DIY radiator fan?

[itm2]

I've been trying out a Speedcomfort double radiator fan, and have been disappointed with how meagre the airflow is, especially for a unit which cost just under £80.
I was wondering if anyone had made a DIY radiator fan - e.g. using PC case cooling fans? In my case I wouldn't need a temperature sensor, as I could control the power using Home Assistant and my Hive heating system.
80mm case fans can be had for a couple of pounds on Aliexpress, but the power supply would be a little more messy/complicated.
Any experiences/ideas would be very welcome.

 

[ericmark]

I did experiment at mother's house, I had a thermostat in a box for brewing beer, a homemade version of one of these

so I switch it on once the radiator got hot, I tried a few fans, and in the main they were too big, as soon as the fan started the water cooled and the fan turned off, used a bit of insulation around the return water pipe holding the sensor onto the pipe, but the fan was other end of the radiator.

This

seems a good idea, however the problem is the TRV needs to open and close not being affected by the fan, so a really weak USB fan seemed about right moving the air a little quicker, but not to a great extent. What we really want to emulate is the old Myson fan assisted radiator

I did have one, and there was no lock shield valve, and it worked well with pre-modulating boilers, with no restriction on water flow, the control was with fan speed not water flow, to work with a modulating boiler, they would need to be plumbed in series rather than parallel.

So I came to the conclusion that to get extra heat, the easiest would be to repaint the radiator mat black. And yes, the Myson fan assisted is mat black, and it is designed more like a car heater.

I looked at the idea of boxing in the radiator

and increasing the chimney effect, but with the heat pump the triple plate radiators are being made, so controller £30, fan £15 and new radiator £400 well worth some experiments, but unlikely to look very pretty, so why not paint it matt black, and try the easy way first.

 

[StephenStephen]

I bought a USB fan for about a fiver - propped at the bottom of the radiator pointing up diagonally between the 2 plates.
I plug it in manually when I put that heating loop on - I'm surprised how much difference it makes.

 

[itm2]

Thanks for the feedback.

I bought a USB fan for about a fiver - propped at the bottom of the radiator pointing up diagonally between the 2 plates.
I plug it in manually when I put that heating loop on - I'm surprised how much difference it makes.

That's interesting. Have you measured the impact in any way? For example, whether it reduces the amount of time that it takes to increase the room temperature by 1 degree?
I tested the SpeedComfort (dual fan version), and found that there was no improvement in this respect. I have a smart thermostat which reports temperature changes, and my heating comes on at the same time every morning (on at 08:45, off at 10:30). I measured the temperature increase over the same 1 hour period on 2 consecutive days - the first day without the Speedcomfort fitted. and the second day with it installed. The outside temperature was the same on both days (give or take 0.5 degrees). The radiator is behind the sofa in the lounge (a 5x4m room), underneath a window. The thermostat was placed on the coffee table in front of the sofa.The results:
- on day 1 (without the Speedcomfort) the temperature increased 0.76 degrees from 09:30 to 10:30
- on day 2 (WITH the Speedcomfort) the temperature increased 0.75 degrees from 09:30 to 10:30
At that point I put the Speedcomfort back in its box and returned it!

 

[foxhole]

Doesn’t increase the temp just marginal speeds up heat distribution.

 

[StephenStephen]

Sorry - I just noticed that the room seemed warmer than without it!

 

[StephenStephen]

The way I think about it is that it cools that radiator, therefore increases the heat loss across that particular radiator.
The situation I'm using it in is turning it on as I turn that zone on for about half an hour, when I want a higher proportion of the heat in that particular room.

 

[itm2]

Interesting. In my case the room neither felt warmer nor did the thermostat show it as being warmer (and we sit literally 12 inches from the radiator). I really didn't understand the point of it :0(

 

[ericmark]

To get more energy out of the radiator, it needs to have a lower return temperature from the radiator. This is controlled with a standard radiator with no fan by the lock shield valve, which is a little multi-function.

So, mother house, I set the lock shield valve by looking at the TRV report. The valve was opened if the target was above the current, and vice versa.

For this to work, the boiler must run, so during the set-up, the wall thermostat was set high, so it did not switch off, but this then means if left like that, during the summer, the boiler would continue to cycle. So it is important to know what the wall thermostat is registering, this was a bit useless, in spite of being a good thermostat, as one it used a mark/space system to stop overshoot, which messed up the modulating boiler, and two no idea of what temperature it is seeing, this does show both current and target, so much better, however since it will not link to TRV heads, still not what is really needed, the Wiser

does show target and current, and will connect to TRV heads, and if you have linked TRV heads, then likely each room will reach spot on the temperature set, so what will a fan do?

The answer is speed, the fan can speed up how long it takes, but the lock shield valve also controls speed. So back to basics, the fan assisted radiator does not have a lock shield valve, and was designed to be used with a non modulating boiler, if used with a modulating boiler, that modulates to gain the latent heat, then unless just one of many radiators like the odd kick space radiator in the kitchen, they need to be in series, not parallel.

So can't really see how the radiator fan is really going to work.

 

[StephenStephen]

To get more energy out of the radiator, it needs to have a lower return temperature from the radiator. This is controlled with a standard radiator with no fan by the lock shield valve, which is a little multi-function.

So, mother house, I set the lock shield valve by looking at the TRV report. View attachment 373337The valve was opened if the target was above the current, and vice versa.

For this to work, the boiler must run, so during the set-up, the wall thermostat was set high, so it did not switch off, but this then means if left like that, during the summer, the boiler would continue to cycle. So it is important to know what the wall thermostat is registering, this View attachment 373339was a bit useless, in spite of being a good thermostat, as one it used a mark/space system to stop overshoot, which messed up the modulating boiler, and two no idea of what temperature it is seeing, this View attachment 373340 does show both current and target, so much better, however since it will not link to TRV heads, still not what is really needed, the Wiser View attachment 373341does show target and current, and will connect to TRV heads, and if you have linked TRV heads, then likely each room will reach spot on the temperature set, so what will a fan do?

The answer is speed, the fan can speed up how long it takes, but the lock shield valve also controls speed. So back to basics, the fan assisted radiator does not have a lock shield valve, and was designed to be used with a non modulating boiler, if used with a modulating boiler, that modulates to gain the latent heat, then unless just one of many radiators like the odd kick space radiator in the kitchen, they need to be in series, not parallel.

So can't really see how the radiator fan is really going to work.

That seems very complicated - how about just using a fan to cool the radiator, and thereby heat the room?

 

[ericmark]

That seems very complicated - how about just using a fan to cool the radiator, and thereby heat the room?

If the central heating has not been set up correctly, then a fan on the radiator may put more energy into the room.

Thanks for the feedback.

That's interesting. Have you measured the impact in any way? For example, whether it reduces the amount of time that it takes to increase the room temperature by 1 degree?
I tested the SpeedComfort (dual fan version), and found that there was no improvement in this respect. I have a smart thermostat which reports temperature changes, and my heating comes on at the same time every morning (on at 08:45, off at 10:30). I measured the temperature increase over the same 1 hour period on 2 consecutive days - the first day without the Speedcomfort fitted. and the second day with it installed. The outside temperature was the same on both days (give or take 0.5 degrees). The radiator is behind the sofa in the lounge (a 5x4m room), underneath a window. The thermostat was placed on the coffee table in front of the sofa.The results:
- on day 1 (without the Speedcomfort) the temperature increased 0.76 degrees from 09:30 to 10:30
- on day 2 (WITH the Speedcomfort) the temperature increased 0.75 degrees from 09:30 to 10:30
At that point I put the Speedcomfort back in its box and returned it!

It seems you have carefully measured the temperature raise, and I have also looked at the idea of faster recovery of key rooms. My boiler (oil) has a fixed output of 20 kW, if I have not used 20 kW the boiler will start to cycle off/on, but with a gas boiler it is not a fixed output, typical gas boiler can modulate between say 8 and 28 kW, I tried in my late mother's house to find a way to know the output of the boiler, and failed.

The boiler was set to start at 7:00 am, so the house was warm at 8:00 am when the carers arrived to shower mother, the rooms would be warm, but we found it 10 am approx before the room was warm. So instead of 17ºC overnight and 20ºC at 7 am, tried setting at 22ºC at 7 am then down to 20ºC at 8 am. This worked, it seems the anti hysteresis software in the TRV heads were OTT.

So it was lifting the room 3ºC in an hour, against you 0.76ºC, but move to this house with micro-bore and the temperature raise is a lot less, easy enough to read as the TRV head records it. also the wall thermostat as you can see they don't seem to match, but same day in the same room, 17th Feb 2025 my living room. The TRV is a Kasa and the wall thermostat Wiser. The Nest Gen 3 is not so helpful

all it does is tell me how long it asked the boiler to fire. Which, since two thermostats in parallel, tells me nothing. Even if only one, would not really help, as calling for heat does not mean the boiler is running, that will only happen if the return water is cool enough.

I find it hard to work out how long the boiler runs for, if the boiler was plugged in, I could use the energy monitor to record how long it has run for, but it is powered from a FCU so unless I sat down in the flat under the main house with a stop watch, hard to measure how long it runs for.

So the big question is how can we measure what a radiator gives out?

 

[Why Not Indeed]

Have a look at this:

DIY Controller for radiator fans

Hey, would you mind sharing your radiator fan setup? I’ve been meaning to strap a few 140mm Noctua PWM fans to the bottom of the K3 rad in my living room, with a PWM controller driving them.

community.openenergymonitor.org

Lots of useful detail in there.

 

[ericmark]

I have looked at these a few times

they are really for UFH, but may fit on a standard TRV head, but never tried. The question is, if set up correctly, how do you allow extra water through the radiator when the fan is cooling it?

As I said, I had a homemade one of these, ⁣

this allowed me to select when the fan runs, so fan at the TRV end so blowing room temperature air past TRV, and the sensor on the lock shield valve end, relied on flow being TRV to lock shield, once the radiator was hot the return pipe would get hot and the fan would start, for around 5 minutes only, and then radiator too cold again, if I opened the lock shield valve to match output with fan blowing then it was open too far to stop hysteresis when fan not blowing, so tried different fans, fan heater set to fan only pushed out far too much air, and computer cooling fan worked OK, but for what difference it made, not really worth it.

In the main, it would seem the best option is to use a larger radiator.

Once the room is up to temperature, the TRV starts to close, with a fan switching on/off, you could hear the TRV motor continually adjusting water flow, so the batteries would not last long.

The fan only helps with the initial warm up, once room is warm not required.

In the main, I found cold rooms were due to the boiler being switched off by a wall thermostat, rather than being allowed to modulate as it is designed to do. So cure was to use a linked TRV head in the cold room.

Yes, I did try with a non modulating boiler, this was even more of a problem, the idea I finally used to speed warm up, was to set a delay on less important rooms. So if we get home and cook a meal straight away, then want rooms in the order of kitchen, then dinning room, then living room, and finally the bedrooms.

If the delay is too long the boiler will cut out first, so found around 10 minutes with first three rooms, this means radiators heat up in sequence. The bedrooms are a problem as around 3 to 4 hours before required, so would turn wall thermostat down for ½ hour before wanting bedroom to warm up, then back up again when bedroom TRV calls, ensuring the boiler is actually running.

But really this is only needed with non modulating boilers, modulating boilers don't turn on/off or at least should not be turning on/off all the time with a well balanced system.

 

[Why Not Indeed]

how do you allow extra water through the radiator when the fan is cooling it?

You don't. The water flow is constant but the air speed is varied according to a curve relative to the room temperature.

 

[ericmark]

You don't. The water flow is constant but the air speed is varied according to a curve relative to the room temperature.

Assuming radiators are trimmed in correctly, the input and output temperature with TRV fully open, should be around 15ºC, once the TRV starts to work, then the difference will increase, so yes if full output is required water flow is constant, it is only once the temperature has reached the set point that the water flow will change.

However, if there is a fan, which extracts more energy, then that 15ºC must increase, it is a long time ago when I studied applied heat, odd I have just found my old text book dated in the 1970s so it was some time ago, but if we start at 75ºC then return is 60ºC, but to get same heat starting at 60ºC then the return temperature with same output will not be 45ºC, but a lot lower, not worked it out, but likely looking at 30ºC.

So the theory is you could double the output, but likely no more. So if the room takes say 1 hour to warm up, instead it will take ½ hour to warm up.

But this assumes the boiler has that spare energy in the first place, so this would not apply if the heating had only just been turned on, it would only apply is a single room had just been turned on. My boiler produces 20 kW, if all motorised valves (including TRV's) are open, this is shared between 15 radiators, of which 1 is therm-syphon, 4 have mechanical TRV's and controlled by one motorised valve and pump, and 10 are from another motorised valve and pump and 9 of those 10 are electronic so have not only the temperature set but also the times set.

So yes in my house one TRV could open at a time when all others satisfied or closed, but all my TRV heads are not linked to the hub, to how fast the room will heat up, it depends if boiler running or not.