Protection from spikes

[#1]

Hopefully someone understands electronics better than I do !  You usually do.
I have a digital timer which controls my boiler - it is not o.e.m.  However, the boiler has its own dedicated power circuit which is controlled by a simple interface relay on the 240 v. house distribution board.  From time to time, the interface relay blows the memory on the digital time-clock, and both the time and the programmes have to be re-set.
I have determined that it is the interface relay which is the cause by disconnecting the boiler from the dedicated circuit, and plugging it into the 13 amp ring main - no further problem.  The reason for the need for the interface relay is that I can turn the boiler on and off remotely using a dtmf signal which activates the interface relay.  I imagine that the coil on the relay causes some sort of surge or spike which sometimes upsets the digital clock.
Is there any way of cleaning the line to stop this happening by fitting a diode or something like that between the contacts of the relay ?  I am not technical and only make this suggestion as a result of "Googling" !
The only other thing that goes through my mind is that I notice that the electrician fitted a 20 amp interface relay (about 10 years ago).  The boiler specification calls for 3 amp protection only.  So it crossed my mind that a 20 amp interface relay might be overdoing it and causing a bigger surge / spike than is necessary.  Is it worth fitting an interface relay with a smaller switching capacity ?  Or is that unlikely to make any difference ?
All suggestions gratefully received.

[#2]

But any change you make would need to be on the boiler or mains side of things?
So it would be a job for a qualified electrician?

[#3]

I have asked a qualified electrician and he says that electronics is not his field and therefore cannot advise.  He said that the the digital time clock is electronic, which it is, and not his area of expertise.  However, if a diode had to be fitted or the relay needed to be changed, he would be happy to do it because it is a five minute job.
So it is not a question of who does the job but rather what job has to be done.
This article seems to address the problem :
http://www.google.com/url?sa=t&rct=j&q=diode%20suppressor&source=web&cd=2&ved=0CE4QFjAB&url=http%3A%...

[#4]

Hi,
First, check that the relay coil is low voltage (less than 25 volts).
If it is, then that document looks like what is required - just solder a diode across the coil terminals of the relay, ensuring that the diode cathode is on the positive side of the coil.  It might be worth asking a local television engineer.
Richard.

[#5]

Richard
Thanks for that. The answer is that the coil is 240 volt as far as I can tell.  There is nothing on the relay to indicate otherwise and the coil terminals connect directly to the a 5 amp fuse in the distribution box.
The circuit is quite simple :  240 volt DTMF signaller  receives a signal from the telephone, and makes a 240 volt connection (max 5 amp) to the coil of the interface relay which switches on the 240 volt circuit to the boiler.  The digital time-clock interupts the circuit in order to control the on / off time of the boiler.  It is all quite simple.
At first sight, it may appear that the 20 amp relay seems to be redundant because the 5 amp relay inside the DTMF signaller is large enough to switch the boiler circuit (drawing only 3 amp) on and off.  In fact the DTMF signaller is on its own circuit protected by a 5 amp fuse.  The boiler circuit is on a separate circuit controlled by a 15 amp fuse, even though it is only drawing 3 amp.  So the interface between the DTMF signaller and the boiler circuit is only through the relay, and the relay is causing the spike which wipes the memory of the digital time-clock.  The easy answer would be to replace the digital time-clock with an analogue time-clock, but the problem then would be that the analogue time-clock would stop when the power to the circuit is removed by the DTMF signaller.  A digital time-clock has a built in battery for the memory, of course.
I have been in touch with Farnell the electronic component distributors, and they suggested the following diode:
http://uk.farnell.com/fairchild-semiconductor/1v5ke160ca/diode-tvs-160v-1500w/dp/1017605?Ntt=1017605
But I am not confident enough to have one installed without a second opinion.  Do you think this is the right diode for the job, because I do not understand the specification ?

[#6]

Hi,
That diode is definitely NOT suitable for connection across a 240V supply.  Neither should a diode be connected across the relay coil if an AC supply is involved.  Unfortunately, without seeing what is there and testing with a meter,  Apologies, but I don't want to advise further - especially as it appears mains voltages are involved.
Richard.

[#7]

I'd suggest a 'MOV' device (metal oxide varistor) http://en.wikipedia.org/wiki/Varistor, commonly used to suppress 'spikes' on AC circuits, particularly across relay coils.

[#8]

How about connecting a resistor in series with the relay coil?
It should be of "moderate" size and able to work at 240V.
It will need to still provide sufficient current to operate the relay, though.
Current = V^2/R and all that...
Should prevent any mighty surges...