Just an FYI ...
At my last job I was involved in the heating and a/c, so got to know "a little" (well more than perhaps I want to) about some of the stuff available. We were using Honeywell valves mostly because that's what the system designer specced initially. Anyway, some random snippets of information that may be of interest.
IIRC, most of the radiator valves worked on a 2mm stroke, while the three and four port valves we used in the heating/ac system used an 8mm stroke - that's important as it means there's more to actuator compatibility than just having the right thread.
On the 3/4 port valves we were using, there was a choice of head - we used the basic thermo-hydraulic (wax capsule) heads because we were tight (and the others would have been overkill anyway without better controls).
Or, there were actuators with a 3-wire motor open/motor close function - simple 2 winding AC motor and a capacitor, apply AC to one winding and the motor runs one way, apply it to the other winding and it motors the other way, in both cases the capacitor provides a split phase to the non-powered winding.
Or most expensive of all were self contained heads with 24V supply and 0-10V control input.
The wax capsule heads are, as already said, not a lot more than a wax capsule (the wax expands when heated) and a heating element. The spec sheet for them showed that from cold they could take 2 minutes to start opening (less if they'd only recently closed), and could take anything up to 20 mins to fully close after removing power (the longer they'd been on, the longer they took to cool). I checked with Honeywell and they said there were no restrictions on duty cycle, so it was OK to cycle them if you wanted some form of crude variable control.
Just had a quick search, and one of the first results was this which is (I think) just what we were using :
http://www.havelockcontrols.co.uk/itemdetl.php?itemprcd=50350510The actuators have been updated since, and now have options with built in switch which we didn't have then :
http://www.havelockcontrols.co.uk/itemlist.php?findchil=1&findclas=136742From the spec sheet, I think they have a PTC thermistor in them to reduce power (and excess heating) once they are up to temperature - the spec sheet also shows the typical opening and closing times. The prices have come down a tad as well since I was looking at them.
It was either this thread or another similar one that got me thinking - not that I'll be in a position to try it in the forseeable future. One thing I have worked out in the aforementioned previous employment is that radiators are a pig to control. When the valve opens, the rad heats up very quickly as it floods with hot water. So if you try controlling them with a stat, you tend to get very hot rads a short time after the stat kicks in, followed by a temperature overshoot, a cooling period as the rads cool off, and finally the cycle repeats.
So I was wondering if there was milage in fixing a temperature probe to the rad itself - and attempt to modulate rad temperature. In effect, have nested control loops - an outer one sensing room temp and generating a desired rad temp, an inner one acting on the desired rad temp and generating a valve control signal. The time constants of the two systems are significantly different which is part of the control problem.
Another thought I've had is related to the thermal store I've fitted in the flat. For that I've fitted a modulating pump for the CH and TRVs on all rads - doing away with the room stat. The flow temp is controlled by a TMV since the store temp at the heating tapping is both variable and often too hot. I was thinking about the practicality of sensing flow rate, and adjusting the flow temperature based on it (by replacing the TMV with a motorised head mixing valve) - it would end up as a closed loop weather/room temp/desired heat compensator. In effect, I figure that if the flow goes down, it means the demand is low and so I could drop the flow temp - so instead of having rads that are hot at one end, I could have them lukewarm most of the way across. Conversely, if the flow rate goes up, then that implies cooler rooms, and the flow temp can be turned up to suit. Unless I've missed something, that should mean better temperature control since I'd be aiming for a more or less constant TRV opening, rather than requiring a variable room temperature to trigger a variable opening.
One of the aforementioned 3 port valves, with a thermal actuator, would do the job - modulated by simple "bang-bang" on-off control to maintain the flow temp at whatever the control loop asked for.
Of course, the ultimate would be to have a circulating pump for each rad, and modulate the loop temperature according to room demands. But I doubt if most people are wanting to buy a separate pump for each rad
The only other problem is cycling of the boiler which is why i designed this system in the first place but i just cant overcome the fact that the boiler is too big.
You may want to consider a buffer tank.
Instead of driving the heating direct from the boiler - have the "boiler top" load a buffer tank, ie just have the boiler put hot water in at the top and take cool water out at the bottom. Then connect the CH using a separate pump, taking hot water from the top and returning the return flow to the bottom. The hydraulic demands of the two circuits are then decoupled - you can have enough flow to keep the boiler happy, and little enough flow to keep the heating happy - including the use of a modulating pump which really does make for a quiet system once you start throttling the flow with TRVs and stuff.
Go one step further, put a coil up inside the big tank and heat your DHW with it - then you've got a "thermal store"
