Nanno's "mule" a TR1 for everyday


Some may or may not know this bike, but it has been part of my life for more than just a few years.

What you're looking at is a 1981 Yamaha XV1000 TR1, with approx. 113,000km, heavily modified 1063 engine, VM38 roundslide carbs, homemade stainless exhaust and quite a few suspension mods as in XS1100 forks and longer XT600 rearshock and quite some mods to the wiring harness, starter system and so on... Well, the kind of stuff you do, when an older bike is meant to be used DAILY.

A while back, I've built a little turbo-setup for Yamaha XVs to have a bit of dragracing fun. And it was pretty good. Unfortunately, I can't always have fun with turbos and in all fairness there's this bike in my workshop, which does HEAPS more miles every week, than the turbo has done in the last two years.

Now, my current engine is a classic "euro-style" TR1 engine build: XV1100 rotating assembly with 750 heads on a set of TR1 cases. This works to my absolute satisfaction and does the job just fine. Realistically it will push out a healthy 70 to 80 horsepower and tons of torque as low as 2500rpm. Putting 750 heads on a TR1/XV1100 engine pushes the compression up to a healthy 9.x, which spices things up quite a bit, but it comes at the price of breathability of the engine or in other words: The powerband ends at 6000rpm.

Originally, when I started this engine build (waaaaay to long ago) this was meant to be a hopped up version of above recipe, but with bigger valves installed in a set of XV700 heads:

I inspected the parts rather thoroughly:

and then basically built the engine until it was almost done.

At this I did a lot of measuring and realized two things, in order to make proper power:
1) the squish has to be tightened up a lot ( on the one hand to improve swirl and thus more power, but also as better squish means cooler combustion)
2) there's no way to fit the big inlet valves in the 700 or 750 heads, without turning them down at least one mm

After a long chat with Sepp Koch and picking his brains and trying to get the info out of him, because quite frankly in the last 35 years the man has tried pretty much everything except forced induction on the TR1 and just KNOWS. As the numbers for the squishband-clearance and (what I consider) sensible compression matched with what I had calculated earlier, I decided to weld up a set of TR1 heads to decrease the volume of the combustion chamber and shorten the cylinders accordingly.

Sounds like a straight-forward plan?

That's when life kicked in, in the form of another BT1100 engine. For those that don't know it: BT1100 is the sporty Buell-style bike, based around the V-star 1100 engine. Bigger fins, nikasil-lined cylinders, lighter rotating assembly(2.8kg lighter !!!) than
the old XV1100 may kill some low-down torque, but should make the ol' girl rev-up even faster and keep the rotating masses under control.

So that's where we stand at the moment. I aim at doing an update approx. once a week, but as I am also working on some other projects (there's a new turbo-build, I've got a new to me XT500 4valve and my XS-Triple sidecar wants some attention too, let alone stuff I do on my dad's and some friend's bikes.)


Gold Coast, Queensland
Re: Nanno's "mule" a TR1 for everyday

Awesome. I was watching the engine build on your blog. Those BT1100 engine sure have some nice covers on the top end.


Well, just trying to get rid of some other projects, so I finally have the amount of time (and if needs be) additional funds to really dive into this build. And yes, this is gonna be awesome. The goal is simple: genuine 90 (euro, not SAE) ponies at the rear is the name of the game. 8)


I am still "a tad" busier than I admittedly like to be, but at the very least I was able to squeeze-in a little tooling upgrade. One of the tools that I absolutely had to build was a decently-sized flycutter to skim the heads after welding up the combustion chambers.

For that reason, one is suppsoed to take an obscenely large (just shy of 5"/120mm) piece of (tool-)steel and chuck it in the lathe and square it up. Any old steel would have been sufficient, but this is what surfaced first in my stash.

Centre-drill it and find out that you grabbed a 17 instead of a 16mm drill...

... and turn a little centre-boss to shrink in.

And as you have to mill down the flycutter's arbor on at least the side where you want to tap it, you might as well balance it and mill the backside flat as well.

And this picture was actually taken BEFORE milling the backside flat, so it was completely out of balance and still, this was the kind of surface finish that was achieved in a single (interrupted) pass. Oh and that is roughly about the worst butter-grade ally that is known to man, so the finish on the much harder aluminium of the cylinder heads should yield much nicer results.

Effectively, now the finish is good enough, so you can still see the ridges, but hardly feel them with your fingernail. Which should definitely do for the job and give the headgasket something to bite in. Next step will be to make a fixture for the cylinder heads so, I can safely clamp them down when milling and also have repeatable results.



Now that this lovely flycutter is in the shop, it needs to be fed. So I grabbed an old XT500 head, which was really only good as a cooling-fin-donor and see if all I had thought up would actually work out in the real world.

The biggest issue with all of this is to get enough heat into the casting and secondly find a filler that will actually work satisfyingly.

Funnily enough, the 4047 (AlSi12) filler didn't work at all. It instantly crystalized and even if I did get it to work, well just look at the last picture. 5356 (AlMg5) was used and wetted out nicely and uniformly.

As can be seen, a lot of "peppering" occured, even though the cast had been brushed and because that didn't yield satisfactory results even ground a bit with a handheld endmill. I suspect, with the real heads, I will have one or two cleaning passes and then grind out what I just welded up, to get rid of most of the imperfections.

The flycutter on the other hand was an instant success. It doesn't care much for the welds sticking out and will take out 1mm just as it will do with 2mm (approx. 1/10"). The head has been decked 0.1mm (2 thou) just to see how close I can get it. It's hard to judge by the picture alone, but the finish is good enough for a headgasket.

Next step is to find the best way to reshape the parts that have been welded up to give the combustion chamber its final shape. Which in the case of the TR1.1 will simply be straight. So I suspect, the main part will be done in the mill and only the last bit will be touched up with a die grinder.



Well, the fixture is done and to be fair even the actual heads have already been welded up, but honestly the documentation is lagging behind a little compared to the actual work, which frankly is a good thing, because the other way round would be way more unpleasant.

So here goes: 4 60mm of 30mm diameter ally roundstock as those XV-heads are one-piece items and as such don't have a valve-cover-gasket surface, one could index to.

Turned down and drilled and tapped on both sides:

Now this is, where the tricky part comes in: How do you clamp something down, when you can't have a stud, nut, bolt-head protrude over the surface? Well, pretty much the same way a milling collet works, but pushing to the outside.

Torquing the countersounk bolts down results in the heads of the studs spreading and thereby safely clamping the head down.

And then in the end you need a baseplate to mount everything to:

Blog-Link (for more pictures and other posts):


der_nanno said:
Funnily enough, the 4047 (AlSi12) filler didn't work at all. It instantly crystalized and even if I did get it to work, well just look at the last picture. 5356 (AlMg5) was used and wetted out nicely and uniformly.
I agree, I always use 5% Magnesium rods with all my bike work and am very happy with them, much better than Silicon.


New Member
And here i was thinking that mg fillers were for mg alloys and si for si alloys.
I wouldnt try to weld up old british dtd424 using AlMg fillers..
Not really working that much with alloy welding, but isnt most japanese castings magnesium alloyed aluminium?


datadavid said:
Not really working that much with alloy welding, but isnt most japanese castings magnesium alloyed aluminium?
As a matter of fact it is and that's the big BUT, I was told by someone REALLY in the know of such matters, that I should try SI-filler, as he had great success on XT and SR cylinderheads and engine cases.

irk miller

You've been mostly-dead all day.
I wish they called these a TR1 here in the US. This is a cool build happening.


I have to admit though (and I hope not to disappoint), but I will actually build two engines. One as started with the heavy 12kg XV1100 crank and one with the much lighter BT1100 crank, as I really want to know the differences (and I have enough parts to build at least two engines, plus it will allow me to begin road-testing the first engine in the coming weeks...)


So why is progress so slow, all you have to do is mill away some ally from those heads...

Well, the everyday TR1, the mule, is my EVERYDAY vehicle and as such stuff breaks. This time it was actually two things: firstly the starter gave up.

Those pearls are where grease from the rear bushing mixed with carbon debris...

Greasy shmoo everywhere!

Luckily I had a spare starter sitting under the bench. Even though it was actually earmarked for the new engine. I quickly pulled it apart, as I have done with literally every starter in the past and look at the picture. That's the pin that is supposed to hold the planetary gear set in place.

The rest is fine and the shaft came greased from the factory.

As I didn't want to drain the oil, I took the actual starter motor off the reduction gearset and installed the new starter onto that. (Not quite by the text-book, but I want to go through this engine rather thoroughly, when it's out and this also means inspecting the oil...)

The other thing that had started bugging me since probably mid last season was some intermittent backfires and irregular cutting out over 6500RPM. I had always attributed that to my exhaust as it appeared pretty much simultaneously. So after replacing the starter and running a few errands it had actually started to get worse and even with new plugs the front cylinder would cut out more and more often. For various reasons I run coil-on-plug coils and as they usually come from four-cylinder-bikes, everytime I buy some, I get two spares. ;)

While I was in there, I decided to give my other ignitech box a try as I bought it two years ago and never ran it. :eek:ps: (It did just fine.) Because I was out there programming away anyway, I decided I try something and that was to have my ignition advance set to zero where I want the rev-limiter to be. Which now results in a super smooth red-line and none of the dreaded backfire revlimiter cut-outs anymore.



Those, who follow this on other channels know that machining the heads was finally on the agenda, so here goes: I started out with CC'ing the stock heads, as I wanted to have a base line. I had calculated the volume of the wedge, I was planning to weld up. But the figure I ended up with, was roughly double, what I ultimately ended up with, so in hindsight this was a very good idea as it would have caused a tremendous amount of head-scratching in the aftermath.

When looking at the pictures, you will notice both silver and black cylinder heads. Let's put it this way: Yes I have two sets of such heads and yes one will ultimately be for sale. Welding the heads up with MIG and ally wire is terribly messy, if you're used to mostly TIG-welding. But as this is done in DC you can get about twice the heat into the material and on top of that with one hundred percent "on-time". So my 110 to 120 Amp welds, most likely correspond to doing the same with 250 to 300 Amp TIG welds. I admit they would come out cleaner though. (The picture shows the head without any cleaning.)

For the very first head, I milled it back before fly-cutting, which caused some issues marking out the cutting line, so in the following heads, I skimmed them first, then marked them out and then milled the wedge to size.

And that's the same head after fly-cutting. The backlight is mercyless (for a reason), so you can see the actual tool-paths of the flycutter. They are between 0.10 and 0.15mm deep and can hardly be felt with your fingernail. I knocked them down with some scotchbrite pad and now the heads feel a touch rougher than stock factory items, which means the stock headgaskets should bite nicely into them and give a very good seal.

Once all of that was done again I cc'ed the heads again and guess my surprise, when I realised I ended up with the same 47-48cc as with 750 heads, meaning with an 1100 crank and standard thickness headgasket and basegasket the compression will be at 9.9:1 and with my shortened cylinders I'll end up at 10.5:1.

So why two set of heads: As you can see in one of the first pictures, these were rather rough to start with, so required a lot of skimming and as such could force me to recut-valve pockets in the pistons. Something, which I would like to avoid. With a stock setup you will have at least 0.5mm more clearance, so it should be nothing to worry about.

I've made some fixtures (inserts for the cylinder-liner) yesterday, to be able to machine the cylinders down to the correct length, so that will be one of the next tasks, that I plan to tackle. Plus some minor jobs of re-threading the engine case for the starter bolts (one was stripped out and the oversized hole plugged).

Blog-Link (there's more pictures on the blog, but I don't really want to write everything twice, every time...):


Before we get into the matter at hand, let me show you a result. Some people asked me, what kind of finish could be achieved with such a big and heavy flycutter on such a small and puny mill. A pretty shiny one, if I may say so.

But heads are only one part of the equation. The other thing is to get the squish-height right and to achieve that the cylinders need to be shortened to get the pistons to sit flush or slightly below. (Actual values when I've done it, which will be tomorrow.) So I picked up two aluminium (AlMgPbXX - soft ally, nothing fancy) discs and turned them to 0.05mm undersize to the bore and tie them together with an M8 allthread. This way I don't have to machine any weird fingers etc.

And that's what they look like, when installed:

And at this point, the world was still a happy place:

And here comes the catch: The cylinder will happily clear the lathe's bed, but it will actually touch the outrigger of the lathe-support. (As can be seen looming in the last picture.) Unfortunately the cross-slide's travel is insufficient to reach all the way to the cylinder's foot. Now there's three potential solutions: Go to someone else or a shop and have it turned down on a bigger lathe. Build a special toolholder with lots and lots of overhang, so it can reach the cylinder or lastly, finally finish the adapter for the rotary table to be able to slap that second hand soviet era lathe chuck on my rotary table and fly-cut the cylinder to the correct length. (Don't worry, I am not going to let someone else do my work for me and the only friend with a bigger lathe is about 200 mls away (one-way), so that's sort of out of question as well.)



Lots and lots of things have gone wrong lately AND now it's exam time at uni, so additionally I have now run a bit out of time. But every now and then I need a quick break, so I sneaked into the workshop to tackle two things.

Step one: Finally fix that last thread holding on the starter motor.

Then carry the engine over to the workbench and find out, whether the piston-to-head-clearance is really as abyssmal as it's propagated in various places. (Spoiler alert: yup.)

Plastilin (red is the fastest colour of course) rolled up and placed on top of the piston (not at TDC) and then put on the head and smash the whole lot against it, by rotating the crank. I could also have used solder, but I wasn't sure just HOW LARGE the gap would be and my thickest (soft) solder is 2mm.

Cut the plastilin back a bit and establish the height with a vernier caliper.

And you end up with 1.65mm of clearance, which indeed is waaaaaay to much to make the squishband effective at all. Literature says that at around 1.00mm you start to see some positive effects and (of course) the tighter you go the better the overall effect in terms of mixing up the gases and cooling down the combustion. In an ideal world this engine could be run with a squishband clearance of 0.80mm or maybe even tighter for race use, but as I foresee a massive lack of time in the coming years and also this isn't a racing engine, but my everyday transport, I will aim for 0.95mm, which will also accomodate for certain variations in head- and basegasket thickness, should I have to pull the cylinders, e.g. when the pistons have to go for a rebore.

Blog-Post (with a bit more text on how to fix up the threads for the starter motor):


Let's put it this way: six weeks. That's how long it took from first idea to the first actually successful testrun. And this is how it is supposed to work (tried on an already thrashed cylinder):

1) The cylinder is clamped between two T-shaped tophats and supported by a live center on the tailstock and in the foreground you can see what will later become the indexable boring bar during one of the first tests to see whether I could actually reach everything.

2)As my QCTP took up some precious cross-slide travel, I had to fabricate a new toolholder...

3) ... which in the end looked like this.

Now all of this looked very promising (see last post), but as soon as the boring pushed against the workpiece, the live-center was in no way capable of supporting the cylinder. A bit more thinking and at least two attempts to flycut the cylinders on the rotary table of my mill had me re-think the lathe setup. (I am sorry, I*ve always been a lathe lad and less of a milling machine man...)

So today in the morning, I modified the backplate, fitted an old 6001 ball bearing and modified an old hex-bit to perform as an axle and finally everything is clamped down properly.

I'll get back to the whole lot on Tuesday and will modified the chuck-sided tophat as currently it is slightly tapered, which is good for alignment, but even low force will make the workpiece shift between the tophats and cause it to tumble, even though this is already rather minimal and a slightly more hairy chested engineer would probably call it good enough.

There's a bit more on the fabrication of all the tools on the Blog:

Additionally, I'll probably post some of the more (in retrospect) amusing failures of my fixtures in the upcoming days on my blog.


Oh the usual... I bowl, I drive around...
Cool stuff! I'm going to be tackling the squish setup on my kz750 twin soon. Interesting to see how you've done it. My problem is the custom pistons I will use are an unknown entity (got them used). Not sure how they will effect compression, I need to find a way to measure them because they are positive deck height.


Easiest way is to cc the heads on the bench and then install them, put the piston on TDC and then fill the cavity with (thin-)oil out of a syringe and take that value down. E presto you have the correct compression ratio.

And well, in this case it was necessary as there are no (or rather no good) tuning pistons out there on the market as the relatively high domes dramatically affect the flow into and the swirl in the combustion chamber.


To my own disappointment my fixture to shorten the cylinders just didn't work. Ever so often the whole lot starts to shift and as a result oscillate. I have an idea on how to do this in the future, but for now I really want to finish this engine and actually enjoy what I've built over the last year.

One of the mistakes I made in the past was to mix up the front and rear sprockets, as I put them in as I found them in the engine (mind you I bought this engine partially disassembled) and as such the front and rear ones were mixed up. That being said, if you mark the teeth of the primary gear behind the rotor, you can very safely assemble everything. I wasn't entirely sure, so I pulled the roto and guess what: All tidy.

With those mistakes tidied up, it was well about time to get back to head work. As I had to fit new valves anyway, I only deemed it appropriate to make them look pretty and polish them. (Opinions are a bit split on that, but I noticed less carbon buildup on polished valves, which is basically good enough for me.) The valves were polished with 120, 220, 400, 800 grit sandpaper.

This is a relatively quick, but nonetheless messy, job in a lathe.

As the stock valve springs create mad seat pressures, I decided to go with the softer inner springs out of a Virago 1100 to prolong the lifes of both rockers and cams.

One of the things I really invested some time in was to blend the valve seats into the ports. The heads I had were exceptionally bad in this respect, as there was a massive lip where the valveseat was undercut.

And that's what they look like now:

I didn't take any picture of the new Viton valve stem seals, but in the past they have without a doubt been worth the money.

More pictures and a bit more on porting on the blog:

DTT Bike Of The Month Gallery

DTT Light or Dark
Top Bottom