A Reclusive Oriental Range:
the Boxer Engines

by Adrian Duncan

Click on images to view larger picture.

    The Boxer 29
    The Boxer Range in the Modelling Media
    The Boxer .099
    The Boxer Engines On Test
    So Why Buy a Boxer?
    A Later Boxer .099 Variant

Those among our readers who sometimes feel that our in-depth articles on model engine history convey too much detail (and we know you're out there!) will doubtless heave a sigh of relief when we state up front that there's somewhat less than usual to say about the subject of this article. We're talking about the Boxer range, about which is so little known, outside its south-east Asian area of origin, that many long-term collectors appear never to have heard of it! Even in Japan, where the engines were undoubtedly marketed, examples are reportedly very few and far between.

The Boxer engines appeared on the Japanese market for a relatively brief period during the decade following the conclusion of WW2. At the time in question, Japanese modelling activities, including model engine production and marketing, went almost completely unreported in the English-language modelling media. At present, we do not even have any definite information on where they were produced or when production commenced and ended.

Accordingly, the best that we'll be able to do here is describe the engines on the basis of several examples in our possession, test them on the bench and offer some highly tentative deductions regarding their origins. We therefore wish to stress that many of the conclusions presented herein should be regarded as conjectural at best and hence requiring further corroboration before being accepted as fact.

It may be asked why we bother to publish information on such an obscure model engine range about which so little definite information is available. Our answer is that if we don't do so, who will? At least this article will make people aware of the existence of the Boxer engines which in itself will be a step forward. This in turn may result in the uncovering of further examples. Moreover, the publication of an admittedly inconclusive article of this nature offers the possibility that someone who knows more than we do will read it and be prompted into coming forward to share their knowledge with the rest of us. As always, any and all contributions will be gratefully received and fully acknowledged!

In order to draw some tentative conclusions regarding the production history of the Boxer engines, it is essential to understand something of the context in which they appeared on the Japanese market. We will therefore present a brief summary of the Japanese post-war model engine production scene for those readers who are unfamiliar with this topic. Readers wishing more details are referred to the introductory chapter of the Hope engine range.

But before going any further, I'd like to acknowledge the considerable support and assistance rendered to me by my valued friend and colleague Alan Strutt. Not only did Alan make available the fine example of the Boxer 29 which appears in this article, but he also read the text in draft form and offered many valuable suggestions. We didn't agree on all points of interpretation, so I must stress that I accept full responsibility for the following text. However, Alan's comments added much to the context within which my own tentative conclusions are drawn. Thanks, mate. OK, on with the story as I see it...


Following the conclusion of WW2, the Japanese economy was in a shambles, as was the country's infrastructure. However, it was not long before reconstruction aid together with the economic boost stemming from the arrival of the predominantly American post-war occupation forces combined to re-awaken the enterprising spirit which had always been a characteristic of the industrious Japanese nation in "normal" times.

This re-awakening extended into all facets of the Japanese economy, including the field of model engine manufacture. It's vitally important to understand at this point that this was in every sense a re-awakening—unknown to most Western modellers, a thriving power-modelling scene had existed in Japan prior to WW2. In response to the consequent pre-war domestic demand, a number of commercially-produced Japanese model engine ranges had appeared in the 1930's, mirroring the pattern established in America and elsewhere.

Most notable among these was the Osaka-based OS marque which had become well-established in Japan following its initial market entry in 1936 and had even succeeded in penetrating the pre-war US market to a small extent. But OS was far from being alone—there were quite a few other pre-war Japanese makers, most of whom are now largely forgotten.

This situation has always been greatly under-appreciated in Western modelling circles because the pre-war Japanese power-modelling scene had been completely invisible to the non-Japanese speaking world due to language and political differences. But the fact is that many of those who became involved with model engine manufacture in Japan after the war actually had considerable prior experience in the model engine field. This was far from being merely a group of talented but inexperienced imitators who were learning on the job, as is often assumed—in fact, we're speaking here of a number of experienced power modellers resuming an activity that they already understood very well indeed.

Following the conclusion of WW2, the founder of the OS range, Shigeo Ogawa, wasted no time in re-commencing his pre-war business activities, with results which are still very much in evidence today. OS soon faced domestic competition from the makers of the Hope range as well as from an offshoot of the Mamiya camera company, which anticipated the East German Carl Zeiss optical company by lending its name to an independently-produced line of model engines for some years. In 1949 the manufacturers of the Fuji range threw their hat into the ring and were quickly joined by the Enya family who went on to create one of Japan's best-known and most durable model engine marques. At some point during this period, the Boxer engines also made their appearance on the Japanese market.

We may as well be honest from the outset and admit that at present we are unable to state with certainty that the Boxer engines were actually made in Japan! They were certainly marketed there, but that does not of itself prove a Japanese origin. However, such limited evidence as we possess, including anecdotal information from present-day Japanese sources, suggests to me that it's more likely than not that the Boxer engines were Japanese-made. We would welcome further evidence on this point, one way or another.

A noteworthy characteristic of Japanese model engine design during the pre-war and early post-war years was a tendency towards a minimalist approach—the engines were designed to do their job and nothing more, with a striking absence of features that did not contribute to that goal. This naturally resulted in minimal emphasis being placed upon styling along with a general avoidance of over-engineering. It also tended to result in engines which performed very well but were somewhat more fragile and utilitarian than their counterparts from elsewhere. It was only in the last few years of the 1940's as the marketing ambitions of Japanese manufacturers expanded in scope that this began to change and styling began to draw more attention from Japanese designers, together with greater concern for adequate structural integrity.

Another thread running through Japanese model engine marketing at this time was the choice of Anglicized names which carried connotations of speed, power and progress. Such names as Hope, Great Leap Forward, Typhoon, TOP, Mighty, and the like appeared regularly in the marketplace, so Boxer with its implications of speed, power and aggression fits right in with this concept. The Anglicisation of the engine's names was doubtless calculated to increase their appeal to the primarily English-speaking markets represented in Japan by the occupation forces.

Due to the limited availability in Japan of both glow-plugs and the platinum-iridium wire needed to make them, Japanese makers were somewhat later than their American counterparts in entering the glow-plug field. The first commercial Japanese glow-plug engines seem to have appeared in the marketplace in the latter part of 1948 as the plugs themselves became more generally available. OS did not enter the glow-plug field in a big way until 1949 when they released their twin-stack .29 cuin glow-plug model which succeeded their earlier 1948 limited-edition sand-cast spark ignition 29 cuin twin-stack design.

When we come to draw a few conclusions regarding the dating of the Boxer engines, we will have occasion to refer back to the above discussion. The two Boxer models of which we are presently aware are the .29 and the .099. Both of these are glow-plug models—indeed, to date we have encountered no evidence whatsoever that the Boxer range ever included a spark-ignition model (or a diesel for that matter). Combined with the obvious OS influence to be discussed below, this seems to place the introduction of the range to 1949 at the earliest.

There is persuasive architectural evidence to suggest that the twin-stack 29 model is the earlier of the two Boxer designs which have so far come to our attention, and accordingly we'll begin with a description of that model.

The Boxer 29

Esarlier, we mentioned the OS twin-stack sparker. The photos here show the glow version of the OS 29 twin-stack, along side the 29 Boxer. Although there are a number of obvious styling differences, there are similarities between the Boxer and OS .29 cuin models which appear too marked to be entirely co-incidental. The underlying designs bear a remarkable resemblance to each other, an impression which is confirmed by a detailed examination. The question immediately arises—in which direction did the influences flow? We'll offer a few speculative comments on this as we go through our description.

In order to do so, it will be necessary for us to draw the reader's attention to the similarities and differences between the OS and Boxer models. It's perhaps most logical to start with the tale of the tape. Here we find that the Boxer and OS twin-stack designs share the same bore and stroke figures of 18.8 mm and 17.3 mm respectively for an identical displacement of 4.80 cc (0.293 cuin). The Boxer is slightly heavier than the OS, weighing in at 229 gm (8.08 oz) as opposed to the 211 gm (7.44 oz) for the OS. It's a little hard to see where the extra weight comes from—perhaps the piston and cylinder units?

As can be seen, the two engines share a common crankcase layout with pressure die-cast cases featuring twin exhaust stacks and rather tall intake venturi for the front rotary valve (FRV) induction. The styling of the two models is however quite different, with the Boxer using aerofoil-section stacks (widely favoured in Japan at this time) as opposed to the more American-influenced "rounded rectangular" shape used on the OS. The internal stiffening column which is cast into the OS stacks is absent on the Boxer. Moreover, the Boxer case is left "as cast" in contrast to the attractive matte vapour-blast finish of the OS—perhaps a "minimalist" feature given the fact that the engine's performance is unaffected by the case finish.

To add further individuality, the outer ends of the Boxer stacks are given a concave form in plan view, while those on the OS are straight with a slight rearward angle. The Boxer also has a somewhat more "dramatic" bell-mouth section at the upper end of the intake. Overall, the Boxer may be said to present a rather more "Oriental" appearance than its more famous OS counterpart.

The name "Boxer 29" appears in rather stylized script cast in relief onto both sides of the crankcase. This mirrors the identification cast onto the case of the OS 29. However, unlike the OS there is no serial number on the Boxer. Nor does the Boxer emulate the OS in claiming to be "Made in Japan". Indeed, the only evidence that we have for a Japanese origin at present is a single documentary reference from Peter Chinn (to be noted later) plus the anecdotal opinions of several present-day Japanese sources.

One possibly significant difference in the crankcase design is the treatment of the mounting lugs. Those on the Boxer are far thinner and seemingly more fragile than those on the OS glow model. In fact, the Boxer lugs are far closed in thickness to those used on the earlier sand-cast OS spark ignition model. This may help us in attempting to understand the relationship between the OS and Boxer models, since this feature appears to place the Boxer more in the "minimalist" design camp which was in the process of losing favour as of 1949.

The porting arrangements for the two engines are identical with FRV induction and semi-radial cylinder porting. The exhaust port flange sits on a recessed shelf machined into the upper crankcase casting just below the lower edge of the twin stacks. A seal is ensured through the use of a gasket. There are two pairs of exhaust ports on either side, each pair feeding into one of the two stacks. The paired transfer ports are located directly below the exhaust ports. They are fed by two bypass passages cast into the crankcase, one on each side. The only difference is that the steel cylinder and integral cooling fins of the Boxer are left in their natural state, while those on the OS glow model are blued. The Boxer also has fewer but thicker cylinder cooling fins than the OS.

The hold-down system for the cylinders and heads on both models is identical, with two long screws extending downwards fore and aft though holes drilled through all the cooling fins to engage with tapped holes in the crankcase castings, and two additional short screws on each side to support the main screws in securing the cylinder head. Interestingly, the apparently-original screws on the Boxer are slot-head items, which generally signifies a relatively early date in a Japanese context. The 1948 OS spark ignition twin-stack model described elsewhere also used such screws, but the later glow-plug version switched to Phillips-head items in keeping with emerging Japanese practise.

The backplates on the two models are of entirely different design. That on the Boxer is a screw-in component similar to the item used on the original sand-cast OS 29 spark ignition model, while the backplate on the OS glow model is secured by four screws at the corners. In both cases a gasket is used to provide a seal.

Another obvious difference between the two designs is the method of producing the cylinder heads and the backplates. On the Boxer and on the earlier OS spark ignition model, both of these components are sand-castings, while the OS glow model uses pressure die-cast components throughout. The cylinder head of the Boxer has fewer and thicker fins than its OS "New 29" counterpart, but is very similar to the head of the original (and quite rare) 1949 OS twin-stack glow model, which had fewer head fins than its 1950 successor.

The head and base seals of my own example of the Boxer (obtained through the kindness of Alan Strutt) were both in very good condition with no detectable leaks, which made me reluctant to disturb them. I was also mindful of the fact that the cylinder fastening system with only two long screws is prone to introducing distortion when tightened down. When the cylinder is first tightened, it tends to take a "set" with some degree of microscopic distortion which is relieved through the break-in process. After the engine has settled down, as this well-used example clearly has, the original "set" is unlikely to be duplicated if the cylinder is removed and re-tightened. Accordingly, the engine must in effect be run-in all over again, with unpredictable results.

For these reasons, I elected not to disturb the engine beyond the removal of the backplate. This was sufficient to show that the crankshaft and rod in the Boxer are very similar indeed to those in the OS The main difference is that the Boxer follows the lead of the earlier OS spark-ignition model in using a conventional gudgeon pin mounted in bosses in the piston as opposed to the separate riveted carrier used in the OS glow models. The Boxer piston is flat-topped with a chamfer around the edge (in effect, a truncated cone) as opposed to the conical form of the OS.

It's possible to confirm without dismantling the engine further that the crankshaft journal diameter of the Boxer is 11 mm, with an 8 mm diameter internal gas passage. These figures are identical to those for the OS twin-stack models and the shafts from the two engines appear to be virtually indistinguishable. The only real difference is that the Boxer prop driver engages with a tapered section at the forward end of the shaft, while the OS uses a single flat to secure the driver. I must say that the Boxer system is far more durable.

I'm unable to confirm whether or not the needle valve assembly with the illustrated example of the Boxer is original or not. It fits perfectly with the jet hole being in exactly the correct location in the centre of the venturi, but it looks very much like an early Enya unit except that the double spring clip which provides needle tension is rather longer on the Boxer than on the Enya's in my collection. In addition, the only other Boxer 29 that I've seen on offer recently had a needle valve assembly which appeared to be identical to mine. That engine was mounted in a car, but the similarity of the needle valve assembly was nonetheless clear. So I'll confine myself to saying that it seems more likely than not that the illustrated assembly is correct.

It's also unclear whether or not the tank on the illustrated Boxer is original or a retrofitted OS item. All that can be said is that it fits perfectly and the red color is consistent with the anodizing applied to the .099 model to be discussed later. So which came first?

We hope that we've demonstrated that the similarities between the Boxer 29 and the various OS twin-stack 29's are too great to be coincidental. This invites us to examine the old chicken-and-egg question—in which order did these engines appear? Here we are unavoidable forced into the realm of informed speculation.

There's little doubt that the first of the Japanese 29 twin-stacks with which we are dealing was the sand-cast OS 29 spark ignition model. So the real question here is whether the Boxer came out next as a glow-plug rendition of the OS sparker by another maker (whether Japanese or otherwise) or whether the OS glow-plug model followed the sparker and inspired the subsequent release of the Boxer version. Either scenario is of course possible, so the following comments represent a personal view only and should not be taken as substantiated fact pending the appearance of further evidence.

When we compare the OS spark-ignition engine with the Boxer, we find that there are some very distinctive indications. The Boxer has the relatively thin mounting lugs of the OS sparker and also uses a screw-in backplate. In addition, both the head and backplate of the Boxer are sand-cast like those of the OS sparker, although the case itself is a quite complex pressure die-casting. The Boxer also uses the conventional gudgeon pin arrangement of the OS sparker as opposed to the riveted internal carrier system used on the OS glow-plug models. The head screws on the Boxer are slot-head items like those used on the OS sparker, generally an indication of an early date in a Japanese context. Finally, like the OS sparker, the Boxer lacks the stiffening columns in the exhaust stacks which appeared in the later OS glow-plug models.

By contrast, the OS glow-plug models use a die-cast backplate which is secured by four Phillips-head screws at the corners. They also feature far thicker mounting lugs on the diecast case. The original 1949 model had a head with relatively widely-spaced finning very much like that of the Boxer, while the later 1950 "New 29" model had more closely-packed fins on its diecast cylinder head. Both models carried the gudgeon pin in an internal carrier which was riveted to the piston crown. They also feature the stiffening columns in the exhaust stacks.

Looking at all of this, I personally feel that the Boxer is rather more easily seen as a derivative of the original OS spark ignition model than as a spin-off from the OS glow-plug design. The relatively thin mounting lugs, the screw-in backplate, the design of the stacks, the treatment of the gudgeon pin, the slot-head screws and the sandcast head and backplate are all direct reflections of the OS sparker. Furthermore, the design of the Boxer case appears more consistent with the minimalist design philosophy mentioned earlier which was beginning to lose favour by 1949.

Attention has been drawn to the presence of vestigial finning fore and aft on the crankcase below the exhaust ports of the Boxer. It has been suggested that these indicate an influence from the OS glow-plug models, which had sub-exhaust finning around the entire circumference of the upper case. In my personal view, the Boxer's vestigial fins may just as easily represent an attempt by the Boxer designer to strengthen the tapped cylinder hold-down lugs against the very considerable wedging stresses imposed by the hold-down screws. It's therefore entirely possible that OS simply carried the Boxer concept a step further by extending the sub-exhaust finning around the entire case circumference. Overall, I don't see this as a dependable indicator one way or the other.

If we were to believe that the Boxer was inspired by the later OS glow-plug models, we would have to believe that the designer deliberately decided to use thinner mounting lugs and also abandoned the sub-exhaust finning in addition to reverting to the less sophisticated sand-cast screw-in backplate and sand-cast head. It would be rather "un-Japanese" for the Boxer designer to have produced a retrograde design of this nature—the tendency was either to "clone" or to improve. That said, we have noted the possibility that the engines were in fact made elsewhere, and if that could be proved to be the case it might be easier to rationalize changes of this nature. As it is, we presently lack any documentary or anecdotal evidence for a non-Japanese origin to balance that which we have in favour of such an origin.

So although I can't prove it, for now I'm personally prepared to adopt the provisional working hypothesis that the Boxer 29 glow-plug model was a Japanese product that was most likely inspired by the original OS twin-stack spark ignition models and may well have appeared on the market in 1949. At that time it would have been quite an advanced design by comparison with some other contemporary Japanese designs like the Hope B and the Mamiya "Chochin" engines. It actually seems entirely possible that it preceded or at least paralleled the OS twin-stack 29 glow-plug model onto the market. I myself have great difficulty in seeing its introduction as having been later than 1949.

I must reiterate yet again that the above conjectural comments represent my own views only—others may see the matter differently. As always, I remain completely open to any persuasive evidence that might point towards a different interpretation.

The Boxer Range in the Modelling Media

The sole English-language reference to the Boxer range that we've been able to find appeared in an article by Peter Chinn in the April 1954 issue of "Model Aircraft". In that article, Chinn discussed the currently-available International class (2.5 cc) engines from various countries and noted that Japan had no engines in that class at the time, the nearest displacement category in use in Japan being the .099 cuin category.

The Boxer .099 was specifically included in the list of .09 cuin engines then in production in Japan. Chinn implied that the Boxer .099 was part of the Boxer "range", suggesting that other Boxer models may have remained in production at his time of writing (early 1954, allowing for editorial lead-time). Rightly or wrongly, Chinn clearly believed that the Boxer engines were of Japanese origin.

However, by the time Chinn published his list of glowplug engines for 1954/55 in the May 1955 issue of Model Aircraft, we find no mention of the Boxer engines. We know from the earlier article that Chinn was aware of this range and we may safely assume that it would have been included if it was still current even in late 1954. It thus appears that the Boxer range vanished from the market at some point during 1954, at least as far as Chinn was aware.

And that, believe it or not, is all that we are able to deduce from the contemporary English-language modelling media. Unfortunately, these references tell us nothing about the introductory date of the range. If anyone out there has a reference that we've missed, please share it with us! So all that remains for us to do is to describe the other Boxer model that has come to our attention—the Boxer .099.

The Boxer .099

It's pretty clear from the architectural evidence that the introduction of the Boxer .099 model post-dated that of the Boxer 29 described earlier. The .099 is a far sturdier and more up-to-date design in a number of ways, displaying none of the minimalist design philosophy that was in evidence in the larger model. It also displays no design influence whatsoever from its larger sibling.

I noted earlier that circumstantial and architectural evidence seem to me to place the introduction of the Boxer 29 in 1949 at the earliest. We have just presented documentary evidence to suggest that the Boxer range appears to have departed from the scene at some point in 1954. At present we have no information whatsoever on the date of introduction of the Boxer .099, but on architectural grounds it appears likely that it followed its larger brother onto the marketplace in the early 1950's, perhaps around 1951 or 1952. At present we can't be any more definite than that.

The only contemporary image of this engine that we have at present appears on an undated Japanese-language promotional leaflet issued by a Tokyo-based company calling itself the Yuhido Corporation. This image is invaluable because it shows the engine in its as-sold configuration. Once again, I'm immensely indebted to Alan Strutt for making an image of the leaflet available.

We obtained a translation of this leaflet in hopes that it might tell us more about the Boxer range, but sadly this hope was not fulfilled. The Yuhido Corporation does not appear to have been the manufacturer of the Boxer engines—instead, the company characterized itself as a wholesaler and retailer of technological products, clearly very much including model aircraft. The Boxer .099 was simply included as the latest model engine on offer from the company, along with a kit called "Mr. Trainer" which could apparently be built in only three hours! Readers of the leaflet were invited to contact the Yuhido Corporation with "any questions about model airplanes".

The leaflet featured the Boxer .099 front and centre, claiming it as a "new standout model engine" which had both "excellent performance and style" and was "eminently suited to beginners". The quoted price of the engine, including shipping, was •1,278, which converts to a cost of US$3.55 at the prevailing exchange rate of •360 to the US dollar established under the American Bretton Woods System in 1949. Clearly it was happy times for US modellers stationed in Japan with the forces of occupation!

The leaflet also stated that the Yuhido Corporation offered "a complete selection of all types, all models and all makes" of model engines. This being the case, it's interesting that the Boxer .099 was specifically featured in this leaflet—perhaps the Boxer engines were a "house brand" of the company?!? Or perhaps it was just the latest engine on offer at the time the leaflet was produced.

The "house brand" scenario plus the absence of any "Made in Japan" designation on the engines themselves once again raises the interesting possibility that the Boxer engines might not have been made in Japan at all but might have been produced for Yuhido under contract in, say, Taiwan (or Formosa as it was then commonly called). It also raises the possibility that the two very different models (.29 and .099) could have been produced by two different manufacturers under separate contract. This would certainly explain the very different design and production styles of the two models. We must make it clear that at present we have absolutely no evidence to support such interpretations, but these possibilities should be borne in mind by future researchers.

One very interesting tid-bit that comes from the leaflet is the fact that the Yuhido Corporation had been established in 1908! This made them a contemporary of Gamges in London, which included what amounted to a hobby department from the early years of the 20th century onwards. There's no certainty that the company was in the modelling business back in 1908, but the possibility exists given the long history of modelling in Japan to which reference was made earlier.

A company called the Yuhido Corporation (Yuhido KK) remains in business today, dealing in mechanical and electrical products. It claims to have been in business for a "long time", and it's entirely possible that it is the same company still going after over 100 years. If so, it outlasted Gamages, which closed in 1972!

The image on the Yuhido Corporation's leaflet shows that the Boxer .099 was originally supplied with a metal back tank. Although the image is not in color, the tank in the image is clearly anodized, and it seems safe to assume that it was anodized red like the exhaust collector ring. Accordingly, the attached images of the sole example in our possession show the engine fitted with a modified OK Cub. .074 tank, which fits perfectly and looks more or less identical to the tank in the Yuhido image. At least it creates a far closer visual impression of what the engine looked like straight out of the box. On this used example, the exhaust collector ring has faded somewhat from heat—the tank color is probably close to the original.

The illustrated example also arrived without a needle, and the item with which it is now fitted is a reproduction based visually upon the component in the illustration on the Yuhido leaflet. The spraybar appears to be original. The thread is a very obscure 3.5 mm x 0.35 mm, and I had to use a home-made tap to produce the split thimble for the repro needle. On the upside, this should give extremely fine mixture control in service.

The Boxer .099 is basically a conventional radial-ported .099 cuin glow-plug motor with one or two unusual features. One such feature in the context of the engine's time, place and displacement is the fact that it's a long-stroke engine. Measured bore and stroke are 12.40 and 13.50 mm respectively for a displacement of 1.63 cc (0.099 cu. in). The engine weighs 109 gm (3.85 ounces) all complete—a relatively hefty figure for a glow motor of its displacement.

The most immediately obvious visual anomaly that leaps out at an observer is the red-anodized exhaust collector ring which surrounds the exhaust port belt. This has large oval holes stamped around half its circumference and can be rotated to any desired position to direct the exhaust in an appropriate direction for the model in question. It is a slightly "binding" fit, which should encourage it to retain its position once set. It can also be removed entirely if desired by slipping off past the cylinder head, and it's a good bet that many of them were removed in this way and subsequently lost. Fortunately, this has not been done to the illustrated example.

The engine is based upon a set of well-executed sand-castings, all of which are relatively massive for an engine of this displacement—nothing in the least minimalist about this design!! The use of sand-casting throughout is another somewhat odd feature—we know that the makers of the Boxer 29 were well able to produce excellent pressure die-castings. So why the switch to sand-castings? Perhaps a different manufacturer was involved, as suggested earlier, we'd welcome any other ideas.

No marks of identification are cast onto the engine at any point. The only identification is the word "BOXER" stamped onto the left side of the integrally-machined location ring which forms the upper part of the main casting and accommodates the female thread for the screw-in cylinder. This stamping was clearly done by hand, since the "B" was initially mis-struck. There is no serial number, and once again the engine makes no claim to having been "Made in Japan".

Another design feature which is somewhat individualistic is the fact that the engine has bolt-on front and rear covers. This was somewhat unusual for a sports engine of this displacement—most contemporary engines in this displacement class had one-piece crankcases with bolt-on or screw-in backplates. Enya and Mamiya of course reversed this arrangement by using bolt-on front bearing housings with cast-in-unit backplates. But as far as I'm presently aware, no other contemporary Japanese .099 used bolt-on components both front and rear. Generally speaking, this form of construction was confined to racing engines.

It seems possible that the manufacturers adopted this route to give themselves the option of developing a rear induction version of the engine. We have yet to find any evidence of the existence of such a variant, so at present there's no reason to suppose that such a model was ever produced.

The backplate seals to the case with a very thin and typically Japanese circular paper gasket without "ears" for the screws at the corners. The fit of this component in the case emphasizes the precision of manufacture—it is a slightly tight plug fit in the case and would almost certainly seal without a gasket simply by virtue of the excellent fit and fine machined finish. An unusual structural feature which should be noted at this point is the unusually short length of the slot-head hold-down screws—there are very few exposed threads to engage with the tapped holes in the case. This is not an engine for frequent dismantling or over-tightening of the screws! Fingers only, please.

Removal of the backplate in defiance of the above cautionary note reveals that the engine uses a stamped steel con-rod along with a conventional steel gudgeon pin and a cast iron piston. The one-piece steel crankshaft has a full-disc crankweb with a modest counterbalance. The journal diameter is 8 mm with an internal gas passage of 6 mm diameter, resulting in a rather minimal wall thickness of only 1 mm. This is probably adequate for a relatively low-powered sports engine of this displacement, but still seems a bit marginal. The induction port (visible though the venturi) is a simple round hole drilled through the shaft wall. The shaft runs directly in the material of the case, no bushing being used. By feel, the shaft is a superb fit in its bearing—in fact, all fits throughout are beyond reproach.

The front rotary valve is supplied with mixture through an early-'50's Enya-styled angled bell-mouth venturi which is cast integrally with the front bearing housing. Throat diameter is 5 mm for a rather fat spraybar having a diameter of 3.5 mm. Clearly the intent here was to promote good suction—a useful characteristic in a sports engine intended for beginners. There's little doubt that substantially more power could be released by judicious "waisting" of the spraybar.

The cylinder head is a plain unfinned unit which is secured to the cylinder using four screws. Unlike the screws used for the backplate, these are of adequate length and utilize all of the available thread in the upper cylinder flange. The use of a long-reach plug is implied by the length of thread provided for that component.

The steel cylinder itself is a screw-in item with integrally-machined cooling fins. Unusually, these fins are cadmium-plated, the plating being applied only above the port belt. Porting is Oliver-style, with three generously-proportioned milled exhaust ports separated by pillars. The transfer ports consist of holes drilled at an angle upwards though each of the three pillars to overlap the exhaust to a considerable extent.

The bypass passages consist of three channels cast into the interior wall of the main casting. The design is in fact very similar to that of the Elfin 2.49 BB Mk I. Using this approach, the three transfer ports clearly have to line up with their respective bypass passages. With a screw-in cylinder, this is a somewhat problematic approach. The Elfin got around this difficulty by dispensing with the screw-in cylinder feature in favour of a cylinder hold-down collar which threaded onto the top of the main casting and engaged with a flange at the cylinder base. However, no such collar is in evidence on the Boxer, which evidently uses a conventional screw-in cylinder design. It appears that the required alignment must have been achieved either through the use of shims or by selective matching of cylinders to crankcases.

In view of the above observation, I elected not to disturb the engine further, since there was no way that I could guarantee that it would go back together in the same alignment. Since the engine was clearly well run in, I judged it best to leave well alone. Accordingly, the above description is as far as I'm able to go unless further information becomes available down the road.

The Boxer Engines On Test

In the interests of making this review as complete as possible, I felt more or less duty-bound to undertake some test running. I had few qualms about doing this as far as the very sturdy Boxer .099 was concerned, but I admit to being a little leery of putting too much of a strain on the Boxer .29! The example in question has clearly had a fair bit of use and has survived intact apart from a set of broken mounting lug "ears" which had been neatly repaired for a previous owner by welding. I was a bit uncertain as to the integrity of these repairs, and as it turned out I was right to worry..........

For the tests, I used a fuel containing plenty of castor oil and 10% nitro. I also chose airscrews that would not have any tendency to lug the engines—always a wise approach when running glow-plug engines given the lack of any positive ignition timing control while running. To provide a bench-mark for comparison purposes I also brought along my "flying" example of the OS "New 29" model which has so much in common with the Boxer .29 in design terms and is almost certainly a contemporary product. As a bench-mark competitor for the Boxer .099, I included a Mamiya .099 of the post-1950 loop-scavenged variety, which was almost certainly a contemporary of the Boxer .099.

The little Boxer .099 proved to be a very good little engine, albeit not spectacularly powerful. It needed a small prime to start, but responded well when that was given and ran very smoothly indeed once speed was increased to above 11,000 rpm. Below that speed, there was a slight tendency towards a misfire which could not be entirely eliminated by needle adjustments. Overall, I gained the impression that the engine was somewhat asphyxiated by the very fat spraybar which blocked much of the intake area, and I have little doubt that "waisting" the spraybar would improve performance quite significantly. The engine certainly gave me the impression of wanting to give more.

The exhaust collector ring proved to be insufficiently tightly fitted to prevent rotation while the engine was running, and I must admit to seeing this fitting as being of very questionable value—it simply rotated continually while the engine was running and thus ensure that the exhaust spent time being fired out in all directions rather than any particular intended direction! If I were planning to use this engine in a model, I would almost certainly remove the collector ring, as many owners no doubt did.

I used the back tank for all testing. There's no assurance that the modified OK Cub .074 tank now fitted is of identical capacity to the original—I merely trimmed it to a length which appears to match that shown in the Yuhido Corporation image presented earlier. Be that as it may, the present tank gave an 80 second run when the engine was propped for over 11,000 rpm—an adequate duration for free flight use.

Both the Boxer and the Mamiya were tested using the same props and fuel. The Boxer was well down on the far lighter Mamiya on all props tested. Actual figures were as follows:

PropellerBoxer 099 (rpm)Mamiya 099 (rpm)
APC 8x4 GF 8,6009,800
Taipan 8x4 GF8,90010,000
Taipan 7x4 GF11,20012,600
APC 7x4 GF11,40012,800

These figures imply a peak for the Boxer of something in the order of 0.11 BHP at perhaps 12,000 rpm—not all that spectacular but a useful enough performance for sport flying. The Mamiya did quite a bit better, and in strictly performance terms it's no contest. In addition, the Mamiya is far lighter and somewhat more compact.

Turning to the larger engines, I first tested the OS, which started and ran superbly as always and delivered a very sturdy performance. When I switched to the Boxer, I immediately found that the compression on this well-used example is marginal for hand-starting, although in my judgment there was just about enough. I was still working away on it trying to find the correct needle setting when I noticed that the engine felt a bit loose in the stand. Upon investigation, it turned out that one of the repaired mounting lugs had re-fractured. So that was it for this test. I was easily able to repair the problem, but this engine is now sitting in my collection in the "non-running" category—I won't try again given the demonstrated risks.

In any event, it probably doesn't matter much. The fact that the engine is pretty much a clone of the OS in strictly functional terms implies that it would likely deliver a somewhat similar performance. On this test, the OS delivered the following figures:

PropellerSpeed (rpm)
APC 10x6 GF10,100 rpm
Taipan 10x4 GF11,600 rpm
APC 9x4 GF12,600 rpm
Taipan 9x4 GF12,800 rpm

These figures imply a peak of just over 0.40 BHP at somewhere in the region of 12,000 rpm—a very useful performance for a 1950 design! The Boxer might not match that, but there seems to be no reason why a good example shouldn't come reasonably close.

So Why Buy a Boxer?

...which is undoubtedly the question that we're left asking ourselves after going through the above exercise! The Boxer 29 is in effect a clone of the OS .29 glow model, albeit of simpler and (it must be said) less sturdy construction. Looking at the two engines side by side, most people would undoubtedly go for the OS—it simply looks far more sturdy and well-finished. In addition, it's slightly lighter, as we saw earlier.

The Boxer .099 is undoubtedly a well-made and attractive little engine which works well and looks as if it would give long and dependable service, but it doesn't come close to matching a typical competitor in the shape of the Mamiya. At 67 gm all inclusive (2.36 ounces), the Mamiya is 1-1/2 ounces lighter than the Boxer and is both more compact and considerably more powerful. Any practical modeller facing a choice between the two designs would undoubtedly go for the Mamiya.

Unless they enjoyed a significant price advantage over the competition, it's actually hard to see how the Boxer engines could have maintained a viable position in the Japanese model engine market. And indeed, they appear not to have done so—they were marketed for at most some five years, and possibly significantly less. Their extreme rarity today is ample testimony to their relative lack of sales success.

A Later Boxer .099 Variant

The greatest reward that we get from the publication of these articles is the receipt of additional information from one or more of our readers. The initial publication of the above article resulted in not one but two such contacts.

Firstly, our valued colleague Tim Dannels, to whom we owe so much for his outstanding research into model engine history over the years, was kind enough to send along photos of a couple of data cards from his archives. These showed black-and-white images of a couple of Boxer .099 engines owned (or now perhaps formerly owned) by friends of his.

The first of these was clearly identical to the example described in detail above apart from missing the exhaust collector ring. However, the cylinder was reportedly a pale blue color instead of being cadmium-plated as on my own example. It was not clear from the data provided whether or not this coloration was original or an owner modification. The owner of that example placed the date at 1950, although the basis for this dating was not supplied. Still, this date is not at all inconsistent with our own estimate presented earlier.

The second image supplied by Tim was far more enlightening in that it showed a completely different Boxer .099 design! This was a more or less conventional loop-scavenged side-stack model, showing a strong similarity to the contemporary McCoy "9" .098 cuin model as well as some Mamiya .099 influence. The date of this model was quite credibly put at 1954, which suggests that it was the final version of the Boxer .099 and followed after the radial-port model described in detail above.

Another interesting piece of information was the provision of the makerís name on the cards to which these images were attached. The name is rendered as "Takara DA" and the country of origin is given in both cases as Japan. Placing the constituents of this name in their English order yields the name DA Takara. This is not a name with which we are familiar, and we leave it to others to follow up on this rather slim lead. However, the name is certainly consistent with the claimed Japanese origin of the Boxer engines.

Unfortunately, the images available to Tim are somewhat lacking in clarity, through no fault of Timís whatsoever. We were therefore doubly blessed to hear from .09 expert Dan Vincent, who recently sold a Boxer .099 of the later type on eBay (Drat! I missed it!). Happily, Dan retained some excellent images of this engine, so we are able to present those images here as a record of what must surely be the final Boxer .099 model, likely dating from 1954. Since we donít have access to the actual engine, we are unable to supply any further details at this time.

Our most sincere thanks go to both Tim and Dan for their invaluable assistance in helping us to tell the Boxer story as completely as possible. Itís responses like this that repay all of the hours that go into the preparation of these articles!


We hope that you've enjoyed this look at one of the more elusive Oriental marques from the early post-war period. If anyone out there can add to our knowledge, please get in touch—all contributions gratefully and openly acknowledged!!



This page designed to look best when using anything but IE!
Please submit all questions and comments to enquiries@modelenginenews.org

Creative Commons License