The Improved Snow

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.


Well-Known Member
Project of the Month Winner
Aug 18, 2008
Reaction score
The Snow tandem double acting gas model engine designed by Doug Kelley is a classic among model engines. It is a bit different and as a result attracts attention at model engineering shows. It is a long build requiring attention to squareness and concentricity. It can be very disappointing when it doesn’t run like you expected, so I thought I would post this thread about my experience with the Snow.

Many of the Snow engines built to the drawings have exhibited a common running issue. I have seen this on at least four other engines, including mine. The engine will start and run for 30-60 seconds and then it slowly loses RPM, then stops. No adjustments seem to help. The engine is restarted and it loses speed and stops again. This repeats few times until it won’t re-start. Let it sit for 15 minutes and it will start and the whole cycle repeats itself. At other times the engine will slow quickly and stop. If your Snow engine has a similar problem I believe there is a solution.

The scale of the Snow model engine obviously presents a multitude of problems related to its scale, but I will focus mainly on the fuel system. The Snow fuel system uses a fuel/oil mix to provide lubrication to the cylinder and piston rod bushings. Neither require much lubrication, but should have constant lubrication which they get with the fuel/oil mixture. Conventional two cycle oils should be avoided. They were designed for high temperature, high speed, hard working engines and the Snow model engine isn’t anything of that sort. The Snow model is slow speed and produces rather low exhaust temperatures so a much lighter oil is all that is needed. Marvel Mystery oil (MMO) seems to do the job nicely. I have used 50:1 and 60:1 fuel/oil ratios of MMO with good results and still have a very wet engine inside. I believe the fuel/oil ratio could be increased to 70:1 or higher, but have not tried it yet.

The carburetor size is important. It must be small. The slow speed of the Snow (400-500 RPM) provides ample time for the needed fuel air mix volume to flow through a very small carburetor opening. I have found that an RC barrel type carb with a throat diameter of .120” (3 mm) is just about optimum for the Snow. A self compensating type is preferred, but due to the constant speed operation of the Snow engine this feature isn’t essential.

Real Snow engines had a fuel mixer adjacent to each valve cage with a common throttle control rod linked to a governor. These mixers used a gaseous fuel so mixing was good and control of mixture was not as important as on a liquid fuel system. The Snow model engine intake manifold is rather unusual. The individual intake runner lengths are all equi-distant to their respective valve cage. They are almost 6.5” long from carburetor to valve cage. When the engine is started the carburetor and adjacent portions of the manifold become quite cold due to the vaporization of fuel in the carburetor. There is little mass to the manifold so it will become cold very quickly well beyond the carburetor junction in both directions. There is little, if any, immediate heat feed back from the engine mass. For a long time I had thought this was the root of frequent running issues with my Snow. I tried using a heat gun on the carb after the engine was started and the results were very encouraging. The engine would run as long as I kept the carb area quite warm. One Snow builder built a jacketed intake manifold that utilized engine exhaust gas heat to warm the intake manifold. It didn't compete with a hot air gun. I suspected that moisture, fuel, and oil were condensing inside the manifold and that condensed fuel was then re-vaporizing inside the manifold upsetting the fuel air ratio after only a short period of running. In my mind, this theory correlated with the engine running for a while and then slowly stopping. But then I began to wonder if the problem was an accumulating slug of condensate itself. And if this was so, how would condensate flow uphill to purge itself from the rising intake runners of the Snow. My wife's old percolator coffee pot came to mind.

I felt a new intake manifold design might help if condensate, or condensate accumulation really was the real problem. The manifold needed should allow any condensate to be carried along into the engine instead of accumulating inside the manifold and restricting intake air flow. It also needed to be capable of providing heat to the carburetor area to offset the significant cooling effect of the vaporizing fuel. This manifold design was a simple flat plane log type manifold installed at the elevation of the intake valve cages. The intake runner from the manifold would be as short as possible. The carburetor would be center mounted as before. No gaskets would be used at the valve cage flanges to get more heat transfer into the manifold. I used a section of K&S .188” x .375” rectangular brass tubing for the main manifold. The cross section area is a bit larger than the original Snow tubular manifold. My idea was to first build a dry log manifold using one rectangular tube for the air intake flow only, and to try it on the engine. Then build one with two tubes soldered back to back with hot return coolant routed through the second tube. This would provide heat along the full length of the manifold, and around the lower part of the carburetor.

The running issues of my Snow had confined it to the shelf for over three years and I was very anxious to get it running properly. The single tube dry manifold was installed shortly before the East Ohio Model model engineering show in Zanesville last October. With the new dry manifold, the engine started easily and it ran,…..and it ran…and continued running for over 30 minutes before running out of fuel the first time on the new manifold. I was cautiously optimistic! Not only did it run like never before, but the minimum smooth engine operating RPM had dropped from 500-550 RPM to 400 RPM. I put about 3 more trouble free hours on the engine before the show. It ran without with problems at Zanesville for at least 3 hours. A few months later it also ran well at Cabin Fever for over 6 hours without any unexpected stopping. It will now run at 300-350 RPM if you watch it, but still seems happiest at 400 RPM. The intake manifold was still getting quite cold near the carburetor junction, but the cold area did not extend very far. The new short intake runners were back feeding heat to the manifold very nicely. I considered the new log manifold a success. My Snow will run continuously for about 40-45 minutes with this manifold before reaching a return coolant temperature of 165 degrees F, but the hotter it gets the better it seems to run.

After Cabin Fever I built a coolant jacketed manifold to see if things could get even better. I was surprised to find that performance wise, there was no noticeable operational difference between the dry or wet manifold. Both run almost the same. That was a little disappointing at first. While I was just as happy with either manifold from a performance standpoint, I noted that the jacketed manifold had three side benefits over the dry manifold. First, it solves all of the coolant flow and frequent coolant spillage problems of the Snow since it becomes a closed circulating system. Secondly, the heat absorbed by the carburetor vaporization contributes to engine cooling and extends the engine continuous operation period indefinitely after reaching about 150 degrees F. I like to run my engines. And third, the fuel consumption seems noticeably reduced. The only drawback to the wet manifold that I can see is that the distinctive Snow overflow tubes and drain funnels must be removed.

In my opinion, condensation in the original Snow intake manifold can become a percolating restriction to intake air flow disrupting and/or stopping engine operation. The flat plane log intake manifold allows condensate to be purged from the manifold as it accumulates allowing for a free flowing intake system and uninterrupted engine operation. It seems to work for me, and I now consider my Snow engine one of, if not the most, dependable running engine I have built.

A few photos of my three intake manifolds are included. I am sure my log manifolds can be improved upon, but these work. I would be very interested in hearing about any other Snow modifications that have improved running performance of the Snow engine.




So is it possible to post a drawing showing how you built the manifold using the cooling water to heat up the intake. I have the same problem not being able to get the engine to run reliable. I agree with you that the intake is to cold and needs some heat.

There is fuel drop out in the intake manifold of any liquid fueled engine. In most engine intake manifolds this condensate drop out is flushed through the manifold or re-vaporized by heat. In the original Snow manifold design flushing through is not possible due to the upward travel in the last few inches of intake manifold. The flat log manifold allows fuel that drops out to flow through without creating any restriction to intake flow. It is not the cold, but accumulated fuel and oil trapped in the low portion of the manifold that causes running difficulties in the Snow.

I suggest you prove this to yourself by building the simple unheated manifold version that I built first which is shown in the top photo of my post. My Snow ran just as well with the dry or heated manifold. I am sorry but I didn't make any drawings of either manifold, but the dry manifold is pretty intuitive. I have a few more photos of the coolant heated version and a cross section sketch of the center carb section that would allow you to fabricate it. It is a bit tricky. But try the dry version first. It doesn't take long to build. Let me know if it works for you. PM me if you have specific questions, or want the heated manifold photos and sketch.
Propane is a great model engine fuel and I use it on all of my other model engines. The tandem double acting Snow engine, however, requires cylinder, piston rod bushing, and piston rod seal lubrication not easily delivered with propane. Point lubrication of these places isn't feasible in my opinion. The intake manifold change has provided very reliable operation of my Snow engine on liquid fuel/oil mix and allows a much slower more realistic operating speed than previously attainable.

The big tandem engines must have had lots of point lubrication but You are probably rigth that it is difficult in model size.
Do You plan to put a video on Youtube?
My video skills are poor as evidenced by the few Youtube video's that I have done (Bruce- MacBeth electric plant, Bates & Edmonds electric plant, and the Titan 50 hp). I have never taken to time to improve them as there so many other interesting things I would rather be doing. If I do a video of the improved Snow I will send you a PM with the link.
Thanks rustkolector! I don't see myself making a Snow anytime soon, skills and tooling aren't there yet, but reading about your diagnostic approach is very enlightening.

Latest posts