Alternative for a relay or switch, switching a 12V 23A motor?

[PloddingDonkey]

Hello.
Sorry for not doing my homework but time is a bit pressing.
A friend has a 12V tire pump/compressor, it has burnt out its switch. This is unsurprising as the switch was rated 10A and the motor has a stated draw of 23A, it must be a brute!
Though the pump will not be permanently connected to a battery its case also has a switched light/lamp, so it would be nice to retain the ability to switch the motor on and off rather than simply soldering the motor leads to the leads which connect the pump to the battery.
I volunteered to fix this thinking of using either a 30A relay or a 30A switch but it turns out there is no room for a relay inside the pump's casing and a switch would protrude and be 'inelegant'.
I am now thinking of using a 'transistor' as switch as I recollect this was a possible usage but the bit of surfing I have been able to do suggests I need a mosfet, what the difference between the two is I do not know and is perhaps irrelevant for a one time solution.
The original wiring had the switch on the +ve side of the motor and I would prefer to stick with this layout if possible.
I assume I would need to include some inductive voltage spike protection for when the motor is shut off.
Could some one suggest a circuit and appropriate components.
Thanks and once again I apologise for my lack of research.

 

[Minder]

You could also look at using a IGBT, most do not include the BEMF protection so a diode would be needed.
How difficult would it be to switch the neg side?
M.

 

[PloddingDonkey]

Thanks for the reply.
I imagine switching the negative side is feasible but at the moment I can not check as the case has decided to play Hide & Seek in my workshop. Switching positive is really just habit from my biking days.

 

[Minder]

IGBT-Mosfet?
M.

 

[duke37]

An FET needs to be switched fully on or fully off to dissipate little power. If the power supply is a little bit low, then the FET will disipate lots of power and burn out unless you have a heat sink much larger than a relay.
To switch the positive side you will need a p-channel FET, these are not as good as n-channel FETs.
A BJT transistor will drop about a volt, at 20A this will give 20W of heat to get rid of.

Are you sure you cannot get a 12V automotive relay in, they are quite small.
A fat diode across the motor is necessary whatever switching device you use.

 

[PloddingDonkey]

Ta Minder, I guess I start reading, which I really should have done in the first place.
I will do that but have been told " 'elegance' doesn't matter, use a switch", LOL.
Is a diode across the motor still a good idea? I assume it would stop some arcing in the switch when switching off
Duke, there is no room in the case for what I assume is a normal 12v automotive relay.
I have, in the past, seen and used smaller relays but, from memory, they were switching small currents, this thing has the biggest draw outside a starter motor that I have encountered, it makes me want to check what mine draws off the cigarette lighter circuit.

 

[duke37]

The small automotive relays that I have are 15*23*25mm with spade connections out of the end 22mm long. These could be chopped down if the wires were soldered.

 

[PloddingDonkey]

Update, phhhhtttttt, I tested the thing last night.
23A my beep beep! unless the motor is duff!
It draws just under 3A with no load and around 6A when the airline is completely closed off.
I had the lamp portion of the pump switched on and in the latter situation its light noticeably pulsed in time with the motor sounds.
The current was measured with an Avo model 12 which might, I suppose, might have averaged out the pulsating nature of the current draw in the latter situation. I am not going to pass this through a DMM just in case there is a transient 12A + current.
I am going to take some voltage measurements across various bits later today as the discrepancy between the stated draw and the measured draw seems a bit extreme.