Controlling the speed of a Fan Heater motor

PhilHornby

Senior Member
I'm embarking on Phase 2 of my project to enhance the temperature control of a Dimplex Oakhurst heater. Phase 1, published HERE makes uses of the heater's Infra Red Remote Control functionality, and is a vast improvement on the in-built mechanical thermostat.

However, the Fan - when it is on - always runs at full speed and I was hoping I could improve on this. It is the fact that the fan turns on and off, that is as much a source of annoyance, as the fact that it always runs at max. speed. I am aiming for a constant, but much lower, background noise level.

The scheme I have in mind, is based on PWM modulating the heating elements to match the room's heat loss and 'somehow' controlling the speed of the fan to match. I don't know if this speed control would involve a simple matching of the average power applied to the heating elements, or if it would be necessary to monitor the output air temperature. As I delved into the workings of the Fan Heater section, I realised I've got other issues to resolve first...

It turns out that I know a lot less about the functioning of AC motors, than I thought I did :confused:

Here are some photos of the Fan and its motor :-

https://www.dropbox.com/s/sfmeims8bd4gkjg/20170719_211335.jpg
https://www.dropbox.com/s/r2q3bx2xp2e2s6u/20170719_211346.jpg

(The forum appears to be out of disk space again, so I can't upload them directly :()

If anyone has any observations on said motor, I'll be glad to hear them.

My various googlings, lead me to believe that since I don't have a big, inertial load, the only workable scheme will be to vary the frequency of the AC voltage applied to the motor. If so, I envisage synthesising a suitable 'analogue' signal with a Picaxe and using that to modulate the voltage applied to the motor - derived (presumably) from the rectified mains input.

If that is correct, it gives rise to the following question: How close to a Sine wave, does the motor control voltage have to be?

I'm pondering the difference between using a Square wave, a Triangular wave, an RC-derived sawtooth, or the signals documented here: http://www.me.ua.edu/me360/spring05/Notes/Topic17-AC_Motor_Speed_and_Other_Motors_sv.pdf (synthetic sine wave or Pulse Density Modulation...).

A supplemental question: Why does this motor always turn in the same direction? Obviously this is a very important feature of a Fan heater motor, but how is it achieved? Is it an inherent feature of the motor design, or are the existing heater control electronics involved?
 
Last edited:

inglewoodpete

Senior Member
The motor is the cheapest form of AC motor: shaded pole. If you take the motor apart (there should not be any unexpected springs etc), you will see that there is a single turn of copper (wire) embedded in the stator. On its own, the mains-voltage (230V in Australia) winding on the stator would only create an oscillating magnetic field - no good for turning the rotor. The shorted turn causes the magnetic field to become 'oval shaped', giving the more rotating magnetic field required to pull the rotor around and spinning the motor shaft. The location of the shorted turn determines the direction of the rotating magnetic field and consequently, the direction of rotation.

I have used a bathroom exhaust fan for "sucking" warm/hot air from under my house's iron roof for heating in (our Australian!) winter. To do this, I took the shaded pole motor apart (down to the last not and bolt) and flipped the stator over before reassembly. The result: the orientation of the shaded pole is reversed and the fan now pulls air into the room instead of pushing it out.

Back to your real question. You should be able to control the speed of your shaded pole motor by adding an AC-mains rated capacitor in series with its winding. You will need to experiment but start with a 2uF or 1uF and see if you get the speed change you want. I trust you can work safely with main voltages and appliances. A 2uF capacitor has a reactance of 1.6kohms at 50Hz.
 

tmfkam

Senior Member
This type of motor, as with other induction motors, runs at a fixed speed determined by the number of pole pairs, and the supply frequency. It may be possible to vary the supply voltage which will reduce the motor torque and possibly prevent it from reaching full speed. It may also be possible to PWM the supply voltage which again may reduce the torque. It may be possible to reduce the time that the motor is energised for (5 cycles on, 10 cycles off) which will give full torque but may once again reduce the speed that the motor runs at due to the air resistance on the fan. The most reliable way of controlling the speed is, as you surmise, to reduce the frequency.

I have built a 240V 50Hz to 240V at 25Hz to 250Hz (with PWM control) to demonstrate these principles to students. This is a non-trivial design and I would hesitate to suggest it due not only to the danger involved with mains voltages, but also the danger of getting the active devices to switch correctly and reliably in order to prevent loud bangs and bright flashes.

I might be tempted to suggest driving an audio power amplifier IC (something like a TDA2020 or similar) from a variable frequency source and feeding that output into a small mains transformer connected with the 12V secondary to the audio amplifier and the 240V primary to the motor. As the shaded pole motors are not that powerful, this should give acceptable results with significantly less danger.

Motors will run perfectly well being powered by a square wave (my 240V 25Hz - 250Hz design was a square wave unit) but you will get reduced efficiency from the motor, and potentially some increased heating of the motor windings. If you were to incorporate PWM into the design, the losses and heating of the windings could be reduced by reducing the duty cycle and thus the amount of time the supply is 'stuck' on maximum with no change in voltage.

Should you be interested I could email you some of the designs discussed. I'd rather not get stuck in a flame war again trying to convince other people that a time proven design will actually work as claimed.
 

westaust55

Moderator
Industrial VVVF drives use PWM square waves but for each half cycle the initially generate short pulses gradually getting longer until quarter of a cycle then reducing in width.
This provides something more like. Sinusoid also current on average and reduces heating within the motor.
If the current is a square wave / non sinusoidal that means ic contains harmonics which do nothing more than heat the motor.

With respect to a square wave voltage from what I have seen in the past, there is typically a spike on the leading edge of each group of pulses that can be more than twice the supply voltage. Industrial motors for us with VVVF drives are typically provide with higher rated insulation.

Adding a reactor to the VVVF output can help reduce the harmonics to the motor but will also slightly reduce the available voltage.
 
Last edited:

hippy

Technical Support
Staff member
Would it not be easier to replace the AC motor with something which can have its speed controlled more easily ?
 

Jeremy Harris

Senior Member
Would it not be easier to replace the AC motor with something which can have its speed controlled more easily ?
That would be exactly the approach that I would take. My choice would be to replace the motor with a brushless DC model aircraft motor and speed controller. A Picaxe can then directly control the motor speed using the servo command.

I've now replaced both the motor in my milling machine with a (large) model aircraft BLDC motor, and the motor in my bench top lathe with another (even larger) model aircraft BLDC motor. These are very controllable, as well as being very powerful for their size, and having no brushes will have a long life. The downside will be the need to provide a low voltage DC power supply, but the chances are that the existing fan motor isn't very powerful, probably only a few tens of watts at most.
 

hippy

Technical Support
Staff member
the chances are that the existing fan motor isn't very powerful, probably only a few tens of watts at most.
14W at 230V AC according to the picture. Not much at all. The blade assembly for pushing hot air out will probably have bearings and be a very light load.
 

AllyCat

Senior Member
Hi,

Yes, substituting a brushless type of motor (to minimise noise) is probably the way to go. I'd look for a computer fan (checking if 5 or 12 volts); I believe the "4-wire" type can be speed-controlled by a PWM signal (rather like a servo), normally from the PC motherboard. Of course the fan must be be "upstream" of the heating coils, perhaps more so than the original motor. It may be "inconvenient" to have to provide a low voltage dc supply, but in an "enclosed" application like this, perhaps it's safe to try one of the ridiculously low-priced "USB mains adapter"s from ebay or even a Pound shop.

Alternatively, the first two answers above are almost correct, but just a little "theory" is needed: Most ac (only) motors (as opposed to the "ac/dc" commutator type used in power tools, etc.) create a rotating magnetic field, usually formed by the current flowing in two or more coils.

You can apply almost any voltage waveform to a resistor and the current waveform will be very similar, with an amplitude dependent on the resistance. But for inductors (coils) and capacitors, that is true only for pure sine-waves and the current is "out of phase" with the applied voltage. With a capacitor the peak current "leads" the voltage by 90 degrees and for a coil/inductor the current "lags" by up to 90 degrees. Also the amplitude of the current depends on the frequency of the sine wave, not just the applied voltage.

"Large" motors use a 3-phase supply directly from the electricity company. The three phases are separated by 120 degrees (or 60 degrees to the negative peak of the next phase) so the motors have three coils at 60 degrees and produce an excellent rotating magnetic field. Smaller motors use two coils at right angles, with one coil driven via a "starting" capacitor so that its current leads the first to produce a field which rotates in a defined direction. Even smaller motors (like yours) use only one coil/magnet, but "split" one of the magnetic poles with a "short-circuit turn" (i.e. inductive load) to distort the magnetic field sufficiently to make the rotor start turning in the required direction.

A "synchronous" motor has a permanent magnet in its rotor and turns exactly at the same speed/frequency as the ac supply (taking into account the number of coils/poles). It's basically similar to a stepper motor. However, an "induction" motor (like yours) doesn't have a permanent magnet, but a "cage", normally made of copper wires, similar to a tiny bird cage. When the copper wires move through a magnetic field, they generate a current which (being short-circuited by the rest of the cage structure) sets up its own magnetic field that interacts with the original field and causes the cage/rotor to turn. But because the the wires must "cut" though the magnetic field lines, the rotor turns slightly more slowly than the rotating field, called the "slip". This is usually just a few percent, rarely more than 10%, so the motor is normally considered to be constant speed.

[Reminiscence] FWIW, I had one of the very first "domestic" VCRs (pre VHS and Beta) where the "head disc" had to rotate at exactly frame rate" (25 Hz). Of course the mains frequency is not exactly 50 Hz and in those days transistorised servo loops and inverters were in their infancy. So the head disc was driven (via a rubber belt to give the correct "gear ratio") by a mains induction motor which carried an additional copper disc on its shaft. The servo control loop generated a dc current which drove electromagnetic poles each side of the disc to act as an eddy current brake. So the control loop was able to slow down the motor by just the few percent needed to synchronise the "head-crossovers" with the video signal. But yes, I did have to call out the "technicians" to make it actually work. :( [/Reminiscence]

The "mains inverters" sold by Maplin and the like normally use "quasi sine wave" which is often just bidirectional pulses, rather narrower than the 10 ms of mains cycles. Some may use more than one "pulse" level each side of zero, but IMHO, if you do want to use the original motor, it will need a variable-frequency inverter, preferably using "Class D" (modulated PWM).

Cheers, Alan.
 
Last edited:

tmfkam

Senior Member
You can apply almost any voltage waveform to a resistor and the current waveform will be very similar, with an amplitude dependent on the resistance. But for inductors (coils) and capacitors, that is true only for pure sine-waves and the current is "out of phase" with the applied voltage. With a capacitor the peak current "leads" the voltage by 90 degrees and for a coil/inductor the current "lags" by up to 90 degrees. Also the amplitude of the current depends on the frequency of the sine wave, not just the applied voltage.
These were the principles the 240V @ 25Hz - 250Hz inverter I designed to prove. While changing the frequency does indeed adjust the speed of an induction motor, whether shaded pole or squirrel cage, this is only within certain limits. Too low a frequency will leave the motor 'stuttering' and/or failing to start, so too does increasing the frequency. I chose the upper limit of 250Hz as the motor I was tasked with using the device with would slow down and stall above 175Hz or so, likewise the lower limit. Other motors may have different limits, but limits all the same. If I recall shaded pole motors had a rather more limited speed range, but varying the frequency to say 35Hz should have the desired effect of reducing the speed and noise noticeably.

I agree with those who suggest changing the motor which could be the simplest solution electrically and significantly safer, if requiring somewhat less electronic ingenuity.
 

PhilHornby

Senior Member
Many thanks for all the responses.

I've read and pondered everything that's been said, which I'll summarise :-

  • It would be far better to have a fan motor that was actually suitable for speed control in the first place. For a variety of reasons (not least being the 'cost' of this heater), I'm not going to attempt to change the motor, but I could be tempted to swap the entire fan heater assembly. The fan heater section is modular and bolts to the underneath of the appliance; it may be possible to exchange it completely - or even, add an external 'blower unit'.
  • Restricting the available power to the fan motor, by inserting series Capacitance, Inductance, Resistance, Triac etc is not likely to produce controlled changes in speed, without some kind of feedback mechanism. That said, I was pondering some kind of measurement of the outflow temperature, so just maybe, that would complete the circle?...
  • Fully-powering the motor for a given number of cycles, then removing power for another period, may produce results. That is what I thought I would do for the actual heat control, so it's a logical experiment to try. I'm contemplating using an SSR as a starting point for this...

I've decided that my hair-brained scheme to somehow synthesis a variable-frequency, high-voltage AC drive voltage from rectified and stored Mains is a non-starter. (I keep ending up with voltage-doublers in the circuit, which scares even me :p).
@tmfkam - I'd be delighted if you would share some of your existing designs - either via Dropbox, or email to DELETED (this is not a general-purpose email address; it will become a "Spam Trap" in the very near future!)
(The first spam email was sent to that address on 15/8/17, by info@cactt.cn - so I thought I'd return the favour ;-)

In the meantime, I've been out and bought a somewhat cheaper Fan Heater to experiment on: http://www.homebase.co.uk/en/homebaseuk/upright-thermal-fan-heater---2kw-167066 :D

UPDATE: I've just stumbled on a treasure trove of information ... on, of all places: the Microchip website!
 
Last edited:

inglewoodpete

Senior Member
* Restricting the available power to the fan motor, by inserting series Capacitance, Inductance, Resistance, Triac etc is not likely to produce controlled changes in speed, without some kind of feedback mechanism. That said, I was pondering some kind of measurement of the outflow temperature, so just maybe, that would complete the circle?...
I'm not convinced that feedback and incremental control of a heater fan is necessary. Probably for the heating element but not for the fan. For a selected heat setting, the energy is being turned into heat regardless of the fan speed. The fan motor does not need a feedback mechanism since the load will be constant: the fan will seek a speed/load equilibrium for the amount of power being supplied.

My 35MJ natural gas space heater (as opposed to the 7.2MJ Oakhurst/Dimplex) uses a shaded pole motor with a 3-position rocker switch: Off-Slow-Fast. It comes that way from the factory: very much set-and-forget, no fancy electronics required.
 
Last edited:

johnlong

Senior Member
Hi
Are totally sold at going down the electronics route to controll the mains
could I suggest if you have the space either within the fire place or at the consumer unit end
the use of an auto transformer.
You could then pick out 3 voltages from it and 1 direct mains to give you 4 settings without altering
the frequency the units where designed around.
You could then select the 4 voltages using a Picaxe to select them dependent on a tempreture reading
The mains on off states for the voltages can then be controlled by SSR's. The CPC44055ST can be operated
via 5volts dc and with heat sinks can pull 20Amps
 

AllyCat

Senior Member
Hi,

I believe the OP's primary aim was to reduce the speed "smoothly", to avoid any sudden "steps" in the background noise level. But...

... uses a shaded pole motor with a 3-position rocker switch: Off-Slow-Fast.
Do you know how that is achieved (additional capacitor, resistor or...) ? After all it might switch in an additional pair of poles (or modify the "winding" in some other way). To me, it seems the "appeal" of the basic induction motor is that it combines the function of a (step-down) mains "transformer" in a single module. So any attempt to add "electronic control" doesn't make much sense.

Going slightly Off Topic; Many, many years ago I was quite surprised to discover that the speed of a car/automobile heater "blower" motor was increased by adding a resistor in series with the rotor (field) winding. This resistor reduces the "back-emf" in the motor, so the stator coils draw more current and the motor rotates faster.

Cheers, Alan.
 

PhilHornby

Senior Member
Are totally sold at going down the electronics route to control the mains
could I suggest if you have the space either within the fire place or at the consumer unit end
the use of an auto transformer.
Thanks for the suggestion - I'll add it to the pot!
 

techElder

Well-known member
You could experiment with using a D.C. field to saturate the core of a transformer. The D.C. current would be easily controlled and the winding would provide isolation.
 

PhilHornby

Senior Member
I'm not convinced that feedback and incremental control of a heater fan is necessary. ... the fan will seek a speed/load equilibrium for the amount of power being supplied.
Further googling reveals lots of contradictory stories, so I think I'm going to have to do the experiments to find out. At least I know what sort of motor I'm dealing with now, and how it works!
 

PhilHornby

Senior Member
You could experiment with using a D.C. field to saturate the core of a transformer. The D.C. current would be easily controlled and the winding would provide isolation.
Sounds interesting - please elucidate :)

Is it a variation on this technique :-
someone @ https://www.electronicspoint.com/threads/shaded-pole-motor-control.122501/ said:
Get a 100VA transformer, something like 240V to 40V.
Put the 240V windings in series with the motor and connect to the mains - nothing happens. Short the 40V windings - motor runs at nearly full speed but you are still working with AC.
Put a bridge rectifier on the 40V windings and short the DC output. Motor runs at nearly the full speed it did before but now you are working with DC. Put a transistor (I normally use a darlington - TIP121 for a 40V secondary and add simple control electronics. You can normally get up to about 90% full speed.
Approximately 10% losses occur in the transformer, bridge rectifier and darlington. This could be reduced by using a higher voltage secondary and a MOSFET when the losses could be 5% or lower. The control electronics side is isolated by the transformer so you are working with low voltage DC.
 

johnlong

Senior Member
Hi
I had a dig back in my notes so here is a bit more for the pot you might like to consider
http://www.tcf.com/docs/fan-engineering-letters/speed-control-for-fractional-horsepower-motors---fe-1000.pdf?Status=Master
the above looks at the diffrent aproches to motor control
http://www.pqmonitoring.com/papers/Transient Overvoltages/pwm drives.pdf
the effect of PWM on the motor windings
https://www.ab.com/support/abdrives/documentation/techpapers/ieee/APEC95 bearing.pdf
the effect on the bearings and arbour.
The consensis is that PWM increases wear and adds harmonic noise as wasted energy
It was due to these effects that stired me towards the autotransformer (variac) route and SSR's
very simple board PIN--Ressistor--OPTOCOUPLER (inbuilt in SSR)--SELECTED TAPPING=Fan speed
As you are in the UK the minimum mains voltage that the fan will operate at will be around 130-140volts mark
to try and add a degree of self balancing I use the following (With great thanks to Weastaust)
Code:
symbol desired_temp=b12
symbol Dry=b14 'temp from sensor
symbol bOptr=20
		
			select case desired_temp 'preset
	case<Dry	peek bOptr,b0 'free up b0
			if b0=7 then:b0=6:endif 'Set the max for b0 to pin  B.7 full power
			b0=b0+1
			b1 = b0 // 4 + 4 
			LOW B.4, B.5, B.6, B.7:pause 20	
			HIGH b1	
			poke bOptr,b0
			
	case>Dry	peek bOptr,b0		
		if b0=4 then: b0=5:endif 'set the min for b0 to pin B.4 low power		
			b0=b0-1
			b1 = b0 // 4 + 4 
			LOW B.4, B.5, B.6, B.7:pause 20	
			HIGH b1	
			poke bOptr,b0			
		end select
Hope this does not overflow the pot to much
 

techElder

Well-known member
Sounds interesting - please elucidate :)

Is it a variation on this technique :-
That's one way to do it. It's variously described as a "variable choke". Nothing about it can be described as energy saving or efficient. This method is usually considered for higher currents (I've used it with 1,000's of amps.)

However, with a little experimentation, you might find the right combination of components for more efficient operation.
 

tmfkam

Senior Member
Is it a variation on this technique :-

Quote Originally Posted by someone @ https://www.electronicspoint.com/threads/shaded-pole-motor-control.122501/
Get a 100VA transformer, something like 240V to 40V.
Put the 240V windings in series with the motor and connect to the mains - nothing happens. Short the 40V windings - motor runs at nearly full speed but you are still working with AC.
Put a bridge rectifier on the 40V windings and short the DC output. Motor runs at nearly the full speed it did before but now you are working with DC. Put a transistor (I normally use a darlington - TIP121 for a 40V secondary and add simple control electronics. You can normally get up to about 90% full speed.
Approximately 10% losses occur in the transformer, bridge rectifier and darlington. This could be reduced by using a higher voltage secondary and a MOSFET when the losses could be 5% or lower. The control electronics side is isolated by the transformer so you are working with low voltage DC.
In theory, this would only reduce the voltage applied to the motor, this may affect the speed, but the theory would suggest not. This would be a useful technique for generating a variable AC voltage as I suggested in an old thread where someone was looking for a way to vary an AC supply for a model railway.

Shaded pole motors were used in almost all BSR and Garrard turntables as the turntable motor for what seems like all of eternity. The reason being that they ran at a fixed speed with the various speeds of 16, 33&1/3, 45 and 78RPM selected mechanically by moving an idler wheel onto differing pulley diameters on the motor shaft.

The Jukeboxes that the place I work for, imported from the US, all have motors designed for 60Hz and so, on 50Hz (after we've fitted a 240V-120V transformer) the records play rather slowly. To remedy this, you need a frequency inverter which gives a 60Hz 120V AC output. The ones we buy in for this don't fill me with much confidence and I have suggested we make our own to my design. As we mostly sell the machines for refurbishment, we don't yet think it is profitable.
 
Last edited:

premelec

Senior Member
OT

@tmfkam - FWIW I would think a mechanical solution would be best - increase the drive pulley on motor by 60/50 - with 3D printing might be a simple custom part... ;-0 [cheap 60hz sine inverters I've found all have lots of RFI - another problem...]
 

techElder

Well-known member
In theory, this would only reduce the voltage applied to the motor, this may affect the speed, but the theory would suggest not. This would be a useful technique for generating a variable AC voltage as I suggested in an old thread where someone was looking for a way to vary an AC supply for a model railway.
My control suggestion was in line with others in the thread about changing motors.
 

tmfkam

Senior Member
@tmfkam - FWIW I would think a mechanical solution would be best - increase the drive pulley on motor by 60/50 - with 3D printing might be a simple custom part... ;-0 [cheap 60hz sine inverters I've found all have lots of RFI - another problem...]
I like that idea! Some do have removable pulleys and we can get 50Hz versions, some of the pulleys are part of the motor shaft but it could be possible to have a sleeve made. Good thinking.
 

RexLan

Senior Member
I wonder if the heating elements will survive with a reduced air flow for needing cooling. In a real furnace they would not and there is always a minimum CFM requirement. What is it for this heater?
 

Circuit

Senior Member
I wonder if the heating elements will survive with a reduced air flow for needing cooling. In a real furnace they would not and there is always a minimum CFM requirement. What is it for this heater?
My concerns are similar. The Dimplex Oakhurst unit that you link to has a sophisticated flame effect and there will probably be a range of plastic components within to enable this. Interfering with the airflow through the unit may lead to overheating within the decorative component. I am sure the manufacturers have done their best in the design stage to keep the noise to the lowest possible level. Reducing either the threshold for fan operation or the volume throughput of air could be hazardous.
 

geezer88

Senior Member
Here's some thoughts on your project:

Mechanically, your fan is mounted directly to the motor shaft, which, in turn is mounted on the fan shroud. This makes it very hard to replace the motor unless you get very lucky in finding a replacement. There are no pulleys or belts, so it would require an inordinate amount of work to accomplish a belt reduction.

The power a fan consumes goes up with the cube of the speed. So does the noise. The result of these facts means that to quiet the fan, it won't take much reduction in speed to accomplish your goal of noise reduction.

You also mentioned that the cyclic nature of the heater operation was as much a problem as the loudness of the fan. You could easily wire a second switch to allow the fan to run continuously while the heater is cycled to maintain temperature.

Shaded pole motors don't allow a very wide range of speeds without stalling. Getting a 20 to 30 percent reduction in speed can be accomplished with a simple triac dimmer control, or a resistor, either variable or fixed. You don't want to turn it down too low, or the motor will stall, overheat, and allow your heating element to overhead as well. So, while available, this option is dangerous because of the potential for overheating. I would not recommend it unless there is an overtemp device in series with the heating elements. Good heater already have this feature. Without doing some experimenting, I can only guess that the resistor would have to be sized wattage-wise to 10 or 20 watts; a big fellow.

Slowing the fan will raise the exit air temperature in proportion to the reduction in flow. The heating elements will run hotter as well. If they are visible in operation, you can judge the temperature rise in the heater elements by how red they get. If they don't run noticeably redder, then they will probably be fine. Life will be reduced, but it would be very difficult to estimate that.

If it were mine, I'd just wire the fan on continuously. The power used would be trivial. And the elimination of the cycling would not disturb my sleep as much.

tom
 

PhilHornby

Senior Member
Thanks for the latest input.

The flame effect uses the heat of 4x50W MR16 bulbs to produce the convection currents that swirl the water vapour around to produce the "smoke" - so even with the fan heater section off, the decorative component runs quite warm.

The intention is to reduce the heat output of the Fan Heater section, to just match the room's losses, which in turn means that the fan speed can be reduced. I know that I can't lower the fan speed while retaining the same heat output.

I intend to take some measurements of the output air temperature - and my final design may continually monitor this, to ensure it is not exceeded. I don't yet know if such measures will be necessary.

My guess is that currently, the air temperature never exceeds 80 °C - it seems to use the type of thermostat commonly found in gas boilers ... though the wiring seems a little odd. (The heater has two separate elements and there are two of these thermostats to act as overload cut-outs. However, they're wired in series with just one element, as far as I can tell...)

See: https://www.dropbox.com/s/o62hrsfyn2ip9c5/20170719_211325.jpg

and some more photos of the innards, for those who are interested...

https://www.dropbox.com/s/4nxokmszzoy6aqr/20170719_211404.jpg
https://www.dropbox.com/s/miv16tffhbmm1o1/20170719_214504.jpg
 
Last edited:

oracacle

Senior Member
If this was me, I would seriously consider dumping the entire internals.
Your application is quite specific. Often modifying something with a different application turns out to be harder than first appearance. It will also most likely pat to look at other heating systems, as far as I cant tell they are thermostatically controlled and simply control the temperature by the amount of heat they give off.

the chances are you will get a better noise reduction the changing the design of the fan itself. There is fair amount of work going on with basing fan designs of that of owl wings and feathers. I quick look shows that just serrating the trailing edge may have some effect.
1 thing I can pretty much guarantee I is that the design of the fan did not have much work put into it when dimplex made this thing.
 

inglewoodpete

Senior Member
Hmm. Seeing all that, it seems to me that you would be better off adding an off-the shelf 10+ amp dimmer to the heater element circuit.

Changing the fan speed does not actually change the heat output, unless you want to have those thermostats clicking in and out all the time.
 

PhilHornby

Senior Member
It will also most likely pay to look at other heating systems, as far as I can tell they are thermostatically controlled and simply control the temperature by the amount of heat they give off.

the chances are you will get a better noise reduction the changing the design of the fan itself. There is fair amount of work going on with basing fan designs of that of owl wings and feathers. I quick look shows that just serrating the trailing edge may have some effect.
1 thing I can pretty much guarantee I is that the design of the fan did not have much work put into it when dimplex made this thing.
A bit more info...

Those two obviously 'thermostat looking' devices aren't actually being used as thermostats - just overheat cutouts (I even wondered if one of them is actually a Mercury tilt-switch?). Within the 'Decorative' section, is a more conventional rotary, mechanical, clickety-pop type temperature control (very fashionable, 50 years ago!). I haven't the faintest idea what it measures, because it is quite isolated from the fan heater ... it must just be sensing heat conducting through the casing from underneath!. This was the reason for my Phase 1 project :) ... which achieves adequate temperature control.

A bit of background.

The Dimplex Oakhurst is located in a holiday-home (for want of a better description). It's a rough-built worker's cottage and has no gas supply. It only has access to a road, by passing over a neighbour's property ... thus ruling out Oil deliveries and such like. The previous owner was doing her bit to warm the atmosphere by burning wood and coal in two stoves. One of these stoves was swapped for the Dimplex and it looks the bees-knees. There is no available wall space in that room for any other heaters. When we're not there (most of the time :(), I position a convector in the middle of the room - ditto the kitchen, which otherwise has a small inconspicuous fan heater. All the convectors are controlled by my Picaxe-based heating system. I've seriously considered placing a convector in the fireplace, behind the Dimplex ... though I'm not sure if it would survive that experience ...
 

oracacle

Senior Member
A bit more info...

Those two obviously 'thermostat looking' devices aren't actually being used as thermostats - just overheat cutouts (I even wondered if one of them is actually a Mercury tilt-switch?). Within the 'Decorative' section, is a more conventional rotary, mechanical, clickety-pop type temperature control (very fashionable, 50 years ago!). I haven't the faintest idea what it measures, because it is quite isolated from the fan heater ... it must just be sensing heat conducting through the casing from underneath!. This was the reason for my Phase 1 project :) ... which achieves adequate temperature control.

A bit of background.

The Dimplex Oakhurst is located in a holiday-home (for want of a better description). It's a rough-built worker's cottage and has no gas supply. It only has access to a road, by passing over a neighbour's property ... thus ruling out Oil deliveries and such like. The previous owner was doing her bit to warm the atmosphere by burning wood and coal in two stoves. One of these stoves was swapped for the Dimplex and it looks the bees-knees. There is no available wall space in that room for any other heaters. When we're not there (most of the time :(), I position a convector in the middle of the room - ditto the kitchen, which otherwise has a small inconspicuous fan heater. All the convectors are controlled by my Picaxe-based heating system. I've seriously considered placing a convector in the fireplace, behind the Dimplex ... though I'm not sure if it would survive that experience ...
you have just described my house (roughly converted 200 year old small stable). Also burning wood isn't an issue. The problem with fossil fuels like coal is the fact that it dumps what was stored solar energy from several thousand years ago back into the atmosphere. most wood will only be putting some back from that last - 50 years or so I would think. then if you stop I think a little more about it, by using electricity you have done is move the fact you dumping the same thing into the atmosphere out of the chimney of a power station - that the problem with electric cars, they just move the problem to somewhere else where its easier to ignore.

I burn wood that I collect through out the previous year, even then I only heat one room once or maybe twice a week (for a couple of hours) in winter and not all between April and September.

Gas can be put into bottles and put into cars, meaning that natural gas heating is easily achievable "Off Grid". while you can't have Calor deliver it, its quite common in my area to have bottled gas instead of oil as its cheaper and constraints on tanks and a likes don't really exist.

I'm a massive fan of, put a jumper on instead of the heating. And I know a lot of people who basically live under electric blankets - they only what needs to be heated and as a result are a lot cheaper to run. I find myself going to some peoples houses and find them overly hot with thermostats set to 21*C.
 

PhilHornby

Senior Member
Initial experiment with SSR

It may be possible to reduce the time that the motor is energised for (5 cycles on, 10 cycles off) which will give full torque but may once again reduce the speed that the motor runs at due to the air resistance on the fan.
I just tried an initial experiment, using my cheapo fan heater and an SSR. I connected the SSR to PinC.2 on an 08M2 and knocked together some code, that varies the ON/OFF duty cycle over a period of 100mS. This equates to 5 mains cycles, but there's nothing in my code or hardware to detect the zero-crossing, so I'm out of synch with the SSR. Time periods much greater than 150mS made the fan audibly "woosh" and then free-wheel. A curious, constant-speed knocking noise can be heard at certain settings.

It definitely achieves speed-control, though I have no idea of the relationship between speed and ON/OFF periods. It doesn't seem linear. Earlier, I ascertained that the fan normally revolves at about 3000rpm (using a phototransistor and my scope), thus making it a 2 pole motor.

Much more work to do!

Rich (BB code):
#picaxe 08m2
#no_data
#terminal 38400
symbol SSR        = C.2
symbol TimeFrame  = 100 ;mS (5 cycles of mains)
symbol HoldTime   = 50   ;maintain speed for this no. of cycles ~ 5 Secs
symbol HighPause  = W10
symbol LowPause   = W11
symbol Speed      = b19
      setfreq M32
do
      ;
      ; Accelerate fan
      ;
      b1 = TimeFrame - 10
      for Speed = 0 to b1 step 10
            gosub Control
      next
      ;
      ; Decelerate fan
      ;
      for Speed = TimeFrame to 10 step -10
            gosub Control
      next
loop
Control:
      HighPause = Speed
      LowPause = TimeFrame - HighPause
     
      sertxd (cr,lf,#HighPause,",",#LowPause)
     
      HighPause = HighPause * 8
      LowPause = LowPause * 8
      for b0 = 1 to HoldTime
            if HighPause <> 0 then
                  high SSR
                  pause HighPause
            endif
           
            if LowPause <> 0 then 
                  low SSR
                  pause LowPause
            endif
      next
      return
[video=youtube_share;M3D8H3qNudg]
 
Last edited:

PhilHornby

Senior Member
A Promising Start...

I re-assembled the 'cheapo' fan heater, connected it to an SSR, added a little 'instrumentation' and then using the following code on a Picaxe 08M2 :-
Rich (BB code):
#picaxe 08m2
symbol SSR        = C.2
      setfreq M32
      do
            High SSR
            Pause 480               ;480 = 60mS - 3 mains cycles
           
            low SSR
            Pause 480
      loop
observed the results...
Aside from some strange noises from the SSR, (and some flickering lamp fittings!), things went reasonably well.

I measured the fan's rotational speed by shining an LED headset through the fan blades and displaying the chopped up waveform from a phototransistor on my 'scope. I measured the temperature of the air emanating from the front grill of the heater using a DS18B20. Sound level was determined using a RadioShack Sound Level Meter (set to response "C"). I measured AC current using an (analogue) Clamp-type Ammeter and the graph was produced using the sensing coil from a power meter, fed into my 'scope.

Mains frequency at the time of the experiment was 49.98Hz (according to this site) and I measured the Mains Voltage as 239VAC.

STATESound levelFan RPMOutlet temp.Measured currentPower
Heat #183dB spl259554°C3.5A836W
Picaxe+Heat #174dB spl171046°C1.75A418W
Heat #283dB spl262575°C6.8A1625W
Picaxe+Heat #274dB spl171567°C4A956W
On the face of it, I achieved what I set out to do - lower the heat and sound output of a Fan Heater; but are quite a few shortcomings to overcome:-

  1. Lowering the speed of the fan in this way introduces some new sounds - in the form of 'knocking' noises. They're not obtrusive, but make the motor sound like it's not happy in its work!
  2. The measured current wasn't stable. My (analogue) clamp Ammeter was flickering in time with the burst firing, but every few seconds it flipped to a different reading. So, where the table says 1.75A, it should really say 1.75A some of the time and 2A the rest!
  3. On fan only, the SSR makes clicking noises!
  4. Controlling the heater power using this 'burst-firing' method introduced 'flickering' in lighting in other rooms of the house. It turns out they weren't restricted to just the Ring Main the heater was on ... so they may not have been confined to my house :eek:

Observing the current flow to the heater, in the Picaxe+Heat configurations shows an odd effect :-

SDS00006.png


  • Why does current continue to flow when the SSR has been turned OFF?
  • Why is the amplitude of the first half-cycle after SSR triggering, lower than the following five?

Presumably, the substantial amount of current flowing after SSR turn off, is related to the burst firing of the heating elements; on Fan only I can't detect any measurable current at all, using this method.

One other thing to mention: I compute the theoretical speed of the Fan motor to be 2998.8rpm @ 49.98Hz. The measured value of 2595rpm equates to 13% slip.
 
Last edited:

PhilHornby

Senior Member
Interference-free heater control...

Would a scheme like this work? i.e. control the Heater Power, without generating RFI or unpleasant lamp flickering...it would only give four discrete settings, but that may be enough.

HeaterControl.png

The Fan Heater has two identical 1KW heating elements, as most fan heaters seem to. They could be turned on fully using a TRIAC, or on half-power using an SCR.

The combinations would therefore be:-

"A" = 500W (Heater 1 activated by SCR)
"B" = 1000W (Heater 1 activated by TRIAC)
"C" + "A" = 1500W (Heater 2 activated by TRIAC and Heater 1 activated by SCR)
"B" + "C" = 2000W (Heater 1 and Heater 2 both activated by TRIAC)

I was wondering about running them in series, to give 250W as well ... but obviously that gets more complicated.
 

newplumber

Senior Member
Hi PhilHornby I am learning from your testing which is cool
just one simple question where would the fan be placed in your last schematic on A-C pins?
maybe you have it explained and i am blind
I ripped apart a few heaters and thinking the fan is connected after the elements but can not remember
and it shouldn't matter with AC but just curious
thanks
Mark
 

PhilHornby

Senior Member
just one simple question where would the fan be placed in your last schematic on A-C pins?
Don't worry, you didn't miss anything - that schematic/doodle was purely for heater control ... the fan speed being controlled separately to match (by some other means). I'd not expected heater control to be an issue at all ... but it seems like it could be.
 

newplumber

Senior Member
Well cool ...first time i didn't miss anything :)
but I have to say ..from my view you have got a complicated project
I made a digital space heater using a lm34 and a pot and running the heater/fan with a cheap ssr from ebay
when I tried it out ...the ssr was getting hotter then the element of course before it decided to choke
so i just used a 20amp 5volt mech relay instead but it means nothing to help your situation
 

tmfkam

Senior Member
As a fan of simplicity...

Why not place a series diode in each of the heater element supply feeds, these could then be 'bridged' with a pair of relays switched by the PicAxe? May be simpler to implement but with almost as much control.
 

PhilHornby

Senior Member
As a fan of simplicity...
Very droll :)

him as well said:
Why not place a series diode in each of the heater element supply feeds, these could then be 'bridged' with a pair of relays switched by the PicAxe? May be simpler to implement but with almost as much control.
They're going to 'click' aren't they? ... at random intervals ;)

Any views on that current flow graph?
 
Top