Building expansion chambers for the 500 Suzuki

Kiwi Glen Morgan has subjected the process of building expansion chambers for the 500 Suzuki to close scrutiny. Read on ….

By Glen Morgan

In the process of working on my own T500 project I’ve met a few others with similar intentions. The exhaust system seems to have presented a problem for a few people. The wide spacing of the frame cradle tubes under the motor make it quite hard to simply attach chambers to the existing header pipes and go racing.

Long slim chambers will work and some people have gone down this road. The orginal exhausts are a kind of long slim chamber with silencers built in. They work quite well, as the performance of the standard road bikes testify.

Standard, welded up and hydraulically formed T500 pipes

The main advantage to making your own chambers similar to the stock “mufflers” will probably be the weight saving. One of the problems, however, will be that your chambers will scrape on the ground just as the stock items do. Take heart though, they will not be any worse than standard pipes and probably better if you tuck them up well.

Having said that, if you are after more mumbo you are going to need chambers with a bit of volume to offset the peakiness that power improving changes to the exhaust ports can bring.

In most cases, a volume increase either means longer pipes or fatter pipes. Longer pipes give peak power at lower revs and they may stick out behind the bike to an unacceptable degree (even on a bike as long as a T500). There are structural problems associated with this, and if you shorten the bike in search of quicker handling it all gets a bit awkward.

Properly designed, shorter, fatter pipes are better if you are searching for more get up and go. However, it has been said that you will know when an expansion chamber design is right because it will drag on the ground or burn your leg! The truth of this was never more evident than it is in the case of the T500. Mount a chamber of reasonable diameter conventionally on a T500 and you will almost be able to park it in the pits by leaning it over on one of its chambers! This is not conducive to cutting a swathe through the swervery, leaving lesser mortals falling off their bikes in admiration.


Nope! You’re going to have to convince the bloody things to nestle up under the engine, between the cradle tubes. This takes a bit of bending of header pipes, but it can be done without too many tears. However, you then have to induce the chambers to diverge again so that there is room for the back wheel between them. This is particularly so if you have shortened the swing arm or fitted one from a T250/350 ( see email from me to Murray somewhere on this site, about this subject). SEE HERE ON READER’S WRITE

Ooops! The back wheel don’t fit

Bending chambers every which way takes some technical savvy and a lot of recourse to the old saying: “Think twice – cut once!” – only I find that I have to think several times. Creating your first chamber all by yourself is quite an achievement and stuffing the thing up is not what you want to do.

If it doesn’t worry you aesthetically, you can extend long header pipes forward, then back under the bike. This will move the chambers sufficiently far forward that the divergence of the baffle cones may give you enough clearance at the rear wheel, or at least minimise the amount of bending required. If you haven’t radically shortened the swing arm, this could be the way to go.

Keep in mind that long header pipes change the power characteristics produced and that the chambers will need to be correspondingly shorter if your chosen tuned length is to be preserved. A long header pipe is not necessarily a bad thing. It can help spread the power delivery of your bike, though it may dampen peak power unacceptably if you over do it.

There are also some problems with conflicting resonances if components of a tuned two stroke exhaust system fall outside certain limits. This article is more concerned with practice than theory, and if you want to know how to go about designing your chambers, get a book like Gordon Jennings’ Two Stroke Tuner’s Handbook or the later Two Stroke Performance Tuning, by Graham Bell.

Bell’s book is a sort of practical up-date of Jennings’ book, without as much theory. Both are excellent and I would recommend Bell to those who just want to do it, not understand it in great depth.

Read Jennings for interest, in fact get your hands on anything written by Jennings if you want to really expand your knowledge.

The other alternative is to use the chamber dimensions that Murray has made available on his T500 Fanatics web site.

You should be trying to achieve performance which is within your technical and riding expertise or your pocket. You will want your expansion chambers to give you power that you can use.

Theoretically you should be able to get over to 45 degrees with the suspension fully bottomed out, without dragging bits of the bike on the ground. 50 degrees may be possible (transiently). In practice, on the tyres we are allowed to use in post classic racing, scraping the pipe could be a timely reminder that we are about to get bits of our anatomy bounced up the road in loose formation.

The picture shows my T500 with the fork springs removed, the stanchions pushed through the triple clamps, the rear shocks replaced by wooden struts and the tyres flattened. This was done to see whether the chambers were giving enough clearance to achieve 45 degrees. They do – just; but I have room to raise the chambers another half inch and I probably will.

Right! Before we get into a bush mechanic’s guide to cobbling up some pipes, a comment about stuff on the net, and I don’t mean the ladies and gentlemen in exotic positions.

There is a lot of excellent stuff written by people who are sharing their knowledge in the spirit which makes motorcycling a fun thing. There is also some stuff on chambers that has me scratching my head. Still, there is more than one way to skin a cat, and there are doubtless other, better ways to do things than those I have suggested.

I am going to assume that you have minimal expertise, cash and equipment. You will need things like tinsnips, pliers, hammer, hacksaw, a firmly mounted bench vice, some scrap water pipe, a bit of patience and access to a gas, TIG or MIG welder.

THE HEADERS (also called the “lead-in” pipes)

In my experience, you can waste good money getting muffler shops to bend up header pipes for you. Unless they have a mandrel bender, the pipes will vary in section. Pressure waves don’t like variations in section, and pressure waves are mainly what tuned exhaust systems are about!

There are still some real tradesmen around who can produce sand bends that look like a pretzel without mucking up the cross section of the pipe. These people are often in retirement and just tickled pink to help a young fella out, if you can find them.

I suggest that you use “donuts” and pieces of strait pipe of the same section to create your T500 header pipes. A donut is a continuous circle of pipe which comes in various diametres and bend radii. They are available from muffler shops and they solve a lot of problems that beset me when I first started tinkering with exhaust systems.


The muffler shop person may not know the internal diameter of a pipe or donut. They generally work in external measurements.

Original header pipes are 1.75 inches internal diameter.

**If you don’t have calipers for measuring, take a header pipe into the muffler shop with you and match it up to the ends of their pipe stock until you get a close external match (better slightly bigger than slightly smaller).

**Next compare the wall thickness. If you don’t have anything to measure it with, just eyeball it. If your “eyeometer” can’t detect much difference, it won’t be significant for what you are doing.

**Now order a donut with the same external diameter. (donuts can be quite expensive in this country – about $45, unless people are kind and give you a trade discount) You should only need one donut to make two T500 pipes. [“TTCO” – think twice, cut once]

Choosing the radius can boggle the mind a bit, but I wouldn’t get too worried about this. Muffler shops that carry a range of donuts in stock will usually exchange them for a different size ( if you have’nt started hacking into them).

** Take the donut home and hold it up by the exhaust port. You will be able to see whether the radius is going to work out pretty easily. Basically you need to be able to clear the down tubes of the frame as you cross over them toward the centreline.

When planning your headers, you also need to make sure that you don’t take them so far forward that you are in danger of touching the front wheel or mudguard (fender if you speak American).


Remember that the wheel moves back toward the frame when the front suspension is in compression and that braking forces can further deflect the wheel toward the frame.

This is particularly important if you have gone the whole hog, and steepened the steering head angle to induce quicker steering.

I am assuming that you are going to stick with the original finned clamps to attach you headers, but if you decide to go the slipper joint way, have a look around for larger capacity PE or RM Suzuki exhaust flanges with slipper joints. I have an assortment of these motors in my storage shed and some of the slipper joints are just the right size to adapt.

Now comes a slight moral dilemma (for me anyhow). I make my own flange where the header enters the exhaust port because I don’t like cutting up original headers that some restorer of T500s may need. I suggest that you be not so fastidious. You can hacksaw this bit off and use it. (But not yet! “TTCO!”)

So, you are going to need this bit with the flange on it, possibly a bit of straight pipe, then a donut bend, then probably a bit of straight pipe, then a donut bend and possibly another bit of straight pipe.


BUT AT THIS STAGE DON’T CUT ANYTHING (“TTCO!”). Leave the headers for now!


(resonators or whatever you prefer)To make chambers, you need cones. You can make your own cones, or you can get them made. If you get them made, remember that all tin benders are not created equal. Most are good, but some may not understand your requirement for accuracy. To work well the cones must be round in section and must diverge at the specified angle.

As this is a twin cylinder machine, they must also both meet the specification pretty closely. If they are more than a little bit different, the chambers will perform differently and one cylinder may be using up power to pull the other one along with it at some or all stages of the power delivery profile of the engine.

You can check the accuracy of the work by making a triangle of the specified divergence angle out of something like plywood and shoving it firmly inside the finished cone. This will show both ovality and conformity to your specifications quite clearly. But do make your test piece accurately.

Now to the business of making your cones without a cone roller, or other cone forming device. As I said earlier, I’m assuming that you know how to project the cone on a flat sheet of metal and mark it out because you’ve read a book about it. However, even the best tuning books that I’ve read make a heavy meal of explaining how to do it. Seek help if you’re stuck. Metal workers and fitters know how to do it and there are sure to be some of these in your motorcycle club. (or if you’re stuck, email me-

I have found that 20 gauge panel steel is about the upper limit for bending cones up by hand. I have used much thinner steel to make chambers for lightweight bikes without problems, but I have brazed them together, not welded them. If you are not a confident welder and not physically strong, I suggest you take this route for your T500. The difficulty with thin sheet though, is that sectioning it to achieve bends gets tricky. If thin sheet is the road you have to go down, you might consider the forward extending headers previously mentioned.

When I first tried my hand at making chambers, I used a vice with six inch jaws, 3 feet of inch water pipe, 3 feet of 4 x 2 inch timber, two F clamps (G clamps will do) and a lot of sweat and time.

I clamped the wood and the pipe horizontally in the vice, then I slid the piece of metal to be formed between the wood and the pipe, with the smaller end of the cone-to-be at the vice end. The next step was to clamp the pipe and the wood together so that the pipe bisected the metal from end to end.

It is important to bend from the middle outward more or less equally in both directions or the edges will creep longitudinally and you wind up with a step where your seam ends should be flush. Too much of this will change the angle of your cone away from you design specifications.


Now, little by little, work the sheet metal around the pipe clamping and unclamping and moving the metal through a wider arc at the larger end. Then reverse the metal so that the small end faces away from the vice, and work the other half round bit by bit.

Finally, when the cone is pretty well formed, you may find that you need to hold a piece of wood flat against the metal and tap it around with a hammer. This is particulary so for the smaller end of the cone. You also have to be careful to work the middle of the edges around properly or you get a belly-out which makes aligning the edges for welding difficult.


I never said it was easy! Now you know why the books suggest you get a sheet metal shop to make your cones. However, with thinner metal, cones can be bent up round just a pipe with one’s hands and a soft faced hammer.

So there it is: if you are a masochist, or just get a kick out of working with metal, what I have described is one way of doing it by hand.

Alright, so you want to weld them yourself too! You can pull the edges together using large pipe clips, or wire loops twisted up tight with a nail, I’ve even done this with string! Of course, you can buy vice grips and welders clamps, but that’s too easy!

**Tack weld the edges. Start in the middle, then do a tack a couple of inches to the right, then a couple of inches to the left and so on out from the middle. There are other sequences. Do what works for you, then weld between the tacks working in a sequence if you must, but I just weld from one end to the other once the tacks are done.

It’s a good idea not to take the welds right out to the ends of the seams. If you have to trim a bit off the ends with snips, it is easier if you don’t have to munch through the weld as well.You can button this seam up completely when you weld the headers and the belly section in.

It pays to make your cones a bit longer than you need at both ends. There is nothing worse than finding that your cone openings are too small at the belly section end and too large at the header pipe end!


Make this longer than you think you will need it to be. You can trim a bit off if you need. Again, you can make trimming easier by not carrying the seam welds right to the end for now. Roll and weld it as for the cones (only it’s a hell of a lot easier).

THE STINGER (bleed pipe)

I hunt around for a bit of pipe of the right external diameter and “dong” the sheet metal around it in the vice with a hammer. A small lap in the joint is fine. Just weld it up (but not with the pipe inside it. You will have the devil’s own job getting the pipe out of the middle!)

Why don’t I just use a piece of exhaust tube of the correct internal diameter? Well, exhaust pipe tubing is relativly thick walled compared to the sheet metal I have suggested. If there is a big difference in wall thickness it seems to concentrate the stress around where the stinger comes out of the chamber and it cracks at this point.

This may just be because of the extra weight bouncing up and down. Certainly, you have less problems if you have part of your stinger down inside the diffuser cone, where the weight is balanced more evenly across the weld that attaches the stinger. The stinger pipe vibrates at a higher frequency with less amplitude on either side of the weld. [Not the sonic frequency, of course, but lets not get complicated. It doesn’t matter as long as the damn thing doesn’t break!]


I have seen people get in all sorts of a muddle cuting up a donut. If the cut does not go right across the centre of the circle formed by the donut, you get an oval shaped opening that does not match up well, and you get a variation in section where you don’t want one.

On the subject of inaccuracy; I have drawn up pipe designs for people. Two things made an impression on me. One was how rough the translation of the design into metal often was, but the second was how surprisingly well some of these pipes seemed to perform despite the indifferent execution of the design. The point being that, while you will gain a useful few percent in performance by being meticulous, you seem to have to make a total botch of a pipe for it to work really badly. So get the pipes made and get racing. You will inevitably want to make a better set further down the track!

But back to cutting donuts: you need to find the centre of the donut with reasonable accuracy. I set a compass to the radius of the inside of the donut’s bend, scribe the circle on a suitable piece of wood and cut it out on my bandsaw. I then put the circle of wood into the hole in the donut and scribe my cutting lines through the centre which is now nice and obvious.

However, I have also found that the circular lids of plastic jars sometimes come in sizes that fit pretty nicely in the hole, and there is often a little mark in their centres. This of course saves time and suggests that God and people who make jam are favourably disposed toward people who build motorcycles.

I suggest that you use a protractor to scribe lines at 10 or 20 degree intervals around your donut. This will help you keep on track if you have to trim a bit off after you have cut the donut into pieces.

At this stage my advice is going to get a bit nebulous because I don’t know whether you are going to bend your pipes all over the place, or go with a less demanding option.

Anyhow, here goes:

** tape the three sections of your chambers together with masking tape or some other sticky tape.

** hold the chambers up underneath the motor both with the back wheel in and with it out.

I am assuming that you have trimmed all the brackets and other vestiges of road bike off already. There are the heads of bungs and other things protruding down under the gearbox on the right hand side as well. You may also need to take this into account.

**once you have decided where you are going to place your chambers (this depends to some extent on what amount of sectioning and rewelding you are prepared to undertake) I suggest that you tape or wire them in place temporarily.

If you can visualise where they will go accurately, ignore this previous advice, but be warned that things can get awfully out of shape if you haven’t done this sort of thing before.

**now you can start to work out where the the headers wil run, what bends you need and what straight pieces you need. “TTCO” and leave a bit to spare if you are not sure about anything.

**as you cut things, you can tape them together. Eventually, you will be able to bolt the taped up headers onto the barrels and see what you have got.

It’s a good idea to tape parts of your pipe together at every stage to see how they fit on the bike.


** When you are satisfied that it will all fit, you should roughly measure the length of the headers along there centrelines to check that your header length is not too far outside the limits for the pipe design you are using. We will have a look at doing this a bit more accurately soon.

**once you know the header length you will begin to have some idea of whether the belly section needs to be shorter or longer to achieve the tuned length required.

Keep in mind that the tuned length is from the piston face to the mean wave reflection point, along the centre of the pipe. When you drew up the design you will have marked this in (I hope!) but locating this point inside the baffle cone can be a bit of a chore. More about this later.

**draw two felt pen lines from metal to metal across the tape that is holding your header pipe together, this will help you find the relationships again when you tack weld the pieces of donut and pipe together. If you use two colours, you won’t get them 180 degrees out by accident!

**as you peel each piece of tape off, extend the felt pen lines so that they join up, and number each piece so you don’t get lost (it is probably 2 am by now and the once sharp intellect can get fuzzy after 10 cups of coffee and not much sleep). I also suggest that you un-tape and weld one header at a time so the pieces don’t migrate from left header to right header. You might not be that dumb, but I have been!

**tack weld the pieces together then bolt them to see how things are going. If it the headers look good and they both check out as being the same length, weld them up now, but if you are not sure about anything stick with the tack welds for now. Then with a few hacksaw cuts through the tacks you can make adjustments.

** now do a final trim of the header pipe ends of the diffuser cones so that the inside diameters match up with the header pipe inside diameters. Check that it all fits again and make your felt pen marks to help you line things up for welding.

You will have noticed that I use a lot of things like string wire and tape to position things. This is because I generally work alone. If you have the luxury of an assistant to hold things in place, you can probably shortcut a lot of this. For example, you could get someone to hold the diffuser in place under the bike while you tack it to the header.

** trim the outlets of the diffusers to fit the diameter of the belly sections, then stand them on end and measure their length by plumbing the centre line with tape or ruler. Don’t worry if they are not exactly the same length, just write the length on them in felt pen. There is a way to compensate for small differences.

**tack weld the belly sections on.

Incidentally, your chamber sections will often come out a bit pear shaped around the seam welds. Ideally you should put the cones and the belly section on the beak of an anvil and strike them flat with a hammer. You probably don’t have an anvil, so put a solid lump of pipe in the vice, slip the sections over it and strike the welds flat on the pipe. This will usually go a fair way to solving the pear shaped probem. You will also often find that the sections go together better if the seam welds are lined up with each other. This also means that all the seams can be positioned where they don’t show (unless you can do welds that you want people to admire, of course).

** now to the thorny question of getting the tuned length somewhere about right – and don’t forget that the depth of the exhaust port needs to be added to the length of the header pipe. (That tuning book that you read will have explained about this.)

**draw a felt pen line along the centre of the outside of the header pipe. Yes, I know this gets a bit”guess and by God”, especially if the header is in more than one plane, but its near enough and believe me you won’t want to know about the other ways of doing it!

**Now use dividers set at a small distance (1cm or ½ an inch will do and you can just use a ruler if you don’t have dividers) to step out the length of the header along the line you have drawn. Both headers should come out as near as dammit the same if you have worked carefully and matched everything up as you went along.

But this is your first effort at this sort of thing, right? So don’t panic if they are a little bit out. If it’s wildly out of whack, you will probably need to go back and adjust one of the pipes anyhow, because they probably won’t fit properly. On the positive side, this is a good thing to know before you get in deeper!

You have already plumbed the lengths of the diffuser cones, so you know what length they are and that they are pretty much the same length. The lengths of the belly sections are easy to measure because the walls are parallel. I usually write the lengths on the headers and other sections with felt pen because Dr Alzheimer seems to have visited me early, and I forget easily! But seriously, a lot of us have to get our bike work done in time that is stolen from Morpheus, because the rest of our lives belong to work, wives and kids. Stuff-ups because you haven’t kept track of what you are doing burn up valuable workshop time. You do need to be a bit careful.

**on the drawing (full size or scale) that you did when you planned your pipe, you will have a triangle that represents your diffuser cone. ( If you haven’t extended this out to a point, you should do this). The mean reflection point, which represents the end of your exhaust system for tuning purposes, is generally taken to be half way along a line that bisects this triangle. (that is, half way up the centre line of the cone) You can measure the length to this half way point on your drawing (don’t forget to scale it up if necessary).


If the diameter of the belly section pretty well matches your drawing, and you trim the diffuser cone to fit, then the length taken from your drawing will be the will be pretty well right.

So now you can add the lengths to see if your system matches your desired tuned length: port depth + header length + diffuser length +belly section length + depth of baffle cone to the mean reflection point.

  • REMEMBER THAT IT SHOULD BE TOO LONG because you made your belly section longer than you needed just in case you stuffed up somewhere! So now you can trim the belly section back till you get the right length.
  • now a sneaky little trick! Somewhere along the line it is highly possible that your pipes may have gained or lost a bit of length relative to each other, and because the T500 is a twin, we don’t want this to happen. Sneak back into the house and pinch the dressmaker’s measuring tape out of the sewing box. Wriggle it down through the pipe from the beginning of the header to the end of the belly section. Pull it tight and mark with your trusty felt pen where it enters and exits. Note this length. Now mark the same exit and entry points on the other pipe and do the same measuring trick.

This is NOT the length of the pipe along its centre line but it will tell how much variation in length between your two pipes there is.

** trim the belly sections so that both pipes are the right length to weld up to the baffle cones and, even though there may be some accumulative variations in the pipes, they will be acceptably close to the same specified length.

** if you’re happy with everything, weld it all up and if you’ve decided on the bend-it-all-over-the-place route, go and get some sleep. You’ll need to be wide awake for that part of the exercise.


If you have looked at the photos with this article and some of the photos in the galleries on Murray’s site, you will have an idea of the problem. I don’t know how long your wheelbase is or how you want to route your pipes, so I’ve taken quite an extreme approach in the hope that it will cover most needs.

You will see from the photos that I’ve kept the header pipes in close. This means (see photos) that the chambers extend well back into the area occupied by the rear wheel. I have also installed a 4 inch shorter swinging arm which brings the chambers and the pipes further into competition for space. Added to this, the chambers themselves are over 4 inches in diameter, so it’s starting to get reasonably crowded.

I’ve assumed that you are not going to chop out, raise and reweld the cross tube behind the gearbox ,and that you are going to stick with the standard rear footpeg mounts for your rearset pegs. I’ve made this assumption on the grounds that you’re keen to stop welding and go racing.

This also means that if you decide that you haven’t got the time or the gear to tackle making chambers, you can order a pair with a reasonable expectation that they will bolt straight on. The further away from the standard frame you move, the harder it is to fit pipes from a distance. A lot of people who make pipes prefer to have the bike and do all of the fitting themselves. After all, there is some pride and reputation at stake!

I’ve taken a pattern off the T500 pipes I’ve made so that I can form them hydraulically. Hydraulic forming has the advantage that you can do “production runs”. Once you have the basic form worked out, you can produce variations to cater for different performance requirements. If you are interested in giving it a go, look at Michael Moore’s Eurospares site on the net. He does a good job of explaining the basics, but keep in mind that making a convoluted two-stroke header and chamber can have a few wrinkles in it, both of the technical and physical kind.

** I’ve taken photos for this article so that they you can see which way the bends go. Take the left side: the header crosses over the left hand front downtube, then goes down under the engine where it joins the diffuser cone. Just before it joins the belly section, the cone has a bend which takes it upward and outward at about 45 degrees from under the cross tube behind the gearbox. The next bend is in the belly section and continues in the same direction, making room for the wheel. The final bend is downward and inward, to tuck the chamber under the footrest bracket and back in toward the wheel. The right side system is pretty much a mirror image of this.

I can think of at least one other way to do this, But however you do it, if you get the pipe nicely positioned under the footrest and well tucked in toward the wheel, the mounting bracket can go straight up to the footrest bracket. You need to think about this mounting business when you plan you pipe.


The tuning books will tell you what to do, but they don’t tell you how. They do tell you that “lobster back” bends are not good because you get a change in section with each cut that goes right across the pipe and the pipe gets progressively shorter. So bang goes your tuned length!

This is true, but I have seen an Ariel Arrow with the headers formed with very neat lobster back bends. It went well, though it probably would have gone better with better bends.

Equally, I recently made a chamber for a large capacity post classic motocross bike because the owner was having problems with a lovely shiny imported pipe. When we measured it up wasn’t anywhere near appropriate. Caveat emptor! Another pipe he had had made to solve this problem didn’t work either. There was nothing fundamentally wrong with the design and it was a well built pipe, but the bends were lobster backs.

A bend in the middle of the belly section so narrowed the pipe at this point that I suspect that it was reflecting the pressure wave and that some peculiar harmonics were operating. This may have been why the owner found it impossible to tune the carb to operate over the whole rev range.

The chamber we fitted solved this problem, but the rider concerned is in the expert class and would now like more grunt higher up the rev range. Can do – but I’m glad he’s riding it, not me.

So pipes can be made so badly that they won’t work, but you really have to try hard to make them this badly…….

Okay, so the recommended bends are formed by “pie sectioning” the pipe with a “V” shaped cut to the centre line and a single straight cut to the apex of the “V” from the opposite side of the pipe (the cuts then form a “Y”). This is easier said than done, and the thinner the pipe and the larger the diameter (in my experience) the harder it gets.

Buy the finest toothed hacksaw blades you can if you are working with thin metal. It is very easy to snag and break coarse toothed blades at the best of times.


Some Tips:


  • Cutting chambers can be hard on your ears, so wear ear protectors.
  • Use fine toothed hacksaw blades and lubricate them.
  • Use softwood pads in the vice. They help to grip the chamber without crushing it.
  • Don’t rush the cutting and try to use minimal pressure as you cut.
  • Try to use the whole length of the blade.

Marking out keeps things on track.

** Make a mark where you want the inside of the bend to be.

** Draw a longitudinal line across the mark.

** Use the dressmaker’s tape to measure the circumference at the mark.

** Make a circumference line right around the pipe where you want the bend to be.

** Divide the circumference by four and use the tape measure to mark the circumference line off into four equal sections, starting from your original mark.


Now you need to decide how big a pie section you need to cut out. Because this depends on how you intend to route your pipe (and its diameter) I can’t give you a firm guide on this. I have found that about the depth of a hacksaw blade seems a good place to start for less extreme bends and twice this really rarks the pipe round. I use two bends if I have to come round that sharply.

** Mark the cutting lines out at equal distances on either side of the mark which indicates the inside of the bend.

The half way through the pipe mark will be the quarter circumference marks in either direction. You need to position yourself above the job so that you can both cut and see how you are progressing toward both these marks. If the vice is mounted high or you are short, you may need to organise something to stand on. Try and get the blade lined up properly from the beginning and don’t rush the cutting.

** When the pie section is out, turn the pipe over and cut through the remaining pipe following the circumference line.

In theory, all you do now is close up the “V” and weld it, then weld the pie section into the space that is now on the opposite side of the pipe. In practice, even if you have done everything to perfection, you have lost the metal in the width of the hacksaw cuts.

This means that you have a hole wider than the section you are about to weld in. This in turn means that if you attempt to fill the gap with weld you may get a nasty, slaggy weld ridge inside you pipe, even if you do get it to look presentable on the outside. I have opened up a few old pipes ( my own as well) What I have found has led me to use the section I have removed as pattern to cut a new section that fits snugly in the hole.

If you decide that this is a bit fussy for your needs, then use the piece that came out.

** You may find it hard to avoid dropping the piece into the pipe through the hole and having to fish around for it. You can get round this by puting two strips of masking tape across the piece, sticking it where you want it to be then putting a couple of tack welds in the middle. You then peel of the scorched pieces of tape.

Small pieces of thin metal tend to warp so you will probably find you have to tack weld, working out from the centre and crossing from seam to seam, before you complete your welds.

You will generally find that a bit of tapping with a hammer and levering with a small screwdriver is necessary to get things to sit properly while you are tacking up.


The first thing of course is TTCO, but another thing is to try the pipe on the bike at every stage after you have tack welded a bend.

If you have haven’t gone far enough with a bend, you can cut through the tacks and try to take a bit more out. This has to be done on both sides equally. It can be a bit fiddly or a real mess. Putting another bend in to bring it round further is sometimes a better option.

If you have gone too far, you can cut the tack welds and make up two pie sections – one for the outside and one for the inside of the bend – this looks quite tidy.

A little trick is to do cuts that stop about one sixteenth of and inch before the marks toward which you are cutting. You then have about an eighth of an inch of uncut metal on either side of the pipe. You can bolt the pipe on and bend it into position using the uncut metal as a hinge. You then make up pie sections to fit the gaps. This also makes the pipe easier to handle on your own because you are not struggling with separate pieces, and if you have got it wrong it makes it easier to measure up what size the new pieces need to be. Watch it with this “trick” though. The hinging action of the remaining piece of metal can be off centre. The result is that the edges do not butt together well when they meet.

If you do wind up having to put new eliptical pieces of metal in the holes, how do you measure up a piece to fit in the hole?

You don’t – it’s too much like hard work:

  • place a piece of clean paper over the hole.
  • rub you workshop begrimed finger on the paper around the edge of the hole. You now have an eliptical outline.
  • cut round the outline with scissors.
  • place the cut-out on a piece of the sheet metal you are using.
  • spray it with a black (or any other suitable colour) spray can and remove the paper pattern.
  • cut round the outline that is left on the metal.
  • bend the cut-out to fit the rounded shape of the pipe.

You should get an almost perfect fit.


If you work your way through the geometry of what happens if you take really big segments out to form your bends, you will find that distortions in the shape and section of the pipe begin to occur. This is another reason for making two bends using small segments, rather than rarking the pipe round in one sharp bend. Stick with smallish segments and you won’t have to worry about this.

Try to make each arm of the “V” shaped cut the same length or they won’t butt up well when you close them together.


Straight pieces of pipe are not too hard to get right, but making good bends in cones requires a little more thought because the walls are not parallel. The three cuts no longer make a symetrical “Y” shape. If they did, one arm of the “V” shaped segment would be longer and they would not butt together properly. The “V” shaped part of the”Y” tilts in the direction of the narrower end of the cone, so that its perpendicular bisector forms a right angle with the surface of the cone. I hope the diagram will make this clear! You can usually “eyeball” this about right if you are aware of the potential problem. If you are not good at eyeballing, set a square up on the surface to help you get lined up right.



Bell’s book will tell you what you need to know about attaching exhaust systems to bikes, so there is no point in me writing about that here. Have a look at some of the pipes around the pits. That will give you a few ideas too.

Well, that’s just about it from me on T500 (and other ) chambers.

I’m sure there are better ways of doing many of these things. I hope what I have written will encourage people to contribute. I’d also be happy to have people email me at – A bit of feedback about anything that doesn’t make sense or is hard to follow would be useful.

There is a copyright on this material, as I hope to publish it along with some stuff on porting and combustion chamber modification (when I get it together). But by all means copy what’s here for personal use.



Original text and images copyright reserved: Glen Morgan 1999



Author: muzza