Douglas - For Sale Items

Douglas 1913 Model P motorcycle

Douglas - Wanted Items

Douglas 1915 3 Spd-Gearbox and Clutch

Support & How to Join

Side Panel Notices

Quick Comments

Reminders, links to interesting topics, bump your own topic, quick comments or any short message of interest to members. Try it out!
Please note - 500 character current limit


Yesterday at 00:09:46
Welcome to the new site!

Recommended viewing for a fast start...
 - Quick Tour of the Front Page
 - Quick Tour of the new Attachments

More info in my post here.

FT35 conversion?

Started by mark1010, 19 Mar 2017 at 14:41

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.



I don't know if anyone can give me any advice on converting an FT35 motor into a viable heart for a lightweight motorcycle.
I acquired the FT35 motor from a genset a couple of years ago. I don't have any of the genset parts, just the motor and having now got it running  I'm thinking I would like to wrap some tube around it and build a lightweight sprint style bike.

I've heard that not many of the parts are interchangeable with bike motors, but as I have a working motor I am more interested in just getting it to a point where it will rev and perform happily enough in a motorcycle.

The things I've been thinking about are: swapping out the genset carb for a couple of amals to allow smooth revving; upgrading valve springs as I have been told the ft35 had softer springs as it wasn't expected to rev?; fabbing some kind of manual advance/retard....

I'm sure that I'm not the first person to have thought of doing this - it's such a nice little motor,  is there anyone out there who has experience of this or any advice they can offer me on the items I've mentioned or anything else I may have overlooked?

many thanks


Hi Mark

I am in the final stages of a Douglas T35 (Generator) motor restoration, converted, with a modest power upgrade into a rather more scary role powering a vintage aircraft (think "Those Magnificent Men"). On the way, we have discovered a few things about what the generator set and the T36 bike had in common.
For various reasons connected with bottle, I feel unable to put too much trust in a T35 in this role, but my friend Julian has the balls to try. The Douglas is from a generator set, with a modest hike in power, and a chain reduction drive to the propellor.

I won't bore you with the rebuild saga, other than to mention that the only available manual, useful as it is, was not quite up to what we needed. We had to discover quite a lot the hard way, including things like re-manufacturing oil splashers that double as nut locks, and copper head gaskets, and rebuilding a magneto from scratch.

The main parameters for the T35 bike engine and the T35 Stationary engine appear identical.
The key clue is from a 1996 article about the T35 [ 2 pages attached] where we discover the innards of the bike version seems near identical to the stationary engine.
The differences seem to be in the shape of the bottom of the crankcase, to make it "bike friendly", and minor changes to the cooling fins.

The use of two carburettors is an obvious difference, and that alone is most of what  it takes to hike the power to T35 bike level, which is not saying much, but it was significant.
We know there was a change to the feeble springs, and Julian has one head with shorter stiffer springs which I will be investigating soon.
If the springs got stronger, then maybe also did the push-rods, but that is only a maybe. Douglas were not into changing stuff if they could re-use current production.
There was a small change to the compression ratio from 6.5to1 up to 7.5to1. I don't yet know exactly what dimension changed to achieve that.

We have, of course, now gathered a lot of information, and have lots of photographs. Probably it will all eventually end up on this site, but if any of it seems relevant to your cause right now, then do come back.


Would love to hear more about this project Graham,



Quote from: nickh on 19 Mar 2017 at 22:27
Would love to hear more about this project Graham,
Hi Nick
By all means. It would probably go down better if I continued to put pictures and description into a final PDF, and in the end, that must happen.  There is no better place to leave such an "extended manual" than on this site. What we have is the best we could do with the information we had, so there is bound to be stuff which others here could add to, or correct, or just show us a better way to do things. The "diary" PDF has some pictures and descriptions covering perforated rocker cover welding magic, cylinder work, magneto reconstruction and various stuff up to about halfway through the work.

Keeping to the original point for mark1010, this was about how much of the Douglas T35 stationary engine information was also directly applicable to the T35 bike. Perhaps some T35 bike owners, seeing this post, can confirm what seems good for the bike also, and maybe point out differences. Everything to do with the gearbox, frame, and all parts of the bike beyond the crankcase are not really things we can help on, and there are probably places elsewhere in the forum where T35 info has been posted in the past.

This engine rebuild has taken me into many deep and dark Douglas places. Douglas had an order for more than 25,000 T35 engines, not including motorcycles.  I think the wartime dispatch riders were given BSA, Triumph, and their favourite was Norton, but there were some Douglas also. The cylinders and heads were of a high grade nickel iron, though still prone to rust. The quality of the castings could have been better, but it was wartime, and they were in a hurry. Post war build quality had a problem, a curious mix of near military spec. and excellence, and stuff that should not have been allowed. The bike did look really good, but was under powered. The Dragonfly, (also under powered) was a sensation when shown off, by Douglas did not have the resources to make a good go of it.

We discover that the valves are made of the same special valve steel used for the Rolls-Royce Merlin. KE965 special Austenitic nickel-chrome steel developed by Kayser, Ellison & Co. Ltd. in Carlisle does not air harden under any conditions!
KE965's maximum tensile stress is 17.2 tons per while it is glowing hot at 900C. That would be bright orange-yellow, and it rings if you whack it. 5 hours glowing somewhat brighter than that at 1000C in exhaust gases still shows no loss of weight. It just does not distort. It is also not made any more. I guess technology moves on, and valve steels are even better now.

Interesting then how one finds the heat, and the constant smacking turns some of the austenite to martinsite in the poppets of the exhaust valves, with only the top of the stems also showing some slight magnetising properties. This does not happen to the inlet valves. This property is not to be confused with valves that have magnetic stems made of different material friction-welded onto stainless poppets.

The KLG spark plugs also, are shared with the RR Merlin. The ignition leads and plugs are fully shielded, to stop interference to radio kit that the machine was undoubtedly powering at times. The Magneto is BTH  MD2 from  British Thomson-Houston of Birmingham.  The questions are.. are the valves in the T35 bike the very same? Is the bike magneto still the the same? I mention this because lots that I found in the magneto is military build to AEI Aviation specification. (I found the overhaul manual)!

I had an abortive adventure trying to re-use an annealed copper head gasket with my 0.006" tiny copper wire O-ring stuck on it until it became clear the sealing trick uses the machined fillet on the cylinder casting spigot to distort the crisp 90 degree edge of a new copper gasket into a perfect seal. It only happens if the hole in the gasket has exactly the right clearance. The trials of trying to home-brew re-manufacture decent gaskets are a separate story!

Perhaps unwisely, I put the part diary PDF on the site here in this post.  it is on the understanding that it is a draft, and is going to change!  It was a sort of quick review snapshot of progress at the time. A better pdf would include disassembly and reassembly help, reflecting our experiences, and a load of good engine data. Very likely, content may have to change because of new input from the folk here. I include the a few pictures of gasket making, diamond lapping valve followers without going through the hardening layer, and some other novel bits. Forgive that I am hung up on gaskets. Those exhaust ones are some kind of black high temperature stuff between layers of tough stainless steel. They are darn difficult to cut accurately, (and they cut you back)! In the end, the holes were made to fit the ports exactly by using a Dremmel grinder.

Do not be confused by the color codings on some nuts. Everything is BSF, and some are high tensile.  The red lines are marked across the ends of any bolt or fixing that has been properly completed with thread-lock in. There are, of course, lots more pictures and info, but for now, you see a bit of where we are.


Very nice work Graham - so this engine is being refurbished for a Flying Flea and there is already one in use on the replica Grahame-
White Flyer shown in the youtube video?

FT350 turn up quite regularly in various states of decrepitude at stationary engine sales for not much money.  I have the remains of one which I thought to restore but on further investigation the condition suggested it had been carefully stored in the village pond for 50 years!



Yes - the one you see is for the Henri Mignot HM14 Le Pou du Ciel. I agreed to try and make it run again, but it would only be for taxiing around when it is being shown, because that particular aircraft is the original Mignot design from before he modified it to make it safe. Now, there are hundreds of these, and their modern variants, and these fly around perfectly safely, but this widow-maker would not be one of them! I suppose I  could make the changes to have it be a safe version, but nobody here wants to try. Getting it certified to fly would be expensive, and finding anyone willing to actually get it off the ground -  maybe a bit challenging!

A Douglas T35 was apparently all it needed to get airborne, and it was reportedly docile, and very easy to fly. The nasty feature was that under some conditions, putting the stick forward to descend  could leave it in an unrecoverable dive where pulling back on the stick either did not work, or made things worse. Know that guys like me can't even read about such stuff without feeling queasey, let alone write good descriptions!

The White Flyer, which looks to me like a normal aircraft simply attempting to fly in reverse, used up one of Julian's several engines. He managed to get it to take off, though not very high, and not for very long. His interest has moved on to his 1916 White Sport Monoplane, also using a Douglas T35. I have a picture of it from before the frame was covered. It now has a chain reduction gear to a large propeller, and the engine mounting arrangements have changed.

I am aware that we have somewhat usurped mark1010's thread, but I include a picture of a Douglas T35.
Surely one of the most beautiful examples, but the interest is in the magneto arrangements. They seem set into the same place as the stationary engine, but have much more elaborate arrangement. I guess this is to make advance-retard available, which is something that a fixed speed stationary engine would not need.

The stationary engine has a set adjustment to rotate the ignition points relative to the magneto, but this is just to set the points to open at the optimum phase to get the most stored magnetic energy. Actual advance-retard for ignition requires to rotate the entire magneto, or somehow otherwise alter the point it delivers sparks.
This is a consideration if one is trying an FT35 conversion to bike use. One needs to figure out some kind of advance-retard.


The magneto on the FT35 has the same advance/retard mechanism as any other BTH motorcycle magneto. The difference is that the camring is set by having the return spring under the plunger and the adjusting screw pushing against it. If you want the timing to be easily adjustable from a lever, invert the plunger and spring, then remove the standard adjuster screw and replace it with the normal BTH cable adjuster - a cable can then be fitted in the normal way.


Quote from: graham-xrfx on 20 Mar 2017 at 18:06 (snip)
I am aware that we have somewhat usurped mark1010's thread,
Yes, but I think the fact that these units are capable of being re-worked for use in replica vintage aircraft will give him some hope that his motorcycle project is at least feasible.



Wow, this all got really exciting all of a sudden!
Thanks for your replies!
Graham, a flying machine is further down my list of things to do - Lilienthal has a lot to answer for! - I saw a flying flea at the Solway Aviation museum a few years back - I think that one had a Citroen? motor - certainly didn't have Douglas on the rockers,  but I looked into it and found the Douglas connection. Very cool! I've flown in Tiger Moth and loved it! But right now I gotto do the stuff on the bench in front of me - a Douglas FT35 running and destined to roll and a Rolls Royce B80 in more pieces than I want to think about.
So it's the Douglas then....
Thanks Graham for your input, I would be interested to know more about the valve springs - sizes etc. any ideas where I could source standard springs for a T35, and if anyone knows whether they will just slip in there nice and easy in a kind of 'too good to be true' way...
Also do you know what carb set up Julian is running on his motors?
Eddie - thanks for your comments on the mag - I have read a little and talked to some folks and I was led to believe that there was no advance retard on them - I will take a closer look now that you have pointed me in the right direction - thanks.
Thanks again guys!
If you have any more thoughts, please post them! - I'm sure that I will be posting more questions!
I really do appreciate it!
Cheers, m


         To uprate the valve springs, all you need are springs and caps from the early T35 motorcycle (don't try to use springs from the later Mk3 - 5 machines - they are bigger). Uprating the springs will give you higher revs, but to get more power , you will need to look at the cams. Unfortunately, the generator engine is set up with the inlets at the front, whereas the motorcycle engine has the inlets at the back, so a cam swap is not so easy - unless you also swap the heads side for side, putting the carbs on the flywheel side (assuming you go ahead with fitting twin Amals). You will also find that the heads will stand quite a bit of porting, so you could end up with a mind boggling 18 BHP!!!. Also, when setting up the engine, gear it a little on the low side - when put together correctly, the Douglas engine will happily 'rev it's nuts off' all day!


Thanks Eddie - I much appreciate the information.
I had another look at the contact breaker housing on the mag, and it appears to be completely free of any kind of tube or fitment I would normally associate with the advance retard mechanism on BTH mags, and searching the web for further info on this matter I found a post on the Scott owners club forum which refers to a book - the 'Radco' book "The Vintage Motorcyclist's Workshop"by Frank Farrington in which it is apparently stated:  "During the last war BTH turned out some very nice single and twin-cylinder stationary engine magnetos with fixed ignition. These crop up brand new from time to time. The lack of advance-and-retard facilities can be overcome by fitting a vintage BTH contact breaker housing, which will usually go straight on without further modification". This would fit with what I can see on the mag. I'm going to try Dave Lindsley to see if they have anything I could use, but if you or anyone else knows of supply options for BTH contact breaker housings, and early T35 valve springs (and perhaps cams?) I would be grateful to hear.
Thanks again for all of you help. I am truly grateful.


Quote from: mark1010 on 24 Mar 2017 at 12:12
The lack of advance-and-retard facilities can be overcome by fitting a vintage BTH contact breaker housing, which will usually go straight on without further modification". This would fit with what I can see on the mag.
Hi Mark
Do not mistake a vintage BTH adjuster for a proper bike-revving-range advance-retard. In the picture of the (un-restrored) BTH type MD2, you can see the settable adjuster with lock-ring that can rotate the entire internal points cam relative to the housing. This does indeed alter the ignition timing, and may be all you need if you could contrive  revs-sensitive way of adjusting it. Manual adjust by cable was onve used.
The real way magnetos work is by storing the magnetic energy from a rising current, and then forcibly interrupting that current at it's peak.
In a conventional coil ignition, the current rises,  limited only during the rise transient by the coil primary inductance, and ultimately by the coil DC resistance.
In a magneto, it generates it's own current, the rise being not quite sinusoidal, much decided by the laminations shape and the magnet. It is an alternating current, and has a definite peak. One tries to catch that peak, with the adjustment. Opening the points at this "sweet spot" interrupts the current. The laws of physics take over. Instant zero current would force instant infinite voltage to keep the current going. The reverse voltage generated is of course, not infinite, (150V to 400V or so), but it finds the handy capacitor across the points to dump the energy into, now storing it as electric field. My point is, there is only one sweet spot. This is  OK to set up for constant speed engines, adjustable for advance a bit only at the expense of spark quality.

If the current can rise high enough, quick enough, it can reach a state where there is maybe enough available "dwell", to allow a range over which one might mess with the timing, to no disadvantage, then rotating the cam a bit is OK . In reality, the magnets were a bit feeble, and when it came to higher revving engines with  more cylinders and multi-pole magnetos, they often resorted to multi-magnetos, and made mechanisms to rotate the entire magneto "housing within a housing". Possibly the mechanism for T35 bike Eddie mentioned is like that. I have a magneto rebuild article that clearly shows a cable operated magneto adjuster. Not a great way, but at least it suggests that some timing alteration could be had without the spark failing.

Since you are perhaps more interested in a T35 engine conversion into some sort of new bike, you may be less concerned with true vintage authenticity. Replacing the ignition with a electronic one, triggered by a low cost inductive probe sensing a drilling or something on the flywheel, or maybe something pointed into the magneto hole, may be the way. Speed-sensitive advance-retard might be provided with (say) one of those Arduino programmable things. There are LOTs of project articles, on how to do it, and nearly all the software is open source.
Or.. you can stay with magneto, and try the cable operated way.


Yes, the best place for the points to open is the point of maximum rate-of-change of magnetic flux, and this is always set (when the magneto is built) to correspond to the maximum advance setting. This is  the "internal timing" of the magneto, and is often missing in "swap meet specials" assembled from mis-matched parts.
When fitting the magneto to the bike it is our job to time the magneto so that maximum advance (and hence best spark) corresponds to normal high speed running of the motor.
But in practice magnetos - particularly those for single cylinder and flat twin - deliver a quite decent spark when retarded say 15-20 degrees (30-40 degrees of crank angle). More-or-less all motorcycle magnetos work like this. A tiny fraction of vintage car magnetos (MEA magnetos come to mind) actually rotate the entire magneto mechanism and so give the maximum spark over the entire advance/retard range, but I can't think of a common bike magneto arranged like this.
By the way, because of cunning shaping of the armature and magnet pole pieces, the variation of the flux is not sinusoidal, but instead very sharply peaked at the maximum flux points. You can feel this if you try to rotate a really good magneto with your fingers.