How straight does a frame been to be?

[531colin]

003 by 531colin, on Flickr

Nuff said

[531colin]

Even with a rear wheel, you can check that it is the frame and not the wheel that is out by putting the wheel in the wrong way round, with the gear side of the wheel on the left. If it is the frame that is out, the tyre will still be 2mm closer to the left seat stay. If it is the wheel that is out, the tyre will now be closer to the other seat stay. Might be worth doing just to double check what is what.

It would only take a tiny something or other at the dropout to generate the sort of misalignment you have, which sounds like the tyre being 1mm further to the left stay than it should be, and 1mm further from the right stay than it should be. You really only want the tyre to move 1mm to the right at that point. I'd be looking with a magnifying glass at the dropouts and the ends of the wheel axle to see if there are any tiny burrs or other anomalies that could be sorted with a file.

Also, because the right stay has to go further from the central line, due to the gears, it angles outwards from the bottom bracket shell more than the left stay does, so complete symmetry in the area of the tyre should not be expected. The right stay will generally be a tiny bit further from the tyre, even if everything is lined up right. You can see that on Colin's post, which follows this one.

No it doesn't
Very few frames have been made asymmetric; a few Orbits I have mentioned.
Traditionally, the frame is symmetrical. The hub is equi-distant between the dropouts, and therefore the cone locknuts, or their equivalent on sealed bearing or thru axle hubs.
Where you have a bank of sprockets one side, the rim isn't equidistant between the spoke flanges; the wheel is then said to be dished.....
...but as you said at the top of your post, you can put the wheel in with the sprockets on the wrong side, and the tyre should still be in the middle of the chainstay or seatstay gap

In the photo, I haven't got the camera lined up absolutely right; you can see the wheel "looks wrong" against the seat cluster, whereas in fact its correct against the seat cluster

005 by 531colin, on Flickr

[Brucey]

IIRC most frames built from tubes are perfectly symmetrical (or are meant to be) until somebody reshapes the RHS chainstay in some cases, to make room for the chainset. That should be the only form of side-to-side asymmetry to be found in most frames.

[pwa]

Well I stand corrected on that. I have always thought the right side chainstay comes out from the centre line more than the left side, never noticing that in fact that cannot be true, otherwise a wheel put in the wrong way round would have the rim in the wrong place, which of course it doesn't. Unless the wheel is dished wrong. It is funny how one gets little misconceptions in one's head and carries them, unchallenged, for years.

[southcraven]

thanks for all the help.

I spoke ellis briggs this morning and they said it is most likely the length of the chainstays being unequal and they would have to unbraze the longer side and shorten the chainstay slightly and then braze it back in.

But now I've got to decide whether I want to go to the expense of repainting it, or touching it in or something

[Jezrant]

Assuming Ellis Briggs is right and is offering to repair the frame as a gesture of goodwill (but won't cover the cost of a full respray), there should only be a small area that would need touching up. A complete respray isn't absolutely necessary for that sort of job.

However, I'd personally still pop over to Woodrup's with the bike and ask them what they think before doing anything (you probably know they have acquired the Bob Jackson marque since BJ closed down). One of the lads in the workshop will be able to tell you what's up with that frame and what to do about it.

[Jupestar]

Have you tried a different wheel - is it the same?

[slowster]

I spoke ellis briggs this morning and they said it is most likely the length of the chainstays being unequal and they would have to unbraze the longer side and shorten the chainstay slightly and then braze it back in.

I would have thought it more likely that the frame was simply out of alignment. The videos below show the frame builder Paul Brodie checking and correcting alignment.

This one shows just the rear end alignment process, from 28.00 mins in:



This one is more general:

[531colin]

thanks for all the help.

I spoke ellis briggs this morning and they said it is most likely the length of the chainstays being unequal and they would have to unbraze the longer side and shorten the chainstay slightly and then braze it back in.

But now I've got to decide whether I want to go to the expense of repainting it, or touching it in or something

If the tyre is 2mm closer to the frame one side than the other, that means the tyre is ONE WHOLE MILLIMETRE away from centre.
Just file the dropout; you only need to take a fraction of a millimetre off the dropout, the wheel radius is much greater than the axle length so the rim moves several times further than the axle end.

But first, check the alignment of the whole frame and forks viewtopic.php?t=59332

[fastpedaller]

Absolutely - Aside from the cost, unbrazing parts and re-doing them for a fraction of a mm is likely to cause more damage than it's worth

[CJ]

What matters is that the frame is in track, or not too far out of track. And half a tyre's width is my rule of thumb for a bike that's knocked about a bit: the point at which I'd try to make it better. I'd expect a new bike to be better than that to start with of course.

The tyre is about 2mm closer to the chainstay on the left. Its a Bob Jackson

That's interesting but I'm afraid it doesn't tell us whether or how far your bike might be out of track. It's quite possible to have a perfectly tracking bike where the wheel doesn't sit central between the chainstays and isn't even 'correctly' dished. Because tweaking the rear wheel dish and skewing the wheel in the frame are two of the ways of putting a bike into track that are easier, quicker and much cheaper than taking it to an expert craftsman to have its bent frame straightened!

Chris Juden's dishy knicker elastic method!

Here's a simple test to tell if your bike is in track, that you can do at home without any special tools. And if not: you can even use it to measure how far out of track a bike is.

1. All you need is a couple of metres of knicker elastic, tied to make a loop (preferably without any twist) that with a bit of stretch goes around both wheels, parallel to the ground at a height about halfway from floor to hub. The bike, of course, should be standing on its wheels (else they'll turn and the elastic flip off into the distance!) and leaning on a wall as near vertical as you like. If the elastic won't grip the tyre treads well enough to stay at that height, make it tighter (i.e. shorter) and/or move it a bit higher so it's wrapped around a less sloping place on both tyres. But don't go so high that spokes begin to interfere with the elastic. You want it touching tyres and nothing else. The wider the tyres the higher the elastic can go and the better this works.
2. You should now have elastic wrapped around the front of the front tyre, rear of the rear tyre and also kissing both sides of each tyre at the other point where it passes, i.e. at the rear of the front tyre and front of the rear.
3. First pay attention to where the elastic passes either side of the rear tyre near the cranks. If by turning the steering slightly left and right it is possible to find a position where the elastic just or nearly touches both sides of the rear tyre: your bike is in track, worry no more.
4. But more likely the elastic never gets quite near enough to touch one side or the other of the rear tyre; and in that case your front wheel is out of track towards the side the elastic refuses to touch.
5. In that case you can even use this simple method to measure how far out of track your bike is! Lets say the elastic refuses to touch the left side of the rear tyre. Turn the steering to the position where elastic is as close as it gets and then turn it very slowly to the left until the elastic just starts to lift from the opposite i.e. the right side of the rear tyre. Hold the steering there and switch your attention to the front tyre. Measure the gap between the elastic and the left side of the front tyre. Your bike is out of track by half this distance.

Note however: if the tracking error is mainly due to one wheel leaning relative to the other, out of vertical, this measurement will under-estimate the error somewhat.

[Brucey]

you don't yet know for sure what is wrong exactly, it might well be that the back end of the frame needs to be cold-set. If so, normally the discrepancy at the chainstays would be accompanied by a similar (but much less obvious) one near the seatstays. So I would certainly check for this before fretting about the need for a new paint job. Even if the chainstays are odd lengths, (which, btw, would not by itself fully explain the 'no-hands' issues) the difference can only be gnat's wotsit, little more than the paint thickness in the dropouts, really. This is something most people would expect to fix with a file, not a blowtorch.

[gibbo1982]

Since my name is getting mentioned so much, here is my 2p worth!

Southcraven got in contact with me today regarding this. I explained that the wheel would need checking for alignment and dish before anything else, then frame and fork would need checking for alignment to ascertain what the alignment issue is, if indeed any at all.

Shortening a chainstay would be a last resort.

Some people have pointed out that filing the dropout is a way of correcting alignment. But what you have to remember is that there are many variables in the alignment process and what you alter in one place affects other areas. Also filing the dropout on vertical end will only affect the tilt of the wheel in the vertical. I wouldn't advise opening the width of the dropout with a file to try and influence its position between the chainstays.

I don't know where the comment about doing it as a gesture of goodwill came from? I've never even seen this bike/frame and I certainly had no part in building it originally!

Those who think you can re-dish the wheel and somehow bring the frame into track?

And as for Orbit intentionally building frames out of alignment to decrease the amount of dish on the wheel! Wow. That just screams a BS excuse for not getting it right in the first place. Orbit were notoriously badly built! Sometimes we would blast one for painting and the downtube was cut too short for the lug and there would be a tiny gap between the lug and the tube which had just been hidden by paint!

Also I would point out that Bob Jackson, like many frame builders who built there frames in a day or so, had a very crude alignment system. So a little bit of mis-alignment is common. To be honest even the tolerances on mass produced carbon bikes leave a lot to be desired. I'm not really looking to correct built in alignment problems on other builders frames, they are what they are.

On another note, Chris Juden's knicker elastic method is good indicator, because tracking literally is about the wheels being inline. But you can't go jumping to any conclusion without measuring every variable to see where the issue lies.

Ellis Briggs

[gibbo1982]

Also I should have added that I expect Bob Jackson frame quality to improve a lot now Woodrup are going to be building them, as I know they have always taken more care, like ourselves.

EB

[gibbo1982]

I spoke ellis briggs this morning and they said it is most likely the length of the chainstays being unequal and they would have to unbraze the longer side and shorten the chainstay slightly and then braze it back in.

I would have thought it more likely that the frame was simply out of alignment. The videos below show the frame builder Paul Brodie checking and correcting alignment.

This one shows just the rear end alignment process, from 28.00 mins in:



This one is more general:

Also on the subject of Brodie's alignment process it is flawed, which is why he gets confused flipping the frame on the alignment table.

EB