Honda-SOHC
SOHC.co.uk Forums => CB500/550 => Topic started by: deltarider on March 13, 2026, 10:55:17 AM
-
What length chain do you CB500 owners fit: 98 or 100 links and why?
-
My own answer is I don't know,
Most chains I buy are to long, so I shorten them by several links
Sent from my SM-A546E using Tapatalk
-
Same as John
-
Depends on amount of teeth on sprockets. I had a 550 rear sprocket on my 500 which needed a 100 links but now changed to the correct 500 one and chain needed a link removing to fit.
-
There is not a correct answer here. I agree with John and Bryan, just put it together and shorten to a point where you have plenty of adjustment space.
-
I may have found confirmation for what I have once read some 45 years ago. It is in this site for chains and sprockets sets: https://www.kettingkits.nl/nl_nl/route
I find most chain sites next to useless, but this is actually a good one. It's in Dutch, but don't let that stop you. It's all about digits, not words.
Select 1. Honda, 2. 550 and 3. CB550. This site suggests what these Honda's originally had; for the CB550 100 links and sprockets 17 and 37. Now this site offers you options to varify the size of sprockets [configureer uw kit] and now comes the interesting bit: if you have the CB550 selected and change the original rear sprocket from 37 into 34, the program reacts automatically by reducing the 100 link chain to a 98 one. Maybe this is a confirmation for what I have read long time ago. Can you guess what it is?
-
Our point is that it is easier, and cheaper, to buy a box chain at 110 and cut a bit off, when Reynolds still produced motorcycle chain dealers bought it in 50ft rolls and sold you whatever you wanted, now bean counters say it isnt worth stocking all the different lengths just one the longest needed and shorten it
-
deltarider that agrees with what i said, i changed from a 37t to a 34t and had to shorten from 100 links to 98.
-
Here is the confirmation I was looking for. With a final reduction of exactly 2 (17/34) and a 100 link chain, you run the risk of uneven wear, as always the same link 'lands' in the same spot on the sprocket. Going to 98 links changes that and wear will be spreaded equally. So the CB500 had 98 links and when Honda decided to market the CB550K3-ED as a touring bike and therefore changed the 17/37 into 17/34, Honda consequentely changed the chain to a 98 link one. Does this make sense or am I overlooking something?
-
Here is the confirmation I was looking for. With a final reduction of exactly 2 (17/34) and a 100 link chain, you run the risk of uneven wear, as always the same link 'lands' in the same spot on the sprocket. Going to 98 links changes that and wear will be spreaded equally. So the CB500 had 98 links and when Honda decided to market the CB550K3-ED as a touring bike and therefore changed the 17/37 into 17/34, Honda consequentely changed the chain to a 98 link one. Does this make sense or am I overlooking something?
Does that in reality stack up to give the outcome suggested ? I'm not immediately convinced that's the case.
If it were a pair of gears, then the 1/2 ratio could have some impact , possibly.
With a chain of X no links, you'd divide the links by the tooth no, wouldn't you ?
And so, 100/ 17 or 100/34 doesn't divide equally to give a repeated engagement "wear pattern" anyway.
Taking the 100/17 would see the repeated engagement through the two at 5.88 ..... and so would see the chain effectively cycle through all of the links in rotation.
-
Wouldn't you also need a chain of 102 links to give any semblance of repeatability ?
-
100:2= 50 (still an even numer), 98:2= 49 (odd). Is it my intuition that tells me there's the answer? ;)
-
100:2= 50 (still an even numer), 98:2= 49 (odd). Is it my intuition that tells me there's the answer? ;)
And what impact does that "50" have ?
Both 98 & 100 aren't exactly divisible by either 17 or 34, making the chain effectively "cycle" around the sprockets throughout its use. Meaning, that each tooth form will interact with a different chain link in successive rotation. That over a period of time until the mathematics repeats, eventually.
-
You seem to be forgetting that only the 550 has a chain length of 100, the 500 is 98. The 500 uses 17/34 sprockets whereas the 550 uses 17/37 sprockets so the maths is different.
-
This reminds me of the old brit bikes when the cam gear is always double the crank,but the intermediate gear had an odd number of teeth to spread the wear,so called hunting tooth idea
No idea if this much thought was invested in the final drive
Sent from my SM-A546E using Tapatalk
-
The reason the rear sprocket size changed on the 550 is because the primary drive ratio changed and not for any esoteric reasoning of chain to sprocket ratio
-
Got to say I disagree with that Bryan. At the time the 550 was in real competition with stuff like the GT550 and the kwack 500 and most of the motorcycle mags tended to be fixated on standing quarter times, so Honda tried to counter that with a bigger sprocket to get that time down.
That’s my take on it.
-
Ken, if you look at the overall ratio the 500 and 550 are almost identical
-
I have the feeling I have not been able to communicate what I meant. There are two things: if a bike can do it with a 98 links chain, why not. The second is: after some time not all links are equal in souplesse. From a designer's pov, you don't want that less supple link to hit the very same spot on a sprocket every rotation (hammering effect). Better is to have it arrive at all spots in turns, so the wear is spread evenly.
-
A98 link wont fot with the 37 tooth rear
-
Easy enough to test though. Mark a link and a tooth on the rear sprocket and see if it lands on it everytime. Get someone to test the 550 and see if it does it on that.
I had a test lined up Bryan to test your theory. Unfortunately my illness meant it has been delayed until I can get back to working on the bike. What I intend doing is fitting a 500 clutch basket on my 500 as standard, do some tests on revs versus speed etc, maybe mpg and then replace the clutch and primary gear with one off a 550. I have a 550 rear gear on a 500 basket so it’s a simple swap, same engine etc so no bias should be apparent. Then do the same tests and see how they compare.
As you know it’s not possible to fit a full 550 clutch on a 500 engine due to crankcase differences. Hence why I designed the test this way.
-
Sounds god Ken, i was working from the workshop quoted figures
-
Johnwebley came closest. I am well aware of the differences between CB500 and CB550 transmissions.
Please do a test by marking, but I predict that with a final reduction of exactly two, which is what you have with a 17/34 teeth configuration, the same link will always land in the same spot of the sprockets. That's why it is best to have a 98 link chain on a CB500. Over the years I had to persuade a seller sometimes to shorten a 100 link chain to a 98 one.
What is nice about the CB500 set up, is that in top gear 5000 RPM equals 100 km/h, 6000 120 km/h, 7000 140 km/h, 8000 160 km/h, 9000 180 km/h and 9250 a theoretical top speed of 185 km/h. The circumference of the tyres fitted - both by size and wear - play a role however and will lead to some deviation.
-
Am I missing something here? If the 500 with standard sprockets needs a 98 link, then the 550 uses a bigger sprocket on the rear and all other dimensions are the same then a longer chain is required due to the bigger diameter? Also the amount of adjustment available is a factor.
-
Just to complicate matters slightly,
When I started fitting modern tyres, the rear only available as 4.00 width, it also had a larger diameter and circumference
So I reduced my gearbox sprocket by one tooth
Sent from my SM-A546E using Tapatalk