Author Options:

Carburettor Heater - need help Answered

Hi All,

Hope this is in the right section.

I am working on a project to heat the body of a carburettor (Bing 84) in order to prevent carb icing.  No commercial solution exists for my carburettor.  My engine (for paramotoring) is air cooled so can't use water from cooling system.

My idea was to strip out the nichrome heating element from a car cigarette lighter and use that.  The heat needs to be applied to the metal casing of the carburettor, preferably being hottest near the air intake side of the carb.  I decided the best place was where the air filter is attached - a 12mm long metal throat with a 35mm diameter The trouble is, I don't know I can transfer the hundreds of degrees of temperature from the nichrome to the metal body of the carb without just shorting out the battery.  If I am to have a layer of insulation between the nichrome and the carburettor, what can I use that will conduct heat but not electrical current, and withstand around 300 degrees of heat?  I also thought about reducing the current and spreading the heating over a larger portion of the metal casing, which would allow the temperature to be lower.

I have also been experimenting with motorbike hand grip heaters that run of 12v but they don't generate enough heat - I need to be pumping in around 8-10A as the internal cooling effect of the fuel vaporising rapidly causes sub zero conditions inside the carb even when the outside air temperature is well above freezing.  I have a 7Ah li-ion which I would only switch on when required (maybe have a latching push button on throttle control).

It may not be doable but its an interesting project nonetheless and something I am going to play around with.  Any ideas?

Thanks, Dug


Buy some insulated ni-chrome wire, and wind that directly on your carb body.


Hi Steve, you can get that stuff with insulation on it? Perfect, thanks!

Wrapping it on directly will save me all sorts of problems.

How much power do you really need ?

Why not draw the incoming air over the exhaust manifold, and pre-heat it ?


Thanks for the suggestions. There are a couple of reasons I would like to avoid joining two parts of the engine or adding what would have to be a fairly large scoop to try to collect warmer air:

The exhaust is on rubber mounts due to the amount of vibration going on. Although the prop is well balanced, it is a single cylinder two stroke and vibration is unavoidable. Even the mounts themselves can break away over time. If I had any sort of connection between exhaust and carb it would eventually break, and the pieces would go into my prop and cause all sorts of in-flight problems.

The black air intake silencer that you can see attached in the second set of pictures actually works very well in reducing noise in flight and I would miss it if it went to make way for something that could sit between the prop and the exhaust. The only place I could collect warm air (due to direction of air flow) is between the cylinder head and the prop, or between the exhaust and he prop (very little space). To get a pipe down to the cylinder head area would be tricky, and I am not sure how it would affect performance.

The other point to consider is that warm air going into the engine will lower performance. Perhaps not by a large margin but every hp counts when you are trying to take off. A solution that I can switch on and off would be best, or an always on solution that heats the carb body.

Its frustrating having so much heat in the exhaust and cylinder head and not being able to access it but I'd be nervous trying to attach anything to fancy to the engine because in flying, everything that can go wrong will go wrong. I appreciate that collecting engine heat if possible would make a lot more sense than electrical heating but I can't see a way to do it that is within my skill set.

What's you maximum flight time ? If we knew that we could work out power availablity better.

The colder the air at the inlet the more efficient the combustion process, so you really only want to just stop freezing/


Hi Steve,

I can stay up for 3 hours maximum but realistically would only stay up for 2 and most flights probably average 60-90 minutes.

I am still unsure whether it would be better to have an hour of constant heat or say 20 minutes of intense heat. There are arguments for both I suppose. The humidity is generally quite patchy up there and it is usually the case that if you can get through 5-10 minutes of icing it will go away by itself (as you have flown through the patch of humid air). The times when I have had it I really could have used some instant and very intense heat for around 10 mins. 2 heating modes would be good - perhaps two circuits with differing resistance?

Batteries are pretty expensive and I am sure that my motor's stator coil assembly puts out a current (although it is unregulated). I will have to do some more research and check exactly how much it puts out and how to fit a regulator. Perhaps I could have it providing constant heating and then engage a a battery for the more intense heating as required...

I'd make it thermostatically controlled for a start - no heat, unless its actually cold. I'd like to experiment to judge how much heat you actually NEED to do this reliably. I'm not sure pulling energy from the magneto is a good idea though - they ain't wound for current.

Add a small alternator ? Say off a motorbike, or moped ?


It seems it actually has a three phase alternator already - it just isn't currently used for anything and the wires are taped up. I would need to add a regulator/rectifier and then test whether I get the required amps.

Thermostatically controlled is a good idea except that inside the carb it is *always* a constantly freezing cold temperature - the question is whether the air is humid enough for condensation and then ice to form (which I guess would theoretically increase temperature fractionally). I don't suppose there is any way to reliably detect humidity? That really would be cool :)

Nope, you wouldn't need a rectifier - AC heats just as well as DC.

How much current can you pull out of it ? What voltage ? Wiring a three phase load for it would be no problem.

You CAN measure humidity/ dew point very easily - there are off the shelf sensors for it

I am told the alternator spits out 20-40 volts depending on rpm. I think that the current might be an issue however as someone on a paramotor forum mentioned that around 2.5A is the maximum to expect (although that is assuming a rectifier/regulator is attached).

Perhaps adding a small battery into the mix would allow me to use a lot more amps for short periods, and it would slowly recharge when not in use.

Thanks for the heads up on the humidity sensors - I found this one http://www.mindsetsonline.co.uk/product_info.php?products_id=267

I think it would be best if the humidity sensor ran on its own circuit and just lit up an LED on my throttle control when humidity was looking high but I retained control on when the heat is actually applied. Or I could have a three stage rocker switch - on-off-auto.

If I am to use a battery I would presumably have to get a rectifier/regulator unit. Do they restrict current at all? Is it correct to say that whilst the volts generated by the alternator increase with rpm, the amps are fairly constant?

You only get "amps" if there is a load to take them ! Your power output would depend on revs, but if we say you can get between 40 (20 x 2 ) W and 80 (40 x 2) W - you should be able to do this entirely with what you already have.

Find yourself some THIN KAPTON tape, to wrap on the body before the wire, and add something to monitor the body temperature.

Are you UK based too ?


Hi Steve, thanks for the info, yes I am UK based! I agree that an external solution would be a lot safer... :)

I found these things: http://uk.mouser.com/Search/Refine.aspx?Keyword=wh50

Which look pretty robust. They might not transfer heat onto the body quite as well as directly wrapping nichrome but with some thermal paste they would be a fairly simple (and cheap) solution for the heating part (and less chance of me setting something on fire).

I wouldn't worry about wrapping wire on, its just a question of getting the right resistance - metalclad resistors are going to get too hot, unless they are exceptionally well coupled to the carb. The wrapped wire is bound to be in much more intimate contact with the metal. I WOULD add a thermostat though....


The resistors are cheap enough that I can get a couple and do some testing. Perhaps I could take thin strip of copper and use it to wrap the resistor onto the exterior of the piston housing to help apply the heat to the carb. I don't think it will overheat once in use due to the volume of air blasting over it and the cooling effect in the carb iteself. I expect it will raise the temp of the carb body by around 5 degrees or so which should be enough.

A thermostat would be a nice addition in case I accidentally leave it switched on when the engine is not running.

A little indicator LED or something similar would help in the "leaving it on" dept. :-)

Thanks Goodhart - the more LEDs the better in my book! I was hoping to have an illuminated 12v rocker switch to insert into the end of my throttle handle, except the diameter of hole I have to work with is around 14-15mm. I can't find any illuminated latching switches at that size - even at 16mm they are hard to come by. I might have to have a separate switch and LED on the throttle handle (I have 2 spare locations for 14mm buttons or LEDs so it should be possible).

I was thinking some of the same things. One certainly doesn't want the wire to wear through from movement, etc, and contact the carb body. Shorts in that area would be dangerous at best :-)

I wouldn't mess with the inside of a carb, ever.

I would add some insulation OUTside my heater though.


I have done some research and found out that commercial carb heaters (particular to specific carbs and not available for mine) use anything from 12w to 60w. Given that 60W is the most that seems to be used, I think that 5A of power use should be sufficient for as-and-when heating.

The other question is how to apply it most effectively to the carburettor. I have attached an exploded view diagram of my Bing 84. The simplest way would be wrapping it around the inlet throat. In earlier images you can see that the ice forms mostly on the piston (item 8) and needle (item 4). I think that although fiddly , it may be possible to insert a heating element inside the piston itself around where the spring sits. When I get my motor back next week (it is currently in another country!) I will be able to get a better idea on where to fit it.


Not being overly familiar with autos nor motorcycles, I am not sure which this is, but in automobiles, icing "normally" occurs when the beast has been garaged, and so condensate forms in the lines and carborator. A bit of a drive in the cold and it freezes. However, if you aren't having any fozen line problems, I am guessing my "just provided info" is of no use to you :-)

Hi Goodhart, the icing actually forms due to the cooling effect of the petrol vaporising and movement of air inside the carb. It can be completely dry before you take off and then if you fly through some humid air and the water from the air sticks to the internal components within the carb (needle and jet) and then freezes. This can happen when its a warm day outside.

The result is loss of power and can lead to an engine outage if the ice is allowed to build up.

The simple solution is to add 2% Silkolene Pro-FST to your fuel, but where is the fun in that?

Chuckle......ok, I hadn't read up on that one....I just kind of figured as best as I could from what I knew :-) Thanks for bringing me up to speed. I love to learn new things.

You could try tapping some of the hot exhaust to heat the incomming air a little, perhaps with a double walled pipe on the air intake. If they are close together, it may be more practical to re-route the exhaust so it passes in front of the carb, warming the air flowing over it.

It is definitely the best way to go, you got hot-gas; use it.


...as a ferinstance

What if you build a wire coil out of transformer wire and shape it, so as it lays flat on the inside of the carb intake. you could drive it with a simple current source so it is hot but not breaking down and it would be switchable.

Hi Skinners,

That would work nicely except that the hot part of the exhaust (where it first leaves the cylinder head) and the outlet is on the other side of the engine :( The other thing I am up against is the fact that it is a 200cc 2 stroke and everything vibrates - so unless things are bolted together right they can shake themselves apart. Connecting the exhaust and carb together would be tricky. The air flows in such a way that to get warm airflow the hot object would have to be between the carb and my back rest. I don't think that the warm air flow would be enough unless to overcome the icing unless the object behind the carb was extremely hot.

The cylinder head is hot, and a bit closer but again I am not sure how I would collect heat from it. I thought about copper pipes but don't think I could attach them in such a way that they wouldn't eventually snap off with all the vibration...

I have attached a couple more images.

I have disassembled the cigarette lighter and done some testing - it draws around 6A. I did some reading on soldering irons and read that they use mica sheeting to insulate the heating element from the metal soldering shaft and I thin that may be the material I need.


2011-11-08 20.34.44.jpg20111120_205943.jpg

From that first picture, it looks as though the insulated part of the exhaust passes close to the carb, but as you say the current routing does not lend itself to easy modification to heat the air intake. However, you could use this to your advantage by constructing an aluminium air-box around the carb and part of the exhaust, so the whole carb is kept in a warm environment.

You may be able to use the element(s) out of a 12V soldering iron, but I expect that heating of the incomming airflow will be fairly ineffective. If you've got the time, it may be more effective to heat an oil filled jacket around the carb. This approach would probably be more effective using heat pipes from the head as you suggested rather than electric heaters.