an RS232 serial link, which I access with a purpose designed program (which I'm still working on). In Windows, you would probably need to write a program to interface between the serial port data and Excel.

--
Andrew Gabriel

Depends how much the error is, if its a lot it is back to an adc and thermistor(s) which would need calibrated, it would be good to collect boiler exhaust gas temperature also by sticking a thermo couple into it.

I rather imagine that if it reads, you'll get the right answer. I don't recall the datasheet being specific on the failure mode.

What might happen is that it runs out of power partway through transmitting back on the bus (as it's powered from a small internal capacitor for this). Another possibility might be quiescent leakage current through several 'hot' devices starts interfering with the reading of all devices in the bus, reducing the number which can be driven in parasitic mode on a bus. All just guessing though...

Also, bare in mind flue gasses can be quite corrosive if you try measuring their temperature. I have a condensing boiler and the condensate is similar acidity to vinegar (not measured it myself, but that was mentioned somewhere in the blurb). Condensate pipework must not contain any metal (except stainless steel).

--
Andrew Gabriel

Some other tips which are based on my experience of using DS18S20's...

Check the temperature you got back is reasonable. Glitches during reading (probably interferance on the bus) often seem to return readings of -88.8C or 85.0C, both of which I just treat as a misread unless the previous reading was near that value. I also disregard the odd reading which would imply, for example, the room jumped in temperature by 20C in a few seconds for just one reading (I think such a reading only ever happened once to me).

The device can oscillate, alternately returning readings of, say,

20.0C then 20.5C when the temperature is on the cusp. I ignore changes of 0.5C for a period (a minute IIRC) to avoid the program really thinking the temperature is oscillating every few seconds when a new reading is taken. Another way would be to average the temperature based on a decaying average of the past readings, but that would be less responsive to genuine fast temperature changes.

(You can read temperature more accurately than 0.5C from the devices, but it requires a different method and the controller I use doesn't do this.)

--
Andrew Gabriel

You might overcome that weakness by comparing the decaying average over say

15+ minutes to a shorter decaying average, say 3 minutes. Any time the delta exceeds an acceptable range activate the air handler or whatever. The logic is simple enough -- a few lines of code. Heat anticipation logic could allow a slow-acting, hydronic or radiant system to activate before a descending temperature reaches the limit and deactivate before the rising temperature reaches the upper limit.

Regards, Robert

=============================>

Bass Home Electronics, Inc

2291 Pine View Circle Sarasota · Florida · 34231 877-722-8900 Sales & Tech Support 941-925-9747 Fax 941-232-0791 Wireless Nextel Private ID - 161*21755*1 =============================>

Note: Parasitic mode does not imply multiple devices on the line. Parasitic mode means that you are really using only two wires and not three (where the Vdd pin is tied to a high current voltage source permanantly). In parasitic mode, the Vdd pin of the sensor is tied to ground. Also in parasitic mode, the microcontroller must supply extra power to the DQ line during a temperature conversion as the charge stored in parasitic mode is too weak to accomplish this without it.

BTW, the 18B20 offers higher accuracy (

Why Control a Central heating from PC?

I thgink you are.

I think, that you and the other's replying from the UK are the same person. Without any design specs they are recommending hardware.. Amazing? Build it first, then design it...

-- Ray