RE: Simple flow switch for HA feedback

["Nigel Giddings" <nigel@xxxxxxxxxxxxxx>]

I would second the idea of a pipe stat. Easy to fit to different size
pipes, my boiler has 28mm connected to the flow and return, and will not
impede flow or cause any warranty issues. It's also a dry job to install
or replace...



The boiler should dump heat once the heating cycle is finished so the
pipe stat will reliably switch off, it can be set to a relatively high
temp (60oC plus) and will not be affected by its surroundings.



Nigel



-----Original Message-----
From: Tim Hawes [mailto:timsyahoo@xxxxxxx]
Sent: 29 June 2007 17:01
To: ukha_d@xxxxxxx
Subject: Re: [ukha_d] Simple flow switch for HA feedback



The alternative (and I don't know why I didn't twig this before, see
*) would be to use a domestic hot water cylinder thermostat clamped to
the boiler outlet pipe and pick up the dry contact from where you
normally connect the thing to the boiler programmer. They're usually
just simple switches with the added advantage that they're readily
adjustable for temperature (the dial on the front) and there's no
additional pressure drop for the circulation pump to deal with. And
they're only a tenner :-)

* I was planning to use one of these on the hot feed to the shower to
start the extract fan but hadn't got around to it yet. Maybe this
weekend if the weather is as bad as predicted :-)

If the temp switch isn't reliable enough (fiddling to get the right
temp settings etc.), consider a mains relay inserted into the supply
to the circ pump and pick up the comfort input from the relay's dry
contacts. This has the advantage of changing state as soon as the pump
switches off, whereas the temp switch would only change state once the
boiler pipe has cooled down sufficiently.

HTH,

Tim.

On 6/29/07, Paul Gordon <paul@xxxxxxx
<mailto:paul%40paulgordon.homeip.net>
> wrote:
> Yes, this was/is something I have also thought of, and may well do
anyway... Not sure it would be OK for real-time event processing though,
depending on the speed of response.. - i.e. How soon after the boiler
has fired up would the DS1820 signal this fact?? - due to the potential
range of temperatures that could be experienced in the environment,
you'd have to configure a fairly significant threshold before
determining that the heating is definitely on... Temps of 30C aren't
unheard of in this country, so anything below that couldn't be relied
upon as a definite. I would think that a threshold temp of at least 40C
would need to be set as the threshold. If the ambient temperature is
20C, then you'd have to wait until a 20C increase was measured. Granted
with the sensor clamped fairly tightly to the bare copper pipe, I don't
suppose this would take ages, but I still think there would be a
noticeable delay, which isn't good if you want to have comfort run a
response when the heating is switched on...
>
> Also, I'm not currently aware of a way to interface a 1-wire temp
sensor directly to comfort... I imagine that it might be possible to
modify an LSM01 board to switch based on a temp input level rather than
a light input level (such a use is even alluded to in the LSM01 docs),
but it seems the LSM01 isn't actually available any more (it's
disappeared from the CHC website), and finally, I've had a hard enough
time getting the LSM01 board to work in its intended configuration, so I
don't hold out much hope for my ability to modify it...
>
> I don't want any reliance on a PC for this function, the communication
must be directly between the heating system and Comfort.
>
> Cheers.
>
> Paul G.
>
>
>
> -----Original Message-----
> From: ukha_d@xxxxxxx <mailto:ukha_d%40yahoogroups.com>
[mailto:ukha_d@xxxxxxx
<mailto:ukha_d%40yahoogroups.com>
] On
Behalf Of Brian Smith
> Sent: 29 June 2007 16:19
> To: UKHA Group
> Subject: RE: [ukha_d] Simple flow switch for HA feedback
>
> If it's monitoring you're after would a temperature sensor not be
cheaper
> and easier retrofit. I was planning, but confess haven't got round to
yet,
> to put a 1-wire sensor on the output of the boiler to monitor when it
was
> actually firing.





[Non-text portions of this message have been removed]