Saturday, 16 March 2019

The next version of the Mk.7 exhaust

This is the third revision of my Mk.7 exhaust design, effectively making it Mk.7.3. Now the original Mk.7 prototype served me very well over the last or maybe even the last two seasons, but despite my best intentions, when building it, it still left some to be desired. First, I knew I could improve on the build quality thanks to a new welder. Second, I knew I could improve on the fit and finish thanks mostly to a much improved welding table, which actually kept the parts flat during welding. And then there was the issue of the old downpipe on the left touching the chain tube and melting it.

So when setting everything up, the clearance between pipe and chain gaitor was deemed essential. (That's about 10-15mm, measured at max. swing of the swingarm.)


But starting from the front, in this case the cylinder head, I wanted to have nice, meaty flanges and location rings to center the downpipes in the port and ensure good seal, as they wouldn't warp easily. I also turned a step into the rings to make it easier to weld them onto the pipe.



A great deal of time was spent on making these two collectors and unlike in previous versions making sure that all the sleeves are pointing the right way in order to prevent leakages.


As I had started from the silencers towards the collectors, the next logical step was to work back from the front in direction of the collector.


Because the tubes had lots of bending marks, a bit of polishing was due. The thing with polishing is though, once you start with polishing at one end, you have to do the whole lot or it'll look a bit odd.


With the two Y-pieces of the collector already being built, the actual collector almost built itself as it was more a pretty simple case of connecting the dots.



The last part that I had to build was the rear downpipe, going to the collector. Building it imposed two main challenges: 1) clear everything, go through the frame and allow suspension movement and 2) be of as close as possible of the same length as the front downpipe. Which if you have a careful look at the bike is quite the challenge in its own right. Suffice to say, making the rear downpipe almost took as long as all the remaining parts.

This is the go-through-the-frame-without-touching-anything-part.



And once it has come out of the frame it "just" had to clear the collector and other pipework and meet up with the sleeve.



If you look very closely at the last picture, you will notice a section, which in reality is about 25 to 30mm long which looks a lot like an afterthought. This was the hardest part to fit on the whole exhaust as it had to compensate for the total sum of all misalignments of all previous cuts.

The last step was to weld in a bung for an O2-sensor (even though I currently do not have any intentions of using one) into the main bend of the collector. The angles and positioning of these are rather sensitive, e.g. the nozzle of the sensor has to point downwards to prevent it from collecting moisture and it also has to be as close as possible (but at least 50cm from the exhaust valve) and there's should be no exhaust leaks ahead of it. If you want to follow all of these requirements, you aren't really left with too many positions to put it and as I didn't want to install two sensors, there was really only one position left.



And then it was ultimately the last time to take the exhaust apart and polish up all the parts.




And this is the final result (admittedly the silencers aren't done yet).



To be fair, there's been a lot of mishaps, as originally this was meant as a completely new generation of my exhausts, I had to redo the sealant on the cylinder foot, when swapping valve springs. Oh and then there was this little episode, where I woke up, because I was absolutely certain that I had mixed up the valve timing on the rear cylinder ...

Sunday, 17 February 2019

The XS Triple Sidecar - new needles and a complete re-jet

The XS-Triple Sidecar is my daily mule during the cold, snowy and salty months. As such it has endured a lot this Winter and to be honest, even after more than one attempt the jetting of the flatslides still wasn't fully satisfactory.

Most importantly fuel consumption was heavy and it has stopped to idle. After doing quite a bit of research, I found out that the 6DP1-needles out of my VM38 carbs are the same length and should give me the much needed richer mid- and top-range.


Finding out  why it had decided not to idle properly anymore became rather apparent, once I took the old pilot jets out. Turns out there's two sizes of idle jets for Mikuni TM-carbs. 


It's very hard to see in the picture, but the needles (6DP1 on the left, 6DGM5 on the right) are actually quite dramatically different. From the midrange to the top, the 6DP1 is much slimmer and thus allows for more fuel and air to flow.


And that becomes VERY apparent when rejetting. Before I was running the bike on 37.5 pilots and 135 mains and it was running rich on the top, now I have upgraded to 45 pilots and 145 mains and when you read this, I will probably be trying out some 147.5s or 150ies. Suffice to say, the power available. grew in the same fashion, so I can't complain. Also the bike is a lot torquier now, which is a good thing as it is geared way too long for sidecar use from the factory. (Hint: it was never meant to pull a chair...)


Other than that, Spring will bring a new exhaust and a modified rear subframe, so I can finally swap rear tyres, without taking the exhaust off, which to be fair is a bit lame. 

Friday, 8 February 2019

Syncing vacuum gauges

I have a suspicion that I know, what you're thinking... shouldn't that read as "syncing carbs/throttle bodies WITH vacuum gauges"? Well, not if you have had a little backfire through the carbs and that properly messed up your gauges.

Now here is the moment, when you realise that you bought tools from a reputable manufacturer, if you can simply buy spares (in this case a new set of clcoks 15 years after you bought them) and they'll basically overnight them to you and respond to your emails at absolutely ungodly times of the day, i.e. late in the evening.

Let's have a little look at this picture, do you see what has gone wrong?


There are copper-beryllium springs inside these gauges, which are very sensitive to even the slightest changes of vacuum (or pressure). Unfortunately these don't take very kindly to being overstretched in either direction. A good indicator is for example #4 which doesn't sit at 0 anymore.

After parting with quite a handful of my hard-earned (definitely more than for a cheap set!), I received four new clocks which had (slightly) adjustable scales.


As you can see with the attached y-piece, in order to sync them, two (or more) gauges have to be hooked up to one vacuum port and then adjusted to show exactly same value.


Which in the end looks something like this and once again proves that sidecars absolutely rule as mobile workbenches. 😀

A quick word on the elephant in the room: Why didn't you just get a set of cheap... Simple really, you can only adjust carbs as accurately as your tools will allow you to measure pressure differences, so the more accurate, the better the job. To be honest, the level of accuracy that can be achieved with these is (usually) absolute overkill, but the other thing about cheap gauges is that you can't sync them to each other inducing another error aside from the already given inaccurracy aside from a less sensitive spring. Also: it's nicer to know you could absolutely nail it and then call it good enough than the other way round. 😏

Sunday, 3 February 2019

The new TR1 engine - squishing things together to make more power (part 31)

Aside from ruining perfectly good engines, there's some more that needs to be done in the advent of the upcoming season. The recent situation of detonation on one cylinder and the fact that the gap between squishband and piston was incredibly large led me to look for solutions to increase the effectiveness of the squish. And the best way to improve things was to get the piston and the squishband closer together by removing the cylinder base gasket.

First the baseline had to be established with some 2mm (super-soft) solder.


This came out as 1.71 to 1.75mm, which was just at what literature would consider the uppermost limit of squish-gap to still be working to a minimal extent.  Time to pull off the cylinder head and cylinder have a look at things.



Quite a few carbon deposits and signs of the engine running a tad richer than I would have hoped, but aside from that all seems to be well. After quite a bit of swearing because I haven't used enough grease on the base gaskets and lots of subsequent cleaning a thin bead of gasket sealer was applied and the engine re-assembled.



So which results did the whole process yield: The base gasket is 0.50mm thick when compressed and that's pretty much what the cylinder came down. Everything spins over freely and nothing seems to touch.


Theory says one should strive for a clearance of 1.00mm or even slightly less for achieving maximum effectiveness of the squish, but this leaves out of consideration that after all this isn't a racing engine and shall still work reliably even 30,000 or 40,000km after this overhaul and without any doubt, tightening up the squishband should help quite a bit in the future.

Saturday, 26 January 2019

The Norwegian Job - etymology "Zweckoptimismus"

From the same people that brought you words like Bratwurst and Zeitgeist, here comes another example of linguistic craftyness: "Zweckoptimismus", which would literally translate as purpose-led optimism. Ach you gotta love "ze Germans".

Last time we left, when the engine was on the bench and the obvious next step is to put it in a frame and squeeze some happy sounds out of it. But in order to put something in said frame, a bit of taking stuff out of the same was due.



Only a few hours later and my workshop is a proper mess.


The reasosn for this taking so long were plentiful, amongst some others, new starter cables had to be made and some new connectors had to be crimped onto the wiring. (Have ever pointed out, that a nicely crimped connector can be a thing of beauty?)


The starter motor cable was crimped and soldered to ensure the best contact possible. 


... and as by magic, the new engine is in the frame and filled with oil.


During the course of action, I also finally found out, why there always was a bit of fine rust-powder in one of my carbs: the glue of the filter had decided to call it quits, as can also be seen by the little flake of glue at the end of the filter.




At this point, things went "slightly" sideways, as in the starter, which had worked just fine, before the engine was set aside... well it didn't anymore. Of course you work this out AFTER you fill the engine with oil.



Oh and as it turned out, that new starter had a slightly oversized shaft, which had to be polished with some emery paper to get it down by about 0.15mm to make the starter clutch slide on and off the shaft without force.


And guess what: Press the button and fire the old girl up. Just like that.


But the story doesn't end on the note of "just like that". What you see here is a spark plug that had a very unfortunate first date with bits of piston and valve.


And a bit of valve stem, which sports a slightly dramatic valve clearance setup.


Alright, so we're dealing with a bent valve. No big deal, right?


Uh oh. Bits of piston stuck inside the carburettor.


Now those engine bits in the inlet are less of a surprise already, if I am being honest.


No point in not digging into the engine and taking of the offending head.


The way the valve stem is fractured is interesting. This very coarse grainy structure leads me to belive, that it had broken away? Also the stem is incredibly brittle, so there might have been some heat involved?


Now, I was still looking for two valve-dishes...


Remarkably, the piston held up nicely, regardless of a huge, gaping hole in the middle, the skirt's virtually undamaged. 


Quite a bit of said piston luckily made it out through the exhaust.


And now to a practical example of what "Zweckoptimismus" is: Looks like the conrod and crankshaft survived, unless the other side with the primary drive gears took a hit.


The cylinder (some actual measurements pending) could be good enough for a 3rd oversize, i.e. +1.00mm bore. As it has taken some damage, but all the dings don't seem to be very deep, there might be a chance of getting away with it.

All in all, there's still a lot of investigation due and I suspect, it'll be best to put a blanket over this engine for a few days and then have another (more calm) look over it and assess the damage. Maybe at least some parts of the heads (aside from the covers, etc.) can be re-used.