Douglas Motorcycle


Your Douglas Photos
Dirt Track Racing Photos

On Road Douglas Photos

About this Site
About the Parts
Engines Frames
Cylinders Tanks
Cyl Heads Wheels
EngineParts Flywheels
Carbies Exhausts
Gearboxes Guards
Casings SmallParts1
Sprockets SmallParts2
Electrical SmallParts3
Footchange More Parts
Triumph 250
Site News
Search this Site
Old Classifieds
Douglas Parts For Sale
Douglas Parts Wanted
New classifieds
Douglas Parts For Sale
Douglas Parts Wanted
Douglas Motorcycle Forums
Privacy Policy

Observations of some Douglas Cylinder Heads

by Doug Kephart

On and off over the last few years I have been drawing up most any pre-war o.h.v. Dougie part I could lay my hand on. The bulk of which has been of the DT models. As one might gather from my previous contributions in the New ConRod magazine (NCR), I have a fascination of manufacturing, and its impact on the evolution of model changes. Phil Manzano's recent comment on gathering together via the NCR of information relating to the pre-war o.h.v. engines prodded me into dusting off this text, finishing it up, and suffering it upon you all!

So starting with the OC head, we have a true hemispherical combustion chamber with valves splayed at an included angle of 90 degrees. This makes for a very deep combustion chamber, which must have had a very high crowned piston to suit. And no, I have not indicated my theory of fused on rocker perches in these cross sections!
Douglas motorcycle cylinder head
click image for larger view
Douglas motorcycle cylinder head
click image for larger view
Next we have the head off a 750cc engine which is simply marked 'J5'. This engine is a bit awkward to date. It was with frame OF219, had been for some time, and the assumption is that it belongs to it. This frame was modified for hill climbing, and its shortened petrol tank and a few other details bear a striking resemblance to a photo of Reggie Pink hill climbing a Douglas at Muskeegon, Michigan. Pink always used 750cc machines (or so he wrote in a letter of 1937 that I have). The fore mentioned photo appeared in the Sep/Oct 1980 NCR, and was a reprint from a pre-war ConRod magazine.

Anyway I digress; this head is quite difficult to classify. The location of the rocker spindles is the same as the OC head (or vice-a-versa, depending on which came first). But the perches are completely different, no longer having the 'joggle' as seen on OB/OC models and including a vertical reinforcing flange 'angle iron' like in section.

The ports are still inclined at a 10 degree angle (in fact all four heads shown here are so) and still intersect 1-9/32" above the head surface. But now the angle between the valves has closed up to 82 degrees and intersects 3/8" off the surface of the head instead of flush with it. Why 82 degrees I have no idea, but I did measure several different ways to see if I could make it conform to 80 degrees but it would have none of it.

Of all the heads shown here, this one has the least support for the valve guides, very little at all really. Another bad feature is the very thin section between the floor of the ports and the cylinder head surface. As a matter of fact, both the front and rear heads are cracked from the head joint through the floor of both ports, from the valve seat out to the ends of the threads for the gland nuts. The gland nut being screwed up tight was probably the only thing keeping the heads together! These gland nut threads are the same as that of the OC head.

The ports are huge, and show evidence of extensive rotary file work. Whether done at the factory or by a subsequent owner I do not know. Perhaps a similar head will turn up to compare it to. It is possible that the ports have be 'opened up' since leaving the works, and that would explain quite a lot as far as the fatally thin port floor. But if that were the case more of the counterbore around the valve guide would have been ground away, which they are not. Unless someone went to the questionable trouble to deepen these, it indicates the ports are the size they were meant to be. The purpose for these counterbores will be debated shortly.
Now we move on to a DT cylinder head. The valve angle has closed up a further 2 degrees to a nice round 80 degrees. At least on the ones I measured, I would not be surprised if some transition heads turned up with 82 degrees as well. The point of the valve's intersection has moved slightly to 13/32". While the rocker spindles are still the same distance from the cylinder head surface at 3-3/4", they are now closer together.
Douglas motorcycle cylinder head
click image for larger view

It is quite possible there may be many flavors of DT heads, early ones, late ones, road racing ones, and so forth. Some have two gasket ring grooves cut to suit either 500cc or 600cc cylinder barrels, and there seems to be quite a few of these about. I would think this was a racing spares economy by the factory. It would allow interchangeability during an event without carrying a full range of spares. And who would be fielding a team of 500cc and 600cc capacity machines at a venue but the factory?

Some have a similar, albeit smaller, gasket groove at the intake and exhaust port flanges as well, apparently done on TT engines I have heard said. The late Bob Jones told me his 750cc sprinter had grooves for flat joint rings, but I can not remember if he said the grooves were in the heads or the barrels. But I wonder if this was not a modification done by enthusiasts to simplify making replacement head gaskets. Unless the opposing surface had no groove at all, which would indicate it had to have been done at the factory.

I have a set of barrels and heads from engine GE6 (R.A.E. glider, 1923) that have the same aluminized coating as Bob Jones' 750cc sprinter. These aluminized components, according to him, were left over from an earlier excursion into the light aero engine market. This would be long before the Sprite engines of the mid-thirties, of course. Mine have the traditional diamond groove, though they are 600cc and not 750cc, like Bob's.

Ever onward, we come to a cylinder head that I unequivocally know the date for, my 1934 OW1. I believe this same head was first introduced in 1931 on the F&G models. (Subject of Patent 357862 granted October 1931, but first submitted in November of 1930.)

Douglas motorcycle cylinder head
click image for larger view

The port geometry is exactly the same as the DT, except that the inside corners are left as machined. Though de-burred, no attempt has been made to blend the transition between the port and valve seat. Nor are any of the machine marks polished out, apparently as a production (and I mean that sarcastically!) model during very penurious times no extra hand work was indulged in. Or perhaps by 1934 they just gave up and stopped caring; whether they lavished the extra attention on the F&G models remains to be discovered.


The internal similarity is belied by its external appearance, which looks nothing like a DT at all. The detachable rocker perch is raised, and the rocker spindle holes spread to suit. As a result the valves need to be about 1/4" longer than a DT. On the road machine the inner half of the aluminum valve cover acts as a spacer (not shown in section view). There is an air gap between the port and the seating for the valve spring, in conjunction with the mica washer under the head of the valve guide, to thermally insulate the spring.
This might stop some of the heat from the exhaust port, but not the heat from the valve, which is shed through the guide.

One thing this head does have going for it is that it has by far the most support for the valve guide. Unlike the poor J5 head mentioned earlier! (As a side note. The 750cc Sprite aero engines seem to use the same head, the valve spring cups being cut away from the casting, or not cast in the first place, to leave the springs totally exposed. Since there are no valve covers to act as a spacer, thick washers maintain the proper perch height.)

There are no counterbores on the port side of the hole for the valve guide. The reason for these counterbores, when present, is a subject of lively debate between Phil Manzano and myself. That is when we are not arguing about gasket thickness! (Which by the way was not .005" nor .007" but .006"!)

One line of thought is that expansion as the head gets hot might apply an unequal, or side load, to the guide if the contact area of the press fit was not terminated with a square end. This could cause the valve to stick or effect its alignment with the valve seat. Another view, and one I favor, is that it is this way because Douglas reamed the holes for the valve guides to final size. To avoid deflection of the reamer you want it to cut evenly around the full circumference of its tip. Starting the reamer in the drilled hole at the bottom of a clearance counterbore, or running out the tip of the reamer into a counterbore (if one if is cutting from the opposite direction) would avoid a situation where the reamer is cutting on one side only.

Whatever the reasons, for the OW1 they decided they could dispense with this feature. One reason may be that the hole for the valve guide is longer, long enough to self support a reamer against deflecting from the side cutting load. This latter and less likely idea could only possibly work if the reamer were started from the valve spring side.

The lack of polishing in the ports for the OW1 allows one to see that the port geometry is simpler to generate than would first appear. Swirl marks from the cutter indicate that the valve throat and the curve beyond that constitutes the outside of the 'bend' were likely done with a simple counterboring tool ground to an elliptical form.
Douglas motorcycle cylinders
click image for larger view

This could have had a pilot that went down the preliminary valve guide hole, or perhaps it even drilled it (and that contentious counterbore when it is present!). Intersect that with a Ø1-1/16" hole on one side, and a 2 degree taper-reamed hole on the other for the exhaust, knock off the burrs and you are done! It could be done very quickly all by a heavy drill press and some fixtures.

The economies did not end there. Measurements show they used the same tool for both the inlet and exhaust valves. The tool was shaped to be a perfect match for the slightly larger exhaust port. The smaller inlet port made due with running out into this larger diameter. At least on the DT the transition was blended a bit to ease the abrupt change in cross section.

Still, when it goes, the OW1 goes like hell. So it is not much of a handicap; something to be said I guess for the theory of port turbulence. Also note that the 600cc OW1 and the 500cc DT share the same Ø1-1/16" inlet port, no allowance being made for the difference in capacity. But then the applications are different, and I have not yet seen a 600cc DT style head or a 500cc variant F to OW style head in comparison, to see if they did bother to match port size to capacity.

Another dimension that is plotted in the cross sections is the offset distance between the valve guide axis and the rocker spindle. The slight difference between the OC head and that of J5 is not surprising as they do not share the same style valve rocker arm. The OC had rockers with the valve lash adjuster built into the end of the rocker which bear directly on the valve stem.

Unlike the later o.h.v. engines where the adjuster is situated at the tappet stem where it protrudes from the crankcase. Having the adjuster on the end of the rocker is an appalling design. The spherical head it required only aggravated wear between it and the flat end of the valve stem. What is odd is the valve to rocker offset change on subsequent engines. This is because J5 is fitted with the same rockers as the DT and OW1 heads. Forging numbers on the stampings are 8530 and 8531, unless some speed enthusiast has polished the numbers off yours! (Oddly, they have a different part number in the DT spares list, but they neglected to update the stamping die!)

Which just goes to show these things are not as critical as one might like to think. As long as it is in the general vicinity to clout the valve without imposing too much side thrust, it will work. J5 does have a 1/4" radius ground on the pad of the rocker where it contacts the valve stem verses 3/8" for the later rockers, and as a result the length of the arm is ever so slightly shorter, as befits the offset. The rockers for the DT and OW1 are identical.

I would be pleased to have the opportunity to examine any other pre-war o.h.v. heads to fill in the missing gaps. Though I would not turn my nose up at examining the interesting K and M types too, they are a little outside the scope of the above research.

© 2002 D. Kephart



Free counters
provided by

All the content of the website is copyright 2000-2007 by except where otherwise indicated.

All rights reserved. No reproduction, retransmission or distribution in any form of any of the content is allowed without written permission.