Homemade 4 Stroke, Opposed Twin Glow Plug Engine - Why Won't It Start?

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.


Dec 28, 2013
Reaction score

I hope somebody here can help me.

I am new to model engines (although my day job is developing car engines) and particularly glow plug engines.

A few years ago, I bought some plans for a 4-stroke, opposed-twin, glow-plug engine from VTH in Germany (see attached picture). Earlier this year, I finally got all the machining finished and the engine assembled (I was having to fit it in around my job and other commitments). I then couldn't get the thing to start: it would cough a bit but not actually run. After some investigation, I came to the conclusion that one cylinder was breathing better than the other so decided to try running it on just one cylinder. I removed the other cylinder and head completely (so that I had, effectively, a single cylinder engine running in a twin's crankcase and on a twin's crankshaft). I finally got it running on one cylinder, using the head from cylinder one and the barrel, piston, rod and cams from cylinder two.

Upon further investigation, I came to the conclusion that the valve seats in the cylinder heads were much too far recessed into the heads, so that the valves were being shrouded when they opened and were not allowing sufficient air into and out of the cylinders (this was due to a mistake I made when fitting the valve seats). This wasn't as bad on cylinder one as it was on cylinder two (hence using the head from cylinder one when I actually got the single cylinder engine running). I also concluded that the cam profiles were probably better on cylinder two than cylinder one (due to manufacturing errors when making the cams) (hence using the cams from cylinder two when I actually got the single cylinder engine running).

I therefore decided to remake the cylinder heads with less recessed valve seats. I have done this and, apart from the valve seats being less recessed, I am pretty confident that the new heads are identical to the old ones. Nothing else has changed (the cams are exactly as they were before) but I still can't get the engine running; and I now can't even get it running on only one cylinder, as I did before.

I've got it set up as single cylinder engine, and it seems to be breathing ok (when I hold my hand over the exhaust during cranking, I can feel air pumping out and, holding my hand over the inlet, I can feel air being sucked in). I've also done a compression check and that seems to be OK. The glow plug glows when connected (I've even checked this by removing the cylinder head with the glow plug in it and connecting it up). The only other difference, apart from the new cylinder heads, is that it is now winter and probably about 5ºC in my garage (about 41ºF). If it was a spark plug engine, I would be checking that I had spark and also checking the spark timing but glow plugs don't have a timing, I suppose. I haven't got around to trying to set the carb up, yet, so it could be fuelling poorly, but I have tried cranking it while squirting the fuel into the inlet using a perfume sprayer (the fuel is a model engine fuel which is, I think, a methanol, nitro, oil mix). It should, at least, cough, while doing this; and, last time, back in the summer, it did run with the carb I am using, so I would hope that the fuelling is approximately ok.

I am thinking of trying two further things: a) using EasyStart, as you use for getting car engines started, when they don't want to and b) warming up the whole engine with a blow torch before cranking it and spraying the fuel into the inlet as before.

Has anybody got any suggestions, at all, as to what is going on and why I can't get the thing started, even on the one cylinder? Or any ideas as to what to do next? I am so unfamiliar with glow plug engines that I am a bit lost. The only knowledge I have of glow plugs is of the ones you use in diesel engines to aid cold starting.

Thank you for any help, advice tips or even just comments that you can offer.


  • Engine_Picture.pdf
    101.6 KB · Views: 543
A photograph of the heads may help.

Glow plug engines work because the plug wire is platinum, which catalyzes the methanol-oxygen reaction. It doesn't always work without sufficient heat, which is why you need to apply power to the plug (unless you have a really high-compression engine -- one of my control line stunt engines will occasionally start without a battery, on a hot day).

Really, the possibilities that I can think of, kind of in order, are:
  • Not enough compression ratio for the amount of nitro
  • Poor compression (as in -- it leaks, either in the valves, the piston/cylinder seal, or the head/cylinder seal).
    • If it gives you plenty of resistance when you flip it, and "pops" over TDC nicely, this probably isn't it. Do this test with a prime so there's oil coating the cylinder.
    • If you can take it to TDC, pause, and it sort of pops out from compressed gasses, this definitely isn't the problem.
    • An engine that's a bit leaky will still do OK at full throttle, it'll just be overly sensitive to the needle and have a tendency to flame out -- and it really won't like idling.
  • Incorrect mixture
  • Plug recessed into the head (it needs to be level with the head, otherwise it just lives in a little pocket of exhaust gas and can never get fresh stuff to catalyze a reaction with)
  • Still too tight, but I think you'd have mentioned this.
  • Bad plugs (unlikely, but plugs can get coated with ash or fuel contaminants and stop catalyzing, even when they're nice and hot).
  • Poor pumping -- I only put this last because you said you checked already, but it doesn't hurt to check again.
I've only built one glow engine, but I've resurrected quite a few, and converted quite a few from RC to control line. A good procedure when you're trouble-shooting a new-to-you engine is to make sure it has a fresh plug, disconnect the fuel feed, give it a way over-generous prime (but not to the point of hydraulic lock), apply battery, and start flipping. It'll clear excess fuel as you flip, and at some point should take itself through a mixture that's good enough to start. Then you should get a little burst while it runs out the prime. At this point you know the engine's basically good, and you can worry about the induction system.

If the engine was designed for an American audience, it's probably intended to have at least some nitromethane in the fuel mix. Probably somewhere between 5 and 15 percent. More nitro means less compression is necessary (in fact, more nitro means too much compression is inappropriate). If you're running straight methanol-oil, you may want to check that. Ditto if you just plain got the compression ratio wrong.

You don't want to use engine starting fluid for gasoline or true diesel engines. The right stuff to use to give a glow engine a bit of help is straight naptha -- old style cigarette lighter fluid works, as does (in the US) Coleman lantern fuel. Just a few drops in the venturi makes magic for a balky engine or on a cold winter's day. If the engine starts on that but won't run, you can try doctoring up your fuel with no more than 2 ounces per gallon of the stuff -- it generally aids ignition, makes needle settings easier, and makes starting easier.
Valves. When I made the valves & cages for my radial (80% done so not running) I learned there are a multitude of ways to make good looking valve components that leak. A good test is by simple hand vacuum pump. Terry Mayhugh has shown this method on many engines, most recent here.
https://www.homemodelenginemachinist.com/threads/another-knucklehead-build.27584/page-2If you can make some sort of adapter for your head & pull vacuum with valve closed (and the valve stem / valve guide annulus temporarily sealed off with silicone tubing or something). If you see the needle draw down quickly you know you have a leak.

Timing. Assume you have checked this independently vs TDC but is it possible gears or cams are are out of phase? Proper tappet/valve clearance gap when cold?

Compression. I haven't done much pressure testing myself. I hear they are kind of subjective. But maybe you could rig up a maximum reading compression gauge to the glowplug threads, remove pushrods so valves closed under spring pressure & turn over (with lubrication). That might tell you relative values between cylinders & heads. Or twitchy needle might indicate head flange seal. Usually when glow engines run and then don't, something is varying with temperature. Metal is expanding but somehow not in a good way.

I don't think moderately recessed valves are an issue from a flow perspective. Many commercial engines have different valve layouts. But where it would impact potentially is compression ratio. Glow engines are quite particular about that. Assume you built to the plans & designer had this figured out. But if you did any modifications, it takes very little dimensional change to make a big CR difference. The usual rule of thumb for CR is 7-9 to 1. The easiest way to verify is fill the volume with methanol at TDC & BDC with syringe & calculate. You have to kind of normalize to the level the glow plug reaches.

Other things.
- Simple leaks can cause engine fits. If the carb is not sealed to the inlet manifold hole the air/fuel mixture can get messed up so it may run at one rpm but not another or at one temperature but not another. Usually requires O-ring between carb. Spray a bit of soapy water around area, close venturi with finger, fast flips, should be no bubbles.
- if the carb is over/undersized it can cause running issues. To begin, have the tank somewhat close & somewhat level to carb inlet. It should draw fine but look for air bubbles moving along when cranking
- if your fuel is old or has been stored in low temperature or humid conditions it can go kind of stale. The methanol attracts moisture. If fuel is good, consider adding nitromethane as recommended.

I would not warm the engine especially with anything like a torch. We used to race in winter to -20C (self imposed finger freezing makes for bad flying haha). These were 2-strokes mind you but nothing special in terms of fuel, glow plug voltage etc. Once they pop the engines come to life and dont know the difference between winter & summer. Some guys would shoot WD-40 or starting ether? over the carb just starting but that was usually for ice. Its hard on glowplug elements.

Good luck!
Thank you very much for your replies.

I shall get some photographs to post on here.

When I compression checked it, I got a figure of 25psi (spinning the engine with an electric drill). That would be low for a car engine but I don't know if it is low for a model, glow plug engine. The fuel I'm using is, I think, 15% nitro (or it might be 25%; I'll double check). When I measured the compression with the old heads, as best I could, it came out at around 7:1, whereas the designed for compression ratio should be 8.5:1. Do you think this could be the issue. I should be able to skim the heads to increase the compression ration.

Thank you, again, for such helpful and detailed replies. I must admit that I am considering buying a spark ignition kit for the engine, although I do so hate to give up on things.
I do not think spinning the engine with a drill method gives you the correct CR unless you have a check valve.
Also consider that any volume added in the plumbing between head and pressure gauge reduce the compression ratio reading considerably, when dealing with model engine small displacement. Not much a factor on a full size engine.
Tim gave you good pointers about compression testing. If it "bounce" nicely is good.
Sometimes a new engine needs to be loosen up by driving it well lubricated. I use my late to drive it either directly or with some pulley.
You do not say how you start it but flipping the propeller get old soon. Rig up a method to start it with a drill.
Don't give up, starting fits are not unheard especially on the first engine you build.
Nitromethane is never required. It does make more power and makes carb needles less critical.

7:1 will run on methanol without nitromethane. Glow engines will run sloppy rich, like raw fuel pouring out the exhausts rich. Significant prime as long as it's not completely flooded, will be fine. A very rich carb setting should be fine as long as the plugs are getting enough current.

It sounds like there is no carb? You won't get but one or two combustion events on prime in a four stroke.

To see if it will fire, I do this. Prime, flip the prop once or twice, energize plug, firmly hold the prop (like your life depends on it) and turn it through a compression stroke. If compression, prime, and plug are working, it will fire and you'll feel a bump in the prop. I'm not trying to start it, I want to see if it will ignite. If I observe combustion, I'll try flipping to start, but it may need priming again at this point.

What is supplying current to the glow plugs? Dull red may not start an engine. What's on the crankshaft, propeller, flyhweel? How are you cranking? Is the valve timing correct?
Thank you all for your very, very helpful replies. I am now focussing on running it as a single cylinder engine until I can get it running.

I gave the engine a good going over today, checking dimensions, static compression ratios and valve timings etc. and I have added a couple of spreadsheets showing these, along with a diagram showing what I meant by a couple of the dimensions. I have also added a couple of photos; on the photo showing two heads, the head on the right is the new head and the one on the left is the old one with the recessed seats. The picture of a single head is the new head, showing how the glow plug is located.

In answer to dieselpilot's question about carbs, I am using a Perry "pumped carb". I'm not sure about this carb, to be honest, and plan to do some reading about setting it up properly in the next few days. However, when the engine ran last time, with the old heads, it was using this carb, so I think it should be ok.

As well as all the checks, today, I heated the whole engine up to about 70ºC, using a blow torch, squirted some fuel (which looks to be 20% nitro) into the inlet port and cranked it and the thing did cough and splutter a bit, so I am now beginning to suspect that the cold ambient conditions are a significant contributing factor. I, once again, cranked it while holding my hands over both inlet and exhaust ports and it does seem to be breathing / pumping quite well. Just out of interest, I squirted a bit of fuel onto my steel welding table, which was really very cold to the touch, and held a blowtorch to it. It tried to ignite but wasn't very enthusiastic and went out very quickly, while there still appeared to be liquid fuel present. I squirted some fuel onto some blue roll and it ignited and burned more happily. Perhaps being in contact with cold metal sucks energy out of the burning fuel and makes maintaining combustion less robust. Also, I do leave the fuel out in my garage and have had it for about 6 months. I know that alcohol fuels tend to be hygroscopic so maybe it has picked up some water. These thoughts are just speculation, of course.

When I turn the engine over by hand, I don't notice any particular resistance at TDC, any greater than the general friction turning the engine over, and certainly no "pop" or feeling of air being compressed and trying to spring back. I suspect that the combustion chamber may not be holding compression that well; maybe poor ring sealing.

My next steps are to try to leak check the cylinder head somehow (perhaps using compressed air or just the fuel) and also to leak check the piston-ring-cylinder assembly (again, perhaps using compressed air or just fuel). I also plan to recut the piston ring grooves, as I noticed that the rings do not really move freely in them and fear this could be having an effect on sealing and compression. I will then do as tornitore45 suggested and run the engine by motoring it (either with my lathe or electric drill) to loosen it up and also, hopefully, lap the rings to the bore to reduce friction and give better sealing (and I agree that the extra pipework on the compression check probably means the compression check figure is not that great, although the tester does have a check valve). I removed the glow plug and put an air line with about 7 bar (about 101 psi) down the plug hole, with the engine just past TDC, and the engine did turn around to BDC but not really as quickly as I thought it might do, again making me think that I should run it to loosen it up a bit.

Anyway, that's it, for now. I shall put an update when I get around to carrying out my next steps (probably next week), for anybody that might be interested.

Thank you, again, for all your comments. They have been very helpful indeed.


  • Comp_Calc_Dims_Explain.pdf
    556.9 KB · Views: 271
  • Valve_Timings_20201230.pdf
    405.3 KB · Views: 248
  • Compression Ratio Calculations_20201230.pdf
    389.7 KB · Views: 186
  • 20201230_100414.jpg
    134.1 KB · Views: 351
  • 20201230_100419.jpg
    138.3 KB · Views: 355
  • 20201230_101754.jpg
    168.6 KB · Views: 353
  • 20201230_104612.jpg
    122.2 KB · Views: 337
I cant quite follow your timing abbreviations but for reference here is a methanol glow radial (Ohrdorf 5-cyl I'm building). A while back we discussed this in conjunction with the Edwards radial which I believe was quite similar just to ballpark.

Re carbs, the Perry website has some recommended sizing guidelines. I just put them on a plot to give you visual comparison to your engine specs. Its another one of those things that seems to vary among individuals, probably because of the wide variations in home made engine designs.


  • SNAG-12-30-2020 0000.jpg
    SNAG-12-30-2020 0000.jpg
    50.3 KB · Views: 232
  • SNAG-12-30-2020 0001.jpg
    SNAG-12-30-2020 0001.jpg
    21.2 KB · Views: 233
  • SNAG-12-30-2020 0002.jpg
    SNAG-12-30-2020 0002.jpg
    20.8 KB · Views: 241
What I was getting at a few posts up - what you guys have called bounce or pop as you rotate through the compression stroke. I totally get this BTW, started RC engines many, many times & its a very reliable 'feel' indicator. A clapped out engine or one that has been run lean & hot feels very different. But what I was getting at is when there are more or bigger unknowns or unfamiliarity. (I haven't done this myself, so follow along & see if its benefit or waste of time).

CR, compression ratio, is a dimensional thing relating the 2 volumes at BDC & TDC. We 'know' that we need to be within some specific range, lets say 8:1 for methanol. We can work it out in CAD or we can validate by filling the BDC & TDC volumes with a fluid & calculate. So we perform this check and CR checks out.

But this doesn't say anything about the actual seal meaning piston (ring), valves, head interface, glowplug thread/washer.... So what I was wondering as maybe the next easiest step - leave the valves closed with their springs on but pushrods off. Hook up pressure gauge to glowplug hole. Lubricate cylinder wall. Then slowly crank through TDC. I would assume with no ignition or valve breathing, we should see something on the gauge proportional to CR * atmospheric pressure. So for example 8 (to 1) * 14.7 = 117 psia theretical peak. So if you got say 60 psia max, that would indicate leakoff 'somewhere'. More I was thinking because he was swapping heads & such so comparing the difference between 2 cylinders all things equal. Is there any merit to this or am I barking up a tree?
Glow engines run fine in the cold. Slightly less cooperative to start, but not much. Warming the engine is common, I don't fly in the cold, but I've heard it all. Hot coffee over the engine, under the car exhaust, adding gasoline to the fuel. We know getting the air fuel mixture to a condition where it can ignite is the problem with cold starting. Alcohols are problematic with low vapor pressure, which is why E85 in the northern US can be 50% gasoline in winter.

Looking at the photo with the degree wheel. Do you have the valves timed to run in the direction the starting ratchet indicates? Model aircraft engines generally turn the other direction and I would guess the plans time the engine that way. I'm sure that you know due to your valve timing measurements. i'm jsut looking at the photo.

Are you using a propeller or flywheel?

Model engine fuel kept in the original plastic jug is not likely to pick up much water.

Valve events just need to be close to run. The measured timing are short to make good power, but should run. The engine has no overlap due to the close proximity of the valves to each other. Compression is also not critical.

Unfortunately, I couldn't find anything about this engine anywhere. There is someone on ebay.de selling some components he makes for this design.
With regards to @petertha's comments:

I would be fully in favor of instrumenting the thing and measuring pressure -- except I don't know what it should be! I just do the flip test, accompanied by pulling the thing through compression by hand. If I get a nice springy "pop" when I flip it, and I don't feel any compression going slow, then I know the engine is worn (or, by repute, has Dykes rings -- I've never tried starting such an engine, but have been told by those in the know that it needs some pressure for the ring to seal, so you're in this catch-22 situation if you try to go slow).

@clemley's comment about observing no compression is worrysome, unless they're pulling it through slowly. I understand that new rings can need to wear in a bit to develop good compression, so I might try a really smart flip to get the speed up, and I might try some straight castor in the cylinder to improve the seal (there is fuel in the cylinder, yes?). Then I might hope that after running for a bit the ring will seal better.

Another comment that @clemey made was about motoring the engine to lap the rings to the bore -- how did you make the rings? I'm stepping outside of my direct experience, but the instructions I've seen on making rings involve holding them sprung as they'll be in the cylinder, then lapping them round. If you left out this step they may not fit the cylinder well enough to succeed.
It's been a while! I've finally got the thing running, both as single cylinder engines (I tried each cylinder separately) and, now, today, as the opposed twin it is supposed to be. There were a few things but the main one was the presence of holes in the second ring grooves of both pistons (see attached picture). I made the pistons exactly as per the drawings but it seems that there was a flaw in the drawings which meant that the second ring groove cut through into the interior of the piston, if you cut the groove to the right depth. I made new pistons with a shallower ring groove.

The other main things, I found, were making sure the batteries driving the glow plugs were always fully charged and getting the mixture ibn the carb right. It seems to need to be very rich to run but I suppose it is mainly methanol which has, I think, a stoichometric AFR of about 6.5:1.

Anyway, now that I Have finally got it running, I have disassembled the engine and am giving it to a friend as a project for him. I now want to move onto my next project: a gasoline-fuelled spark-ignition engine (rather than glow plug engine). Anybody got any suggestions as to where I can get some plans or am I going to have to do my own? If so, I will start with a single cylinder one.


  • Hole_In_Piston.PNG
    845.5 KB · Views: 213
I was going to post a video of it running but it looks like we can't post videos on here. Have I understood that correctly?
Thanks for the reply. I might try that.

I forgot to mention, in my earlier post, that the engine runs fine while the glow plugs are powered up but, as soon as I switch them off, it stops. I thought glow plug engines only needed the glow plugs powered up for starting. Could this be because the engine is running very rich and so cool?
Glow engines should run just fine without the battery. It's part of their utility in models.

There's a lot of reasons why a glow engine may not run without power to the plugs -- the two main causes (assuming a good piston seal) is that the thing is undercompressed for the fuel, or that the glow plugs are recessed in their holes. You can fix the former by skimming the heads (or, if it's close, increasing the amount of nitromethane in your fuel). You fix the latter by using a longer plug, or modifying the head.

We had a big discussion about this not too long ago with someone concerning their first engine. They had both a recessed plug and too little compression. Once they modified their head, the engine worked fine.
Thank you. The compression is certainly high enough: you can really feel it when you are trying to turn the engine over. I reckon it is the recess of the plug.

Building a glow plug engine has been a real learning curve. Very interesting indeed.
The rule is that the business end of the plug should be flush with the inner surface of the head. If it's proud by more than a thread or so it'll cause detonation, or at least mess up the combustion chamber geometry -- and there's usually not a lot of clearance in there anyway, especially if you're using an RC plug with a bar. If it's recessed by more than half the diameter of the hole, then that recess can become a reservoir unburnt gasses that shield the plug from the fresh charge coming into the engine. Certainly if it's recessed by more than 1/8" then you've got problems, and you need to either rework or select a different plug.
Four stroke plugs are different from 2 stroke.
It is common for 4 stroke plugs to have a much smaller hole where the element resides, also they can be a longer reach.
Most are installed with a bias towards the exhaust valve.
All allowing for heat retention.
2 stroke plugs tend to be 1/2 the cost of 4 stroke.
All above considered, I tend to go for a 2 stroke plug, run on battery for idle speed and transition and zero nitro.
Nitro rusts the guts out of an engine if the residue is not attended to after running plus it's expensive.

Latest posts