digitane

on my track bike: the chain is rather loose when the right crank arm is straight forward; and the chain is very tight when the arm is straight back. rear hub, cog, chain, BB, chainring are all new; the crankset i picked up used on ebay... the set is suntour superbe pro 172.5mm. the square tapers look perfect, so i am assuming the crank arm and/or spider is bent, but i cant see it other than in the chain tension. whats the best way to actually measure whether it is indeed bent, and is there something i am overlooking?

HillRider

A bent spider or bb spindle would probably be noticed as lateral run-out rather than changes in chain tension. Either your chainring is not perfectly round or it is not mounted perfectly centered on the spider.

Try loosening the chainring bolts and turning the crank to center the ring, then tighten them again.

digitane

i actually did this right before posting the thread. the ring floated maybe 1/2mm as i turned the crank with the rings loose: chain acted the same, then i retightened the bolts: chain acted the same. its a TA specialties track ring, i measured and studied it: doesn't appear out of round at all, and it is brand new. also took the crank off to check tapers and BB spindle, no problems found.

is it possible that the spider could be off-centered?

HillRider

I guess the crank spider could be slightly asymetric or the mounting holes not perfectly centered. Since the crank is that old, you really can't do much with a warranty claim so the only "cure" is to use a round file to either elongate the holes in the chairing and flat file or modify the chainring's mounting tabs or to file the spider itself.

BTW, is the rear cog round?

San Rensho 

Could be the CR is bent. Take the chain off and spin the cranks as you look at them. The chainring should be true, no wobble side to side, as you sight along the legnth of the CR. If it has a lot of wobble, lay the CR on a glass table and see if it is fairly flat. If it is, then one of the spider arms could be bent in or out. You can bend them by very carefully and slowly bending the arm with a BFAW (big f@cking adjustable wrench). Grab the arm with the jaws of the adjustabe and slowly bend.

digitane

yeah, after inspection cog is round too, brand new EAI cog. i had anticipated the answer might be having to file the holes and/or mounting tabs. i'm still wondering: is there any way that a mildly bent crank (perhaps from use on the track over the years) could cause the chain tension problem?

digitane

ok, found a good flat surface, doublechecked ring for flatness (its flat,) put crank and chainring back on to track lateral movement of the ring again, took my time getting a good solid control point to track it against and there is a slight 1/2mm movement inward right at the top spider arm that i didn't notice before. the chain goes from tight at that point to loose opposite that point with at least 10mm of up-down movement. could that little movement cause the chain tension problem?

Soil_Sampler

could be were your chains rollers make contact with the ring that is out of round,
not the tips of the teeth.

how new is your chain?

digitane

chainring is good on all fronts, chain is new too.

i dont think that tiny lateral movement in the ring is causing such a big tension problem. i just put the chain back on and theres no way thats the problem. chainline is straight btw.

i think Hillrider nailed it with the idea of filing the ring and/or spider to get it centered. i think i'll chase after some very precise measurments of the spider holes against the spindle hole and the ring holes against its center to figure out what is truely off and what should be filed. busy now though so it will have to wait. i'll post again later once i get it figured out.

thank you all

HillRider

One last thought; try reorienting the chainring around the spider. You have 5 choices. Maybe you can find a position where the radial run-out of the chainring compansates for the radial run-out of the spider and that fixes the problem.

digitane

good idea, ill try that too before any filing, thanks again

G piny parnas

your wheel is not centered,,, or your hub has become uncentered--- randys doghnuts... ahhh.

Grand Bois

Makes sense.

DannoXYZ 

Uh no. An off-center wheel will still spin a circle around the axle. The rear-cog's teeth will NOT be moving forwards and backwards as the wheel rotates. Unless the cog is ovalized, in which case it'll cause tension-variation even with the wheel centered/

Most likely it's a sloppy fit between the chainring, the chainring-bolts and the spider.

1. How much radial-play do you have between the chainring-bolts and the hole in the spider?

2. How much radial-play is there between the chainring-bolts and the chainring?

3. Is there a ledge on the spider-arms where the chainring sits to centre it?

buzzybelmondo

try Sheldon's method:


https://sheldonbrown.com/fixed-conversion.html#tension

digitane

good call buzzy.

so, i thought the chainring was good, but hillrider was right on the first post: the chainring is out of round and needed to be centered properly on the spider. here's what i did:

i measured the crank and found the spiders holes are precisely centered around the spindle hole. so i left the crank alone. then, i took hillriders advice and changed the chainring position on the spider to see if that helped, and it did: i found that the chain didn't have nearly as big a tension issue after rotating the ring 2 positions on the spider. so, something in the way the ring sat on the spider helped there. last, i used sheldon's method of centering the ring; that helped a lot too. there is still a slight tension change when i rotate the cranks, but, compared to what it was like, its hardly there now at all. after a test ride around the block: chain doesn't bind or get loose, rides smooth, and its quite

problem solved for now as far as im concerned, but i'll be investing in a(nother) new chainring soon.

thank you very much everyone!

G piny parnas

I disagree with the notion or "fact"that a hub that is not centered will "not" cause elipse--
I have seen this with my own three eyes that when a wheel is not tensioned properly --- wsp
on the non drive side--- it caused elipses( a condition that can be cured with a wheel building class and a couple of shots)

G piny parnas

bent rings( bent back to sight) cause slop--- cranking backwards for stopping causes dead spots--- crispy creams and coka cola in the morning causes me to nap in the afternooon

DannoXYZ 

Sure if the spokes aren't tensioned correctly the rim will be elliptical, but that doesn't affect the hub or cog in any way. You can make a rim square or triangle if you want, but the cog and cog's teeth will spin around the axle in a perfect circle (assuming hub bearings aren't loose).

G piny parnas

I humbly disagree --- If the issue is a "dead spot" where the chain becomes slack at any given
point around the cog--- I see this as a factor to be looked at--- is the frame bent?
are the stays straight-- I know everybody gonna jump on this here post-- but I built a few ss's
and I sense something simple--- also I never had no perfect line on no bike yet--- always gonna hve some slop--- unless it be a big money bike-- you all talk funny