I have two strap-on cylinder thermostats, as normally used for DHW cylinder
but in this case applied to a thermal store. I understand (right or wrong)
the Switching Differential to be the temperature difference between 'On' and
'Off'.

I've had a Danfoss ATC, with a quoted SD of 6-10K approx. However, for my
installation I measured the peak& trough at about 30degr. I though it must
be broke, so I replaced with a new Honeywell L641A with a stated SD of 10
approx. But it turns out this also gave me a peak-to-trough difference of
about 30degr. Seems to be mounted OK, flat to the cylinder (and yes I dug
out the insulation to the copper). I can understand these being a _bit_ out
but not by 300%. Can anyone shed any light on this?.- I'm baffled - I can't
have two broken stats?.

The boiler control circuit just has this 'stat wired in series - electrical
connection between 'C' and '1' for make-on-rise & break-on-fall. The stat
location is a bit over halfway up the cyl. The SD needs to such that the
boiler doesn't frequently switch on/off (undesirable for varoius reasons) or
the CH doesn't go cool for long periods. The supplied built-in immersion
'stat is too sensitive and switches the boiler too much, hence the strap-on
stats. Actual SD of 12-15C would make the boiler run every hour - about
right.

One way out of this would be an adjustable cyl stat - preferably one with
the control dial connected by a wire to a cylinder sensor (to make it
accessible). I haven't found one yet (and I'm dreading the price tag if I
do). Or, take pot luck with the Horstman cyl stat (SD of ~3) and see where I
end up there. Just seems like something simple has somehow become
difficult!.

I don't quite know where I'm going with this so any thoughts would be
welcome.


Egremont.

PS - I know there are better and more expensive/complicated solutions which
I'm trying to avoid if I can.

"Egremont" wrote in message news:4592d94a.0@entanet...
I have two strap-on cylinder thermostats, as normally used for DHW cylinder
but in this case applied to a thermal store. I understand (right or wrong)
the Switching Differential to be the temperature difference between 'On'
and
'Off'.

I've had a Danfoss ATC, with a quoted SD of 6-10K approx. However, for my
installation I measured the peak& trough at about 30degr. I though it must
be broke, so I replaced with a new Honeywell L641A with a stated SD of 10
approx. But it turns out this also gave me a peak-to-trough difference of
about 30degr. Seems to be mounted OK, flat to the cylinder (and yes I dug
out the insulation to the copper). I can understand these being a _bit_
out
but not by 300%. Can anyone shed any light on this?.- I'm baffled - I
can't
have two broken stats?.

The boiler control circuit just has this 'stat wired in series -
electrical
connection between 'C' and '1' for make-on-rise & break-on-fall. The stat
location is a bit over halfway up the cyl. The SD needs to such that the
boiler doesn't frequently switch on/off (undesirable for varoius reasons)
or
the CH doesn't go cool for long periods. The supplied built-in immersion
'stat is too sensitive and switches the boiler too much, hence the
strap-on
stats. Actual SD of 12-15C would make the boiler run every hour - about
right.

One way out of this would be an adjustable cyl stat - preferably one with
the control dial connected by a wire to a cylinder sensor (to make it
accessible). I haven't found one yet (and I'm dreading the price tag if I
do). Or, take pot luck with the Horstman cyl stat (SD of ~3) and see where
I
end up there. Just seems like something simple has somehow become
difficult!.

I don't quite know where I'm going with this so any thoughts would be
welcome.


Egremont.

PS - I know there are better and more expensive/complicated solutions
which
I'm trying to avoid if I can.

I am assuming that the stat is wired with just two wires live and
switch-wire (plus earth).
Inspect the stat carefully, is there provision for a neutral, that has never
been
connected? If so, connecting it will improve the responsivenes of the
thermostat
by switching in an accelerator (actually a high value resistor) that
pre-heats
the bi-metal strip causing it to switch off sooner than it would do
naturally
thus reducing the "hysteresis" of the system.
The neutral needs to come from the same CU circuit.

A far better solution would be a digital control unit that would just need
a thermistor to take the temperature of the cylinder.
--

Graham.
%Profound_observation%

Thanks for that. I checked the new Honeywell L641A and docs, but no 2nd
neutral connection. Apparently nothing for the Danfoss ATC either. So it
looks like I'm stuck with the 30C switching for these, which seems insane!.
Perhaps the slow thermal draw-off in comparison to, say, running a bath is a
problem?

I looked at Digital alternatives like the Danfoss remote stat, but haven't
yet seen anything that will allow me to set both the 'On' and 'Off' temps
(or the differential). Without this I'm back to square one.

Egremont

Graham wrote:

I am assuming that the stat is wired with just two wires live and
switch-wire (plus earth).
Inspect the stat carefully, is there provision for a neutral, that has never
been
connected? If so, connecting it will improve the responsivenes of the
thermostat
by switching in an accelerator (actually a high value resistor) that
pre-heats
the bi-metal strip causing it to switch off sooner than it would do
naturally
thus reducing the "hysteresis" of the system.
The neutral needs to come from the same CU circuit.

I'd find it strange if *cylinder* stats had this kind of compensation.
An insulated cylinder full of water heated to the desired temperature
will surely lose heat much more slowly than a room.

OK I've come to realise the next thing to try is to get hold of a latching
relay and wire this up as a two-stat system, with one stat turning the
boiler on & the other off via the relay (as described before in here).

I didn't really want to get into this, and I was put off by the wayward
strap-on stat behavior, but perhaps the activation setting will be fairly
consistent (though whether it closely matches the setting on the dial
remains to be seen!). If it works I'll be able to set the difference between
'On' and 'Off' via the two stats.

Just hope it goes smoothly.

Egremont.

"Skeggy" wrote in message news:45940c9a.0@entanet...
OK I've come to realise the next thing to try is to get hold of a latching
relay and wire this up as a two-stat system, with one stat turning the
boiler on & the other off via the relay (as described before in here).

I didn't really want to get into this, and I was put off by the wayward
strap-on stat behavior, but perhaps the activation setting will be fairly
consistent (though whether it closely matches the setting on the dial
remains to be seen!). If it works I'll be able to set the difference
between
'On' and 'Off' via the two stats.

Just hope it goes smoothly.

Egremont.



.... now known as 'Skeggy'

On Wed, 27 Dec 2006 20:06:50 -0000, "Egremont"
wrote:

PS - I know there are better and more expensive/complicated solutions which
I'm trying to avoid if I can.

It's about a hundred quid or so for a thermocouple pocket,
thermocouple and fully configurable PID controller with relay output,
maybe only half that on Ebay.

Just how cheap and nasty do you want to be?


--