Rumbling noise, chain/ring or motor bearings?

[steve_sordy]

I was cleaning and lubing the bike today and when turning the crank, I noticed a low rumbling noise coming from the chairing/motor area.
My first thought was that it might be that the chainguide had shifted and was rubbing on the chain, so I removed it, no effect. So I took the chain off the ring and that silenced it. The motor spun in either direction nice and quietly, no rumbling. I reassembled everything and the rumbling returned.

The Shimano XT chain has 0.3% wear on 1500 miles (as has the XT cassette), the steel chainring (Shimano CRE80-12-B, 34T) is original equipment and has 3776 miles on it. So it could be the chainring, they don't last for ever. But as I was leaving the garage, it occurred to me that the chain tension will be putting pressure on the crank bearings and could be causing the rumble.

Anyone experienced this, or any useful comment? Any other tests I could do to isolate the problem?

 

[Montana St Alum]

You did the trouble shooting I'd have done. Does it make the noise in different gears, and when you turned the cranks with the chain off of the front ring, was the motor powered? If so, on all gears, I'm guessing it's the chain. Side to side wear on chains aren't measured by stretching, so it may be that's the issue. I'd throw on another chain and see if that fixes it.

 

[Stihldog]

My first thought was that the chainring was wearing and developing chain-suck. My next thought was the plastic chain guard was rubbing against the chain.

If the chainring is secured and torqued on properly …maybe the motor bearing? (Btw, I’m not familiar with your type of motor)

I removed my plastic chain guard after the cuff of my pants did a complete revolution. I no longer wear my stripped bell-bottom pants during a ride.

 

[Mikerb]

I would guess the chain is pretty worn.( ps I use the chain wear checker recommended for SRAM Type chains even on non T Type and the measure is far more accurate than the original type of chain checker measuring as it does the distance between opposite sides of the rollers rather than the distance between rollers).
I normally expect non Type chains to last a maximum of 600 miles and I clean the chain after every ride. It is also likely your chain ring is worn. If you put a new chain on the rumbling is likely to be worse...which would indicate that the chainring is worn. So I would put a new chain on............and then if indeed the rumbling is still there, or worse, change the chainring.
If you motor bearings are causing the rumbling, you would probably notice an increase in feedback through the cranks when you apply increasing crank torque. So unless that is the case I would suggest it is drivetrain.

 

[steve_sordy]

You did the trouble shooting I'd have done. Does it make the noise in different gears, and when you turned the cranks with the chain off of the front ring, was the motor powered? If so, on all gears, I'm guessing it's the chain. Side to side wear on chains aren't measured by stretching, so it may be that's the issue. I'd throw on another chain and see if that fixes it.

Yes, I went through all the gears as I was lubing the chain and mech. It made the same noise. Motor not powered.
I confess that I had not considered side to side wear on the chain. Although only at 0.3% using a steel tape measure, the chain doesn't owe me anything at all. Its surface finish is looking a bit scabby, so as I have a new XT chain in the spares box, I'll play safe and change it.

 

[RickBullotta]

Just throwing out the possibility that with the chain on, there is pressure/force pulling back on the chainring slightly, and maybe there is some issue with a bearing that only shows up under that load? Also, double check that it isn't the jockey wheels/pulley wheels on the derailleur and perhaps just vibrating through the chain.

 

[steve_sordy]

My first thought was that the chainring was wearing and developing chain-suck. My next thought was the plastic chain guard was rubbing against the chain.

If the chainring is secured and torqued on properly …maybe the motor bearing? (Btw, I’m not familiar with your type of motor)

I removed my plastic chain guard after the cuff of my pants did a complete revolution. I no longer wear my stripped bell-bottom pants during a ride.

The motor is a Shimano EP8. and it has been flawless since purchase in Jul'21. Even the knocking sound when descending that many complain about has been absent since shortly after purchase. (Well, I don't hear it).
The chainring is secured nice and tight.

 

[steve_sordy]

Just throwing out the possibility that with the chain on, there is pressure/force pulling back on the chainring slightly, and maybe there is some issue with a bearing that only shows up under that load? Also, double check that it isn't the jockey wheels/pulley wheels on the derailleur and perhaps just vibrating through the chain.

I agree that it might be the chain tension pulling on the crank bearings. I cannot hear the noise when I'm riding the bike because the rumbing is too quiet for me to hear above the sound of boulders splitting under my ferocious riding!

I had not thought to look at the mech end of then transmission because the noise appeared to be clearly coming from the chainring end. But its a very good suggestion. The mech has only done 340 miles, but I go through mechs quite quickly, always from damage, so its well worth a good look.

 

[steve_sordy]

I would guess the chain is pretty worn.( ps I use the chain wear checker recommended for SRAM Type chains even on non T Type and the measure is far more accurate than the original type of chain checker measuring as it does the distance between opposite sides of the rollers rather than the distance between rollers).
I normally expect non Type chains to last a maximum of 600 miles and I clean the chain after every ride. It is also likely your chain ring is worn. If you put a new chain on the rumbling is likely to be worse...which would indicate that the chainring is worn. So I would put a new chain on............and then if indeed the rumbling is still there, or worse, change the chainring.
If you motor bearings are causing the rumbling, you would probably notice an increase in feedback through the cranks when you apply increasing crank torque. So unless that is the case I would suggest it is drivetrain.

The chain may be worn past its useful point, but it has not stretched beyond 0.3%, as measured with a steel tape measure across 100 links. I used to average in the region of 600 miles when I was on 7, 8, 9 and 10-speed chains. But once I went to 12-speed chains my average shot up. At first I thought it was simply the chain design, but I now firmly believe that it was chain quality. I used to buy the cheapest chain I could find, but the first 12-speed chain I got came with the bike and it was a Sram XX1 Eagle. That had 1400 miles on it when I sold the bike and it had only 0.25% "stretch". I have been a fan of higher spec chains ever since. I have always been scrupulous about chain cleanliness. Before every ride, at the least, the chain is run through an old towel and lubed. It regularly gets a more intensive clean and lube.

The motor is still quiet when I am riding it, or more accurately I never hear anything! But I don't feel anything either. Also my riding buddy hears nothing either and his hearing is 100%.

I am definitely going to change the chain and I may also change the front ring as It owes me nothing and I've never had one last this long before (3776 miles), so it is probably contributing.

 

[Mikerb]

All my bikes are SRAM. The only difference between a GX chain and the next one up the line ..X01....is an additional hardened material.....but it is twice the price so needs to last twice as long....not sure that is the case. The biggest change to SRAM chain design came with the SRAM Transmission (T type) which has bigger rollers and straight top edge chain links. One of my bikes uses T Type AXS ....it will be interesting to see how long it lasts given the expensive components.

 

[RickBullotta]

All my bikes are SRAM. The only difference between a GX chain and the next one up the line ..X01....is an additional hardened material.....but it is twice the price so needs to last twice as long....not sure that is the case. The biggest change to SRAM chain design came with the SRAM Transmission (T type) which has bigger rollers and straight top edge chain links. One of my bikes uses T Type AXS ....it will be interesting to see how long it lasts given the expensive components.

Actually there was some formal testing done and the jump from GX to X0/XX was indeed significantly longer life.

Chain Testing - Zero Friction Cycling

zerofrictioncycling.com.au

 

[Mikerb]

Actually there was some formal testing done and the jump from GX to X0/XX was indeed significantly longer life.

Chain Testing - Zero Friction Cycling

zerofrictioncycling.com.au

Yes I saw that but I suspect there are too many variables for that to be conclusive. When you look at chain design and the fact that wear is measured in terms of distance between links the main area of wear is between the rollers and pins. That is the most difficult area to clean effectively and is only sealed by the presence of the chain links on either side. So once a chain loses its longitudinal stiffness it becomes easier for dust and mud to get into that area and act as a friction paste. So actually the degree of sideways flex may be as important as longitudinal wear when checking chain condition........and effective cleaning together with cassette design/ chain line the biggest chain wear differentiators.

 

[Mikerb]

Yes I saw that but I suspect there are too many variables for that to be conclusive. When you look at chain design and the fact that wear is measured in terms of distance between links the main area of wear is between the rollers and pins. That is the most difficult area to clean effectively and is only sealed by the presence of the chain links on either side. So once a chain loses its longitudinal stiffness it becomes easier for dust and mud to get into that area and act as a friction paste. So actually the degree of sideways flex may be as important as longitudinal wear when checking chain condition........and effective cleaning together with cassette design/ chain line the biggest chain wear differentiators.

......and T Type also varies from other set ups by having a 55mm chainline ....most others are 52mm for boost rear end. Even so I avoid using gears 1 and 12 ( I really do not need 12 speed on an emtb!!)

 

[steve_sordy]

................ So actually the degree of sideways flex may be as important as longitudinal wear when checking chain condition....................

Good point! I will attempt to find a way to measure sideways flex on a new chain that is repeatable and accurate enough for the job. It will be interesting to see how it changes with mileage and "stretch".

 

[Jackware]

Come on Steve, you've had all afternoon to change the chain, we want to know if you've solved today's mystery!

 

[Stihldog]

Maybe it’s been mentioned, or you have checked, but if the chain has jumped one cog on a narrow- wide chainring, that could cause a rumbling noise as you peddle.
Just a shot in the dark?
And I like mystery’s also.

 

[steve_sordy]

Come on Steve, you've had all afternoon to change the chain, we want to know if you've solved today's mystery!

I was out on the trail for five hours today, the chain can wait.

 

[steve_sordy]

Maybe it’s been mentioned, or you have checked, but if the chain has jumped one cog on a narrow- wide chainring, that could cause a rumbling noise as you peddle.
Just a shot in the dark?
And I like mystery’s also.

Yes, a relevant concern. I have seen people do that. But it wasn't me!

 

[steve_sordy]

I changed the chain today. I fitted the spare Shimano XT chain I had in the garage. The rumbling stopped! So no need to replace the chain ring.

In line with @Mikerb's comment about sideways flex and that I had decided to see if I could measure the flex in a repeatable manner.

What I did was to lay the chain sideways on a flat horizontal surface. Poke 10 links out and measure the sag. I positioned the worn chain (0.3% stretch after 1400 mile) next to it and it sagged more. Getting a photograph of them both with a tape measure next to them proved to be a problem and believe it or not, the pic below was the best I could do after almost a dozen attempts! The whole picture is at an angle, but you can use the counter top in the back ground as a guide.

I printed off the pic and did some measurements.
The new chain sagged under its own weight by 3.0 cm.
The worn chain sagged by an additional 1.4 cm.

As expected, the worn chain sags more, but now we have a datum.

 

[CaptChip]

Get a Park Tools CC2 chain checker, it's easy peasy and super accurate. On my 2022 Giant Trance X Adv. +0, I get 500 miles of single track out of a chain and change its out at .5. I use KMC E12 chains and I find that even a brand new chain reads .25. Also I get 1K miles on a cassette and 2K on a chain ring.

 

[Shorty4]

I've had the same problem on my Jam with E8000 which is just under 10,000 kms now and while fitting a new chain dropped the noise level it was still fainlty rumbling so changed the chainring and that cured it completely. I've been fitting alloy chanirings ever since and change them when I change chains

 

[steve_sordy]

Get a Park Tools CC2 chain checker, it's easy peasy and super accurate. On my 2022 Giant Trance X Adv. +0, I get 500 miles of single track out of a chain and change its out at .5. I use KMC E12 chains and I find that even a brand new chain reads .25. Also I get 1K miles on a cassette and 2K on a chain ring.

If you measure over 100 links of a new chain on a flat surface the chain WILL measure 50", ie the pitch is 0.500".
One hundred links of a brand new chain WILL NOT measure 50-1/8" (50 + 0.25%)
If I measured the same new chain with that gauge and it told me it was 0.25, I would examine my measurement technique, read the instructions carefully and try again, or suspect that the gauge was faulty.

 

[Mikerb]

Like a lot of chain checkers it measures between rollers over a few links. I use the SRAM TType chain checker for both T Type and normal 12 speed Eagle. T Type rollers have a bigger diameter than normal chains and measures distance between the back side of one roller to the front side of the other and that seems a better system for all chains. The roller pin through the chain link is a press fit.....the roller pin through the roller has to allow the roller to rotate.

 

[CaptChip]

If you measure over 100 links of a new chain on a flat surface the chain WILL measure 50", ie the pitch is 0.500".
One hundred links of a brand new chain WILL NOT measure 50-1/8" (50 + 0.25%)
If I measured the same new chain with that gauge and it told me it was 0.25, I would examine my measurement technique, read the instructions carefully and try again, or suspect that the gauge was faulty.

I see the problem; you should measure the chain under tension, not statically. It should be measured on the bike not on a table,

 

[steve_sordy]

I see the problem; you should measure the chain under tension, not statically. It should be measured on the bike not on a table,

Of course I had it under tension.

Just having the chain still on the bike does not guarantee that it is under tension.
Besides, irrespective of the measuring equipment, measuring a longer length is more accurate.

 

[TimC7]

When the chainring gets worn and hangs onto the chain at the bottom you get a rumble and you can see it bouncing coming off the bottom under power.

 

[Stihldog]

When the chainring gets worn and hangs onto the chain at the bottom you get a rumble and you can see it bouncing coming off the bottom under power.

Sometimes that is called “chain-suck”. There’s a number of ways this can happen. Worn chainring cogs, excessive worn chain, a bent chain link, poor alignment.
My original chaining developed chain-suck. I replaced that with the ethirteen aluminum model, however the alignment was slightly more and caused rumbling against the chain guard. When I removed the brass O ring, behind the chaining, alignment was perfect. A bearing seal kit was installed also, at least once a year.

I eventually removed the chain guard when the cuff of my pants tried a full rotation through the chaining. The plastic guard broke.

Of course I had it under tension.

Just having the chain still on the bike does not guarantee that it is under tension.
Besides, irrespective of the measuring equipment, measuring a longer length is more accurate.

I agree. The small chain measure tools are only a warning. Accuracy can be measured by hanging the chain and measuring the overall length or comparing old with new.

If you have a piece of vertical wood trim in your bike cave you can easily nail 1,2 or 3 small finishing nails high up Let gravity stretch the chain and mark the bottom length for the correct length. (Masking tape and a sharpie). A number of different lengths can be marked for different bikes.

Doesn't take up much room and looks cool

 

[RustyIron]

That is the most difficult area to clean effectively and is only sealed by the presence of the chain links on either side. So once a chain loses its longitudinal stiffness it becomes easier for dust and mud to get into that area and act as a friction paste.

Word!

My last chain had 1,728 miles on it when I sold the bike, so I know a little about maintaining a bike where I live and on these trails. Back in the day, I would clean my chain religiously, because that's what everyone said to do. No matter what I did, I could always feel some grit in the links when removing it from the solvent. That's like lapping compound in there. Those were in the 9 Speed days, and I didn't get a lot of mileage out of my chains.

Nowadays, my "cleaning" consists of holding the chain in a rag as I rotate the cranks. Then I get my wax-based lube, and squirt it on the chain as I rotate the cranks. I'm not one of these dribbler boys--I give it a healthy stream as the chain goes by. And that's the extent of my chain maintenance.

If I look closely at the chain, there's a buildup of wax at the edges of the link plates. I think the wax dam keeps particulates out of the link pins. One could argue that it also inhibits the migration of new lubricant. I can see that, but I'm content with the life of my chains.

On an exposed plain bearing that's lubricated by grease, the grease is pumped into the middle and oozes out to the end. It's this little bit of grease on the outside of the bearing that provides a barrier against dirt and moisture getting in. We want the same thing on our chains.

 

[Mikerb]

we seem stuck in a time warp!! Nobody with any engineering integrity would choose a chain to connect a drivetrain given it will be used in extremely hostile conditions. Given the advances in materials a belt is a far more practical and logical option. Of course it is not going to work with a cassette or derailleur...........but those components are also from the arc!! The problem we have is that too many suppliers benefit from our unique technique of destroying chains, cassette, chainwheels and derailleurs.

 

[steve_sordy]

@Mikerb I used to work for Renolds in the 1970's. At that time it claimed to be the largest manufacturer of chain in the world. It may have been true, who knew what the Chinese were up to in those days?

They had a thing about belt drives, ie "not wanted on voyage". There were still people in the business who were family members of the original Hans Renold and they (allegedly) insisted that all belt drives be replaced with chain drives. Full stop, no argument! To stamp out the billions of chain links they had a lot of very large presses. All had a large motor on the top, connected by a belt drive to the mechanism that powered the press's reciprocating motion. All these belt drives had been replaced with a chain drive. Because of the confined space and the size of the pullies, a multi-strand chain (x4 if I remember correctly) was required. The location on top of the press was not a good environment for a chain drive, neither for cleanliness, nor for maintenance access. Consequently, the chain drives failed frequently, but on an irregular basis. Something had to be done! I was a mechanical engineering undergraduate on assignment to the drawing office for 8 weeks. I got the job of designing, and having made, a chain case for the drive. This had to be capable of being fitted and removed by no more than two men. It had to be sealed against oil leaks and also contain a quantity of oil to cool and lubricate the chain drive. I got it done and it worked, but it was the most stupid thing to do! A belt drive was the perfect solution and we should have kept them. But I didn't get a vote.

PS: It was still in place two years later when I went back for a visit after graduating.