PDA

View Full Version : Debate of the Day: Experimental Chain Tensioner



Stevo
11 September 2018, 2024
I designed this chain tensioner to push down on the chain in order to remove the chain slack. When I viewed the video, as expected, the chain tightens on acceleration, slackens on deceleration. Now I'm debating on whether or not the tensioner should be pushing upwards from underneath the chain instead of downward from above. The logic being that the chain slack will still be removed while the chain stays in the acceleration position, preventing cogging and also reducing wear at the swingarm pivot/slider. I'm interested in opinions.

https://youtu.be/uzIG-G3nhNU

Electro Flyers
11 September 2018, 2111
Whenever possible, the tensioner should be on the slack side(unloaded side) of the chain during acceleration. The one drawback is during regenerative braking, the tensioner will act like the tensioner in the video, and the chain will get looser. If looseness is a problem, I guess a second tensioner could be used opposite the first one.

Ted Dillard
12 September 2018, 0316
ummm... Explain to me why a chain tensioner is necessary? I just woke up and this isn't a fully-formed thought, but it seems to me that you want constant tension on a drive chain so it rides on a consistent radius on the sprockets. That is, you don't want it riding up and down, which, even with a chain tensioner it's going to do... You're essentially running a loose chain and hoping the tensioner is going to keep it where it should be?

Lemmee get some more coffee and wake up a little more, but I'm wondering why you can't just adjust the chain to where you want it and call it a day.

Richard230
12 September 2018, 0600
I like the concept. It reminds me of how Buell tensioned their final belt drive, but I believe their tensioner was located below the belt run. However, you do have to ask yourself why such a tensioner hasn't been installed on other chain drive motorcycles in the past. :confused: After all, motorcycles have had chain final drives for around 110 years, yet automatic chain tensioners have never managed to make it to the market. Do motorcycle manufacturers know something that we don't?

My favorite chain tensioning system is the one that Zero uses. Placing their drive belt drive sprocket in line with the swing arm pivot. :)

Ted Dillard
12 September 2018, 0647
Wait, Richard, are you talking both chain and belt drives? I thought I remembered the Buell being a belt, and the Zero as well. A belt is a different animal, and pretty much needs some sort of tensioning system AFAIK.

...now that I've had my coffee, to speak to the original question, I think the best advice is to position the tensioner so it pushed into the chain oval, which would maximize the contact with your teeth on both sprockets. But re-reading now, I see it is, and would go with Electro Flyers' sage advice. But you know, us trolls (and band members) like to hang together. :cool:

Electro Flyers
12 September 2018, 0844
ummm... Explain to me why a chain tensioner is necessary? I just woke up and this isn't a fully-formed thought, but it seems to me that you want constant tension on a drive chain so it rides on a consistent radius on the sprockets. That is, you don't want it riding up and down, which, even with a chain tensioner it's going to do... You're essentially running a loose chain and hoping the tensioner is going to keep it where it should be?

Lemmee get some more coffee and wake up a little more, but I'm wondering why you can't just adjust the chain to where you want it and call it a day.

Because the axis of the drive sprocket is forward of the swing arm pivot axis, the chain tension moves through a slack-tight-slack action as the swing arm moves through its travel. On ICE bikes, normally this action is not excessive because the drive sprocket axis is close to the swing arm. Look at the position of the drive sprocket on this bike. It's much further forward, increasing the slack-tight-slack action and increasing the need for a tensioner.

Slack-tight-slack action. I think I just found a name for our next album, Ted.

Ted Dillard
12 September 2018, 0856
Gotcha.


Slack-tight-slack action. I think I just found a name for our next album, Ted.

Thinkin' a Reggae groove? :cool:

Michael Moore
12 September 2018, 0908
There are plenty of automatic chain tensioners on motorcycles, but most of them are on cam and primary chains. However, chain tensioners on drive chains are not uncommon on dirt bikes, especially long travel dirt bikes that have more problem with the "slack" at the ends of the travel. Spring-loaded drive chain tensioners were OEM fitment on Bultaco trials bikes starting about 1969.

A front sprocket/pulley concentric with the swing arm pivot gives constant chain/belt tension, but it may not give the chain squat/anti-squat effects the chassis designer desires. The designer has to decide which feature is more important to them.

cheers,
Michael

Stevo
12 September 2018, 1111
Gotcha.



Thinkin' a Reggae groove? :cool:

LOL... I was thinking New Wave genre, maybe Alternative, but Reggae works too!!

Michael hit the nail on the head. Because of chassis constraints, the drive sprocket is at it's location as close as possible to the swingarm pivot as I could get it. I do have a smaller chain tensioner on the bottom run also. I could try to run it without the top tensioner and video it again to see how much difference the slop in the chain makes. I was inspired by the ATK chain tensioners of the past, not Bulltaco. I may also try to flip the spring around and try it from under just to see what happens.

Michael Moore
12 September 2018, 1138
With a rear suspension set up like a modern dirt bike (or even many modern street bikes) there is no slack-tight-slack because the axle/pivot/counter shaft are never in line. Instead, the high pivot means the chain is always tightening as it goes from droop to bump. In this case you can often find a point for a stationary roller on the lower run that will take up the slack at full droop and as the suspension compresses and the chain gets tighter the lower run lifts away from the roller. If you can arrange that then you don't need to have a spring-loaded tensioner.

The ATK system was designed to have an effect on squat characteristics which the spring loaded tensioner won't have. Here's Leitner's patent:

https://patentimages.storage.googleapis.com/cf/7f/b7/a1e882d5164a3c/US4299582.pdf

cheers,
Michael

Stevo
12 September 2018, 1152
It looks like his design pivots in the center allowing for chain tightening as the rear suspension compresses. Applying my tensioner's spring load the opposite of the way I have it now, will be a similar effect, I think.

Electro Flyers
12 September 2018, 1553
With a rear suspension set up like a modern dirt bike (or even many modern street bikes) there is no slack-tight-slack because the axle/pivot/counter shaft are never in line. Instead, the high pivot means the chain is always tightening as it goes from droop to bump. In this case you can often find a point for a stationary roller on the lower run that will take up the slack at full droop and as the suspension compresses and the chain gets tighter the lower run lifts away from the roller. If you can arrange that then you don't need to have a spring-loaded tensioner.

The ATK system was designed to have an effect on squat characteristics which the spring loaded tensioner won't have. Here's Leitner's patent:

https://patentimages.storage.googleapis.com/cf/7f/b7/a1e882d5164a3c/US4299582.pdf

cheers,
Michael

MMs right. I forgot this is a dirt bike with the always or nearly always downward angled swing arm. Like he says an adjustable but non-articulating(no spring action) idler pully(no teeth) pushing up on the bottom run of the chain would probably be the way to go. I'm thinking you first want to to establish the best drive ratio and then size the sprockets for minimum chain contact with the swing arm slider?

Spoonman
13 September 2018, 0339
Because the axis of the drive sprocket is forward of the swing arm pivot axis, the chain tension moves through a slack-tight-slack action as the swing arm moves through its travel. On ICE bikes, normally this action is not excessive because the drive sprocket axis is close to the swing arm. Look at the position of the drive sprocket on this bike. It's much further forward, increasing the slack-tight-slack action and increasing the need for a tensioner.


Not sure I agree with this - the alignment of the swingarm pivot, drive sprocket and driven sprocket has a bearing certainly; assuming good alignment, for any reasonable measure of suspension travel, the distance from the swingarm pivot to the drive sprocket does not (as long as it's still much smaller than the length of the swingarm itself). Hence, the need for a chain tensioner does not arise as a result of the nature of the mechanism, but rather on account of an angular deviation from 'straight' between the line made by the swingarm under static sag, and line between the swingarm axis and the drive sprocket at the same time.

As as result, the notion of a tensioner is a tailored solution to deal with a consequence of a given drivetrain layout, and not something that would ever see broader application unless there some very good reason why, in design, the alignment issues couldn't be addessed.

Eyeballing the alignment in that video, I can see no good reason for the amount of slack present in that chain.

Ted Dillard
13 September 2018, 0832
A drawing would be nice... :rolleyes:

(My visual imagination is on the fritz right now lol)

Spoonman
13 September 2018, 0929
As requested Ted (and in nice Crayola colours too ;) ).

It all comes down to the ratio of the lengths and the manitude of the Cos scalar. As long as the L1 << L2 then you're never going to be too sensitive to any reasonable deflection.

https://lh3.googleusercontent.com/OUeRuFMZMOutcmmBuNP6JDfbKTOa5FYvTrJBcyKGW_JQA4_aJI V4_cVEv0TpuwYje2XWmasEKjVXzSb4YEql8eLX5OAUR7f3BQv6 9CR5YXDS1lGq8JNO85AZXWNVy2LKMZ-0wmC9C3XMRykEFLcjW21tY5jRA_pbWfb_DtkASzvXKgMHATGis F4V14OmcfWd3lvqsNlcr42RiJDdqOw9MjGKAcGXaXl9lBea1lJ jkIKs8S7lrotBKWVGkD-2_q1TAHUYAnYxtBPVaP7nFyX_kyMgtNU1R3n0x9fLH33fhk3ib pmfD0EkXbfachkydSe2Aijl8hBghaf7_nMwsfqDgFlSIlKM9Ly Ap1j_6XdFpkLDyl3jChEgLq1iplfqDwtaAzEL3XrAm7RGnl3rb D3TD-LuShiBFPjr4qHfXXv_H6JjxQalSNQP0SCMlZc6qeQQ6RMeDV0t OoWV0cqts25meOfh4UjqdfdnJOepCKGi_SnkW7ZiIEnj8S0XZL XI8XYxzZt3juAk6cXc1dm8eGwHHtmVeaEXFJYeKFoV17jF0j3b adel03z757yNc8gSQVRTveXhSObc3ZYdJhi1Djy-YcMk1v64UNCmzM5n_G5Up7IcLxeeWKA8cgeBl0NkKR1D=w1522-h822-no

Stevo
13 September 2018, 0954
That's nice Spoonman!
Here is my setup (not to scale):
7958

The stationary "idler" is there to guide the chain between the swingarm pivot and chassis frame crossmember in order to prevent chain wearing on either. There is and needs to be chain slack. As the rear wheel travels up into compression, the chain tightens. That's why when you adjust chain tension for routine maintenance, there must be set slack and your adjustment shouldn't be too tight. Right??!!

Ted Dillard
13 September 2018, 1012
HIS drawing is more better than yours is.


;)

(well... this IS fight club, RIGHT?)


yeah, feeling kinda stoopid today, forgive me...

Stevo
13 September 2018, 1017
Who's arguing?!
I asked for opinions... that could be flaming as well as constructive criticism or suggestions. I really don't know what the optimal way is for me to go with this. I do know that there is a lot of chain slack that needs to be there... I just need to figure out how to deal with it.

Aww Hell! Maybe I'm overthinking this too much!!

Ted Dillard
13 September 2018, 1041
LOL

Well, my first reaction, which isn't at all helpful, is that it sure looks like that chain goes around a lot of corners. Like I said, not helpful...

furyphoto
13 September 2018, 1049
As requested Ted (and in nice Crayola colours too ;) ).

It all comes down to the ratio of the lengths and the manitude of the Cos scalar. As long as the L1 << L2 then you're never going to be too sensitive to any reasonable deflection.

7959

Can I just say than my understanding of swingarm alignment, and chain deflection just increased about 10000% from seeing this diagram.

I am in this phase now, I need to add new suspension mounts on my swapped out swingarm, and finalize my motor mount for the drive sprocket position (a tiny bit forward to allow space for future motor upgrade) so this has been on my mind. Now I'm excited to figure out my AUW neutral suspension position, and see if the chain will rub on deflection.

In your formula, do you have an ball park idea of what actual ratio is represented by "much less than" ( << ) for an averageish motorcycle, before slack becomes excessive?

Stevo
13 September 2018, 1049
LOL... ya but they're round corners!!

Stevo
13 September 2018, 1051
Glad to see you back at it Andrew! Cheers!

Richard230
13 September 2018, 1254
What I like to do is to disconnect the rear suspension, place the bike on a stand so that the rear wheel is off the ground and free to move and then move the rear wheel and swing arm through its maximum travel to visualize what is happening with the slack at the two extremes of movement and when the axle is in line with the swing arm pivot and the counter shaft sprocket. With that information you can set the chain slack appropriately when the rear wheel is resting on the ground or wherever you typically check and adjust the chain slack.

Stevo
13 September 2018, 1329
Correct, for initial set-up without any chain tensioners. I need to find a replacement for the stationary idler when it wears down...
How about these, Ted Dillard and the ElectroFlyers? https://www.ebay.com/itm/Skateboard-Wheels-102A-50mm-Rasta-Tri-Color-with-HYBRID-CERAMIC-BEARINGS/391773764287?epid=1576196842&hash=item5b37893abf:g:T3kAAOSwcZFbLBm3
;)
Rasta Man!!

Richard230
13 September 2018, 1444
Correct, for initial set-up without any chain tensioners. I need to find a replacement for the stationary idler when it wears down...
How about these, Ted Dillard and the ElectroFlyers? https://www.ebay.com/itm/Skateboard-Wheels-102A-50mm-Rasta-Tri-Color-with-HYBRID-CERAMIC-BEARINGS/391773764287?epid=1576196842&hash=item5b37893abf:g:T3kAAOSwcZFbLBm3
;)
Rasta Man!!

Very colorful. ;)

Ted Dillard
13 September 2018, 1528
OMG SO AWESOME

Electro Flyers
13 September 2018, 1743
As requested Ted (and in nice Crayola colours too ;) ).

It all comes down to the ratio of the lengths and the manitude of the Cos scalar. As long as the L1 << L2 then you're never going to be too sensitive to any reasonable deflection.

https://lh3.googleusercontent.com/OUeRuFMZMOutcmmBuNP6JDfbKTOa5FYvTrJBcyKGW_JQA4_aJI V4_cVEv0TpuwYje2XWmasEKjVXzSb4YEql8eLX5OAUR7f3BQv6 9CR5YXDS1lGq8JNO85AZXWNVy2LKMZ-0wmC9C3XMRykEFLcjW21tY5jRA_pbWfb_DtkASzvXKgMHATGis F4V14OmcfWd3lvqsNlcr42RiJDdqOw9MjGKAcGXaXl9lBea1lJ jkIKs8S7lrotBKWVGkD-2_q1TAHUYAnYxtBPVaP7nFyX_kyMgtNU1R3n0x9fLH33fhk3ib pmfD0EkXbfachkydSe2Aijl8hBghaf7_nMwsfqDgFlSIlKM9Ly Ap1j_6XdFpkLDyl3jChEgLq1iplfqDwtaAzEL3XrAm7RGnl3rb D3TD-LuShiBFPjr4qHfXXv_H6JjxQalSNQP0SCMlZc6qeQQ6RMeDV0t OoWV0cqts25meOfh4UjqdfdnJOepCKGi_SnkW7ZiIEnj8S0XZL XI8XYxzZt3juAk6cXc1dm8eGwHHtmVeaEXFJYeKFoV17jF0j3b adel03z757yNc8gSQVRTveXhSObc3ZYdJhi1Djy-YcMk1v64UNCmzM5n_G5Up7IcLxeeWKA8cgeBl0NkKR1D=w1522-h822-no

Here's my take on this debate:
Your analysis and formula oversimplifies the relationship between the swing arm(SA) pivot axis and axes of the drive and driven sprockets. Think of there being a virtual SA(VSA) between the drive sprocket axis and the driven sprocket axis, keeping the chain at a fixed, ideal tension. In the set-up you show, the VSA would always have a greater radius of its swing than the actual SA. And, of course, the length of the radius would increase with the distance the drive sprocket axis is away from the SA pivot axis.

It follows that for any normal given vertical travel away from the inline position, the position of the points of travel of the driven sprocket axis of the VSA would be further away from the actual SA's pivot axis and the points of travel of the axis of its driven sprocket. A longer radius forms a bigger circle. This means that slack in the chain increases, within limits, with the swing of the SA(the actual one) and the distance the drive sprocket axis is from the SA pivot axis. The difference in the slack would not be great. But, the ideal tension of a chain is usually determined by a couple of mm or a small fraction of an inch of adjustment.

I guess this would be covered by a variation of angle alpha in your formula.

Spoonman
14 September 2018, 0604
That's nice Spoonman!
Here is my setup (not to scale):
7958

The stationary "idler" is there to guide the chain between the swingarm pivot and chassis frame crossmember in order to prevent chain wearing on either. There is and needs to be chain slack. As the rear wheel travels up into compression, the chain tightens. That's why when you adjust chain tension for routine maintenance, there must be set slack and your adjustment shouldn't be too tight. Right??!!

OK - now I see the problem.

...and even with the humour intended, the below indeed outlines the missing factor (the undesirable design feature if you like) which I wasn't seeing.


LOL

...it sure looks like that chain goes around a lot of corners. Like I said, not helpful...

Except that it is helpful.
The slack that's appearing in your chain under compression of the suspension is as a result of the need to be able to accommodate the deflection of the chain by that lower idler when the suspension is extended (this is obvious of course, I simply wasn't aware of it/hadn't twigged it). So we can't model your swingarm system in the straight forward sense I've used above, you have a additional component in play. In design, the ideal solution would be to remove that component; but given we're dealing with a hack on an existing design we need to work with it - which explains the need for the tensioner.




In your formula, do you have an ball park idea of what actual ratio is represented by "much less than" ( << ) for an averageish motorcycle, before slack becomes excessive?

I don't but it will be geometry dependent and you can work it out quite easily* if you know how much net shortening you're willing to accept in the compound length (hypotenuse of the triangle formed by the three points with the swingarm at maximum displacement).

*edit: not since I revised the formula you can't - I'll try and come up with a restatement of the equation which will relate a single term angular deflection of the swingarm to net length of the system (which will make it a lot easier to use)... expect there's an elliptical equation which will do it.. it'll likely be something second order in any case but leave it with me.

Spoonman
14 September 2018, 0623
Think of there being a virtual SA(VSA) between the drive sprocket axis and the driven sprocket axis, keeping the chain at a fixed, ideal tension. In the set-up you show, the VSA would always have a greater radius of its swing than the actual SA. And, of course, the length of the radius would increase with the distance the drive sprocket axis is away from the SA pivot axis.


Agreed



It follows that for any normal given vertical travel away from the inline position, the position of the points of travel of the driven sprocket axis of the VSA would be further away from the actual SA's pivot axis and the points of travel of the axis of its driven sprocket. A longer radius forms a bigger circle. This means that slack in the chain increases, within limits, with the swing of the SA(the actual one) and the distance the drive sprocket axis is from the SA pivot axis. The difference in the slack would not be great. But, the ideal tension of a chain is usually determined by a couple of mm or a small fraction of an inch of adjustment.

I guess this would be covered by a variation of angle alpha in your formula.

You're right!
I had actually meant to encompass that into the formulaic result but I've not actually done that successfully in that first equation.
Despite this, for L1 'much less than' L2 (must be some forum code overlap with the double less than notation, can't get to display properly) it would still produce quite an acceptable result but the more accurate approach is as follows:

L=sqrt[(L1 + L2Cos(alpha))^2 + (L2Sin(alpha)^2]

Where:
L -> net length between sprocket centers
L1 -> distance from drive sprocket center to swingarm pivot center
L2 -> distance from swingarm pivot center to driven sprocket center
alpha -> deflection of swingarm from 'inline' with line formed by the pivot and drive sprocket centers

I'll revise that graphic and reupload.

Spoonman
14 September 2018, 0631
Proper formula amended.

https://lh3.googleusercontent.com/OU4hj8qSFf91-PpLVyDobOmIS3cCv5ESyFXGYhF54qh8eCnCg1Ai-MZOWrNg7mJs48NxQAsZSX8t2pSXYWpCHBa8ElR2DpClObD2m-17TVOZ7Eq-LvzEce2UseqlV7Af8ZN8SKBHewxmVadlJSLIs10XRV3a-725SZFAXjP4oX7BXqLJu2gLqjd9RkXLZoFsPzAMDO3wdu4TMr4 e0Ttm8H-Q75vqGqRUyNH5TwUXAmasr-MbB9leStf1tzHrn72FeQdqQRLPq48a6J1gpUt6NkzUvV2PJNJ3 EBNhI88r1qVXhBof5hmYRLZ2KuYBaIau6j4ZSKPU98bD2TpVB8 ency9JMlbljUHeTpb_dVQSKJjzHZtJ8vIdIHDRFonaWPXV9c8R d9qC_779JnBJKOgxLiJuXENcnLQry4L2AabxpqgiQJ5n2IyO-uF3CV6ls3iwgvmM0n4PlOIjPeAtdZOxS8qbUdgDWbmJKRY6Ewp InR5IeKjSoIiuFdH4WeITtrAnoFfJ5mgmOjfaROFcXe0FjFMPk _6YUUDLY1B4Bw17M_F8BwZh4WcHqEgOfb-Lan4YaWIoJN2hEr_5ebJNo_xxzYzb_OKF9b6iOHOkV7ilydQfJ yJax3u_oVN2tih6QKxE=w1522-h822-no

Spoonman
14 September 2018, 0637
Stevo: that idler pully - is that just to take the chain over the top of the center stand support bracket?

...if so, is just cutting that out altogether an option?
Can't imaging that chain tensioner (and the maintenance and noise that will come with it) is any better value than an abba stand...

Ted Dillard
14 September 2018, 0752
is any better value than an abba stand...

DID SOMEBODY SAY ABBA????


https://www.youtube.com/watch?v=xFrGuyw1V8s

(oh god i'm so sorry. hell, it's friday and i haven't even started drinking yet... ;) i probably better put up bail money right now.)

Electro Flyers
14 September 2018, 0957
DID SOMEBODY SAY ABBA????


(oh god i'm so sorry. hell, it's friday and i haven't even started drinking yet... ;) i probably better put up bail money right now.)

That's it Ted. I'm going solo the next album. I think you're going to start drinking because you're worried about how to incorporate, and explain this complicated chain formula in the next edition of your book.

Stevo
14 September 2018, 1056
Listening to Abba tunes will definitely lead me to drinking...:rolleyes:

I'm more of a visual guy, but I know the importance of math, so thanks Spoonman for your effort!

The idler is a preventative measure... without it, the chain will chainsaw a hole through the frame cross support in shortorder.
There was originally a stationary idler on the top run above the swingarm pivot, mainly just to prevent the chain from rubbing the rear shock reservoir. I'm thinking of utilizing this as another "corner", which I suspect will control the slack... see new illustration :
7960

I suspect pushing upwards on the chain would be better than pushing down on the chain, or removing this chain tensioner idea altogether, as it was oem style. It just seems like there is a lot more slack now than there was before with the new motor install, which I believe is due to the increase in L1 in Spoonman's calculations.

Spoonman
14 September 2018, 1255
It just seems like there is a lot more slack now than there was before with the new motor install, which I believe is due to the increase in L1 in Spoonman's calculations.

In the case of this arrangement, an increase in L1 will have a disproportionate impact on the chain slack on account of the narrowing effect it has with respect to that window of travel in the swingarm before the chain starts to get deflected by the idler pulley - so yes, the increase in L1 will most certainly be the source of the problem alright.

I really would be considering moving or refabricating that crossing brace in the frame to be honest - it *is* the best solution even if it's a good bit of work. You'll end up with a quieter, more efficient, and lower maintenance drivetrain as a result.

I don't expect it's actually doing a whole lot more than carrying the center stand, particularly given the proximity to the swingarm throughbolt.
The bike would also have likely originally had a couple of engine mounting through bolts not too far away from that location as well so I reckon its importance as a structural member is limited.

If you're worried about it then just weld in a new one 4-6" forward of it, or if you want to retain the support at the corner then maybe something in a 'U' shape such that the actual cross-tie is still closer to the motor and hence not fouling the chain.

Electro Flyers
14 September 2018, 1341
In the case of this arrangement, an increase in L1 will have a disproportionate impact on the chain slack on account of the narrowing effect it has with respect to that window of travel in the swingarm before the chain starts to get deflected by the idler pulley - so yes, the increase in L1 will most certainly be the source of the problem alright.

I really would be considering moving or refabricating that crossing brace in the frame to be honest - it *is* the best solution even if it's a good bit of work. You'll end up with a quieter, more efficient, and lower maintenance drivetrain as a result.

I don't expect it's actually doing a whole lot more than carrying the center stand, particularly given the proximity to the swingarm throughbolt.
The bike would also have likely originally had a couple of engine mounting through bolts not too far away from that location as well so I reckon its importance as a structural member is limited.

If you're worried about it then just weld in a new one 4-6" forward of it, or if you want to retain the support at the corner then maybe something in a 'U' shape such that the actual cross-tie is still closer to the motor and hence not fouling the chain.

Ditto on the cross brace mod or move for chain clearance. Looks like a steel frame-easy to weld. Although it looks like you've done a fair amount of aluminum welding. Something like this a good mod: 796279637964

In this case, this is a top view of a section of a swing arm round tube replaced by a bent piece of heavy section flat bar. The tube was in the way of a jack shaft sprocket mounted concentric with the swing arm pivot for an ongoing E-bike project. For strength, keep the piece that replaces the tube fairly massive. Although massive, you'll only add a small amount of weight for a small replacement piece.

Stevo
14 September 2018, 1717
It's hard to visualize from the pics how the belt travels through that jackshaft and drive pulley. Since it's a belt drive, there must be a tensioner somewhere in that design.

Sent from my Pixel 2 using Tapatalk

Stevo
14 September 2018, 1724
I disagree with both of you about not needing the cross support. The next place up the frame that cross supports right and left chassis sides is at the top rear shock mounting point. I already welded a cross support 6" forward to support the battbox bottom mounts. There is no centerstand....it's a dirtbike.

Sent from my Pixel 2 using Tapatalk

Electro Flyers
14 September 2018, 2012
I disagree with both of you about not needing the cross support. The next place up the frame that cross supports right and left chassis sides is at the top rear shock mounting point. I already welded a cross support 6" forward to support the battbox bottom mounts. There is no centerstand....it's a dirtbike.

Sent from my Pixel 2 using Tapatalk

I'm not suggesting you REMOVE the cross support. I'm suggesting you MOVE or MODIFY it. If the chain is rubbing on the cross support, is this because of the large rear sprocket? If so, you could put a fixed idler on the top part of the chain to keep the chain from rubbing on the cross support. Because it is on the main load side of the chain, it will be subject to heavy loads, possible making a lot of noise and early failure. As I recall, you wrote there was an idler in this position originally to protect part of the shock. What I am suggesting is that if you are confident in your welding and fab skills, you could move the cross support or modify it to clear the chain so that no idler is needed. If there are other rub points those too could be moved or modified

I successfully do this kind of stuff all the time. If you don't feel like you have the skills to do this properly, don't do it. Maybe, you could find someone else to do the work.

Stevo
14 September 2018, 2124
As you say, the main load side of the chain is the top run (witness the video of chain tightening on acceleration.) The bottom run may see light load on Regen, which is programmed to pretty low levels, almost insignificant, hardly noticeable. I don't see the bottom idler or the cross member being the issue, as the bottom chain tensioner controls the bottom run's slack without any issues whatsoever. It's the top run that's proving to be the headache.

Sent from my Pixel 2 using Tapatalk

Stevo
14 September 2018, 2136
No I'm not moving that bottom support. I agree with the original builders (the Vertimati brothers) that is where that support belongs.

Sent from my Pixel 2 using Tapatalk

Electro Flyers
15 September 2018, 1224
No I'm not moving that bottom support. I agree with the original builders (the Vertimati brothers) that is where that support belongs.

Sent from my Pixel 2 using Tapatalk
I should have written MOVE or MODIFY
Do you mean the cross piece that interferes with the top run of the chain? What I've done in these situations and have seen the factory do to clear the chain is
(1) angle the cross piece up to clear the chain
(2) put a S curve in the X piece (use a larger diameter, stronger piece of tubing if you do this)
(3) if there's room and the chain has a master link, put the X piece between the chain runs
(4) Like in the photos in post 36, on a swing arm in this case, replace part of the X piece tube with a piece of large cross section(thick) angle iron or bent large cross section(thick) flat bar

Electro Flyers
15 September 2018, 1325
It's hard to visualize from the pics how the belt travels through that jackshaft and drive pulley. Since it's a belt drive, there must be a tensioner somewhere in that design.

Sent from my Pixel 2 using Tapatalk


No tensioners. The primary belt from the motor to the jack shaft is tensioned by moving the motor in slots under the motor mounting bolts. The secondary belt (the sprocket in the photos is the front sprocket for that belt) from the jack shaft to the rear wheel sprocket uses the stock slots in the swing arm to slide the rear axle back to tension the belt. With the jack shaft turning concentric with the swing arm pivot, the belts' tensions remain the same through out the swing arm travel. I've noticed some OEM electric dirt bikes use a similar system, but with a chain secondary drive. This bike is waiting for the right motor/controller/battery combination, and of course $, to get finished.

Stevo
15 September 2018, 2335
I now understand your setup better now. I'm trying to keep my drive as close to oem as possible. There is no crossbar on the top run...the only things are the rear shock reservoir and rear break line cross over... which are protected by the idler wheel attached to the frame. The chain makes contact with that wheel on extreme load and compression. There has to be a pivot there in order to protect those other components. You gotta remember this is dirt bike... a lot more travel than a street bike. So I'm waiting for some parts and I will shoot some more video with a few different solutions I have in mind.

Spoonman
17 September 2018, 0238
OK, so even if the lower support was gone, you still need to limit upwards travel, which means you need an additional element in any case and so the whole cycle reverts.

Have you considered mirroring the idler that's already in place?

I'm not certain about it, but I expect that would remove L1 entirely from the chain length calculation and effectively bring your chain pivot point into line with the swingarm pivot meaning that you get zero (or very close to it) net change in system length with +/- ~50deg travel about the aligned position. Reckon that's enough even for a dirtbike... and may remove the need for a tensioner.




Of course you could likely achieve very close to the same net effect by increasing the stiffness of the pulley you have in your video and taking a link out of the chain...

Stevo
18 September 2018, 1033
So I got to thinking about chain derailers on mountain bikes/street bicycles and how the "double-ider" is utilized to remove/control all of the required chain slack in those drive systems. The wheels are spinning (in my head).

Spoonman
20 September 2018, 0247
Dunno about that notion dude.
With the weight of the chain and its velocity at even relatively slow speeds, I expect it'll just be a mess - you'd be better staying as you are at that stage.
I still reckon you could stand to loose a link in the setup you have, trim it out using the adjustment available, and up the tension on that tensioner a good chunk - it doesn't seem too far from a workable solution as it is.

Ted Dillard
20 September 2018, 0541
^ What he said.

Have you checked out conventional industrial solutions? I'd think they'd be a lot more to the point than bicycle products. See this: http://www.brewertensioner.com/tensioners-positioners.html and this, for starters: http://www.roll-ring.com/?lang=en

I know you like the movies, so here:


https://youtu.be/-qhVOW--QSs

...also this:


https://www.youtube.com/watch?v=QPaz7qYTYWM

(complete with Happy Music)

Stevo
20 September 2018, 0951
That Rollo-Ring is a pretty nifty invention, but I don't think applicable here. And gocart chain adjustment is not the same animal, without 6" suspension travel

What I really really need to find is a 520 sprocket adapter for the 7/8" keyed electric motor shaft. Nobody makes a solution AFAIK.
I was going to meet with my local machinist about prototyping one up. I ordered a JT sprocket for a BMW G450X... the inner diameter should fit over the shaft diameter no problem. They are available in 13, 14 and 15 tooth configs.

Stevo
27 September 2018, 2309
Today I had time to shoot some video of the new chain installed by Richard's suggestion of removing the shock to see how the chain slack changes with rear wheel travel and no chain tensioners installed. The chain is at it's tightest length when the suspension is fully extended, and gets much more slack as the wheel moves upwards. The first part of the video is at HWY speeds of 65 mph, the second part is during fast acceleration runs over big ass potholes and roots. I wanted to see how the chain behaves with big suspension compressions. My experimental chain tensioner has been modified to push upwards on the chain instead of downwards as before, but I didn't have time to video that test run yet. The only chain tension being used on this video is the bottom run, which has a chain tensioner pushing up.

https://youtu.be/NJ3UrSPcEMI

From my own analysis, and since I posted this thread in The Fight Club! please feel free to add your own critique, is that with the chain adjusted at the bottom of the shock stroke, there continues to be a considerable amount of chain slap on the top run. The bottom run looks smooth as butta. I like it. But the top run has too much chain slap for my liking. There is a protective roller located on the bike frame that you can't see in the video, to keep the chain from hitting the shock reservoir. I noticed some chain damage to it on inspection. I'm gonna try and get some video of it from a different view. The next video should be interesting. I gotta go make a gopro mount now, but first, back to fighting....

One potential fight is the belt vs chain argument. Another potential fight could be the 520 vs 525 vs 530 chain size argument. I chose 520 chain over belts. I wanted to keep as close to OEM as possible. It's easy to get rear sprockets made but man is it PITA to get a front sprocket! Nobody makes what I need! Travis can get a 10 and maybe 11 tooth in 520, but I think a smaller sprocket would compound my chain tension issue.
Another fight can be over sprocket choices. I got a 15t Martin on the drive now. I like the acceleration and the throttle feel at low speeds. A 14t has waaaay better acceleration. Exhilarating acceleration is how I would describe it. But it's also scary, as the back wheel can break loose by twisting the throttle :p The 15t will too, but it has a better feel to it. A more controlled feel, but noticeably slower to top speed, which is around 80ish, the 14t will take it to 85ish. A 13t??? I didn't try!

frodus
28 September 2018, 0809
What I really really need to find is a 520 sprocket adapter for the 7/8" keyed electric motor shaft. Nobody makes a solution AFAIK.


I have 10t 7/8" for 520 in stock:
https://www.emf-power.com/product/10t-520-sprocket-for-7-8-shaft/

Stevo
28 September 2018, 1040
I have 10t 7/8" for 520 in stock:
https://www.emf-power.com/product/10t-520-sprocket-for-7-8-shaft/

Thanks for your support Travis. I think it would be too small for my liking. I was editing my post above^^^ while you posted this, so you probably didn't see the context.
I did find, ordered and received a 14t JT sprocket for an idea I have. I'm drafting up a sprocket adapter for these 7/8" keyed Motenergy shafts. My idea was to be able to get an over the counter model specific sprocket and use it on the adapter. The problem is I only found 2 sprockets that I thought would work. This sprocket that I just got is for a BMW G450X dirtbike. It fits nicely on the shaft and only a keyway needs to be cut out. The problem is that the sprocket's center splines need to be a much larger diameter in order to be mounted to an adapter. My idea wont work because the sprockets will have to be proprietary and made to order, which defeats the purpose of my original idea. The only larger sprocket spline diameter I could find was for a Honda RC51, and that would only give 2mm of meat to play with, not enough.

I can use the BMW sprocket by cutting a keyway, then welding it to a 10t Martin, which I just happen to have in my garage.:rolleyes:

frodus
28 September 2018, 1539
Can you machine down a #50 sprocket (530 chain size) to what you need?

www.surpluscenter.com has some under "Power Transmission"and "finished bore sprockets"

Michael Moore
28 September 2018, 1828
520 is good enough for Superbikes and probably MotoGP as well, so go with that instead of the wider chains. Of course, that presumes you get a good 520 chain, Regina actually makes 6 different 520 chains in their "GP" line of chains for racing, and other 520s for lower level uses.

cheers,
Michael

Michael Moore
28 September 2018, 1836
Can't you buy some sort of adapter made for the shaft, bore the center out of a sprocket and weld the two together? I've retoothed rear aluminum sprockets by turning off the teeth and boring out the new sprocket and welding them (intermittently) at the join.

Carbide will cut through the case hardening on the steel sprocket. You can also fit the two together and drill holes at the border between them and put close-fitting steel pegs in the holes to take the drive loads and then weld everything up if you want a little more security. Another way to do that is to cut a wavy "sine wave" pattern on both pieces so the weld isn't taking the loads, just holding things together.

If you've got room leave a shoulder on the adapter for the sprocket to mate against to help keep it running true.

Stevo
28 September 2018, 1844
I agree with you Michael. Switching my 530 chain out to 520 on my VFR probably shaved off a few lbs in unsprung weight. The difference in strength is arguably minimal. I'm actually kinda surprised nobody else on here has started a discussion regarding the lack of available sprockets for these motors. 520 is probably the most common chain size used on motorcycles. 530 the next most common.
What I've done in the past is use a cutoff disc and slice a sliver off of a Martin #50 sprocket, then grind the teeth profiles all by hand. It works. It's cheap. It's also a PITA.
Not to mention that the surface hardened heat treat is ruined when you do this. The heat treat will probably be ruined if I weld a sprocket onto a Martin also. Is there no other better options available to us? :mad:

Michael Moore
28 September 2018, 1934
Sprockets aren't rocket science, you've just got to find a place willing to make them at a price and quantity you can afford. For big production numbers sprockets are hobbed or shaped, but I've seen mention of places doing BMX/MTB stuff that just CNC sprockets (I suspect aluminum) on a mill. I've got software that will generate a DXF of different pitch/tooth count sprockets and I've got 520 x 14-18T DXF already on hand (those are actually just 5xx curves, I extrude them as thick as needed for the xx width). I've attached various 219/415/428/5xx dxfs.

Use at your own risk!

If you need a hand with the CAD I can probably assist a bit.

cheers,
Michael

Michael Moore
28 September 2018, 1947
A bit of Googling found ANSI 50 taper lock sprockets starting at 15T

https://www.rainbowprecisionproducts.com/sprocket-ansi50#1

and they use these bushings

https://www.rainbowprecisionproducts.com/taperlock-bushing#1

It looks like $26 gets you a 15T with a 7/8" bushing.

cheers,
Michael

Stevo
28 September 2018, 2012
I've seen the taper locks before too, but they are still sized 530 and not really a solution .

Your other suggestion of CAD design help may work with Aluminum. They will probably need to be replaced more often. But we could collaborate and design a prototype sprocket and adapter pairing to have it cnc'd somewhere. Are you game?

Sent from my Pixel 2 using Tapatalk

Michael Moore
28 September 2018, 2048
530 needs 1/8" turned off one face to turn it into 520. That doesn't touch the hardened surface on the teeth where the rollers ride and would be much easier to do than making sprockets.

An aluminum chain sprocket (as opposed to an aluminum belt sprocket), especially one with too few teeth per the chain manufacturers recommendations, coupled with the torque of an e-motor, sounds like a bad plan to me. Every stress cycle on aluminum brings it that much closer to fatigue failure. Rear sprockets get away with it by having a lot more teeth engaged but even so an aluminum rear sprocket won't last as long as a steel.

I'm willing to assist you on the design but modifying a decent sprocket/hub that someone else already makes in quantity is probably going to be a much cheaper and faster to realize solution, as well as a better one for longevity. You could buy a sprocket and hub and put the sprocket in a lathe (or on a mill table) and have it thinned in 10-15 minutes. It is what I'd do if I were going to run a chain.

On these keyed shafts I hope the hubs are a tight reamed fit. A hub that also had a taper lock on the shaft (like an ER or 5c collet) strikes me as a good idea if you can't clamp the sprocket on tight enough with a nut/bolt to keep it from moving back and forth under load reversals causing fretting.

cheers,
Michael

Stevo
28 September 2018, 2101
I agree... I don't believe Aluminum is the right material to use for the sprocket or the carrier.

Sent from my Pixel 2 using Tapatalk

Richard230
29 September 2018, 0628
I agree... I don't believe Aluminum is the right material to use for the sprocket or the carrier.

Sent from my Pixel 2 using Tapatalk

I agree. Instant wear. Anyone recall the days of the nylon rear wheel sprocket and oil-less, non-O ring, no maintenance chains? The motorcycle magazines were touting those clean and light chain drive systems for a few years about 20-30 (?) years ago. :cool: Then everything went silent - because the nylon sprockets wore out very rapidly or stripped their teeth, followed quickly by the self-lubricating "sintered" chains. :O Another bright idea and start-up industry down the drain. :rolleyes:

Stevo
29 September 2018, 1016
I do remember those too. That was basically an idea straight into production without any real world testing and R&D. It questionably may have worked on a bicycle, but I doubt it. Ya I was dumb and was suckered into buying into the test group:rolleyes: ...tried one on my 1st bike 8027 circa 1980
It didn't last a week LOL I returned it for refund.
I thought Teflon may have worked, but probably ultimately not tough enough. Maybe graphene infused teflon??

Stevo
05 October 2018, 2308
It looks like this is the best I can do with controlling the chain slack. It appears to be working pretty good... I'll see how long these ceramic bearings last.

https://youtu.be/ws7MASKnSQc

Ted Dillard
06 October 2018, 0348
That's kind of brilliant... in a sick and twisted way, of course! :D

Good job - it will be interesting to see what, if anything, fails. Looks pretty beefy to me, though.

Richard230
06 October 2018, 0602
That is a neat solution to accommodating chain slack as the suspension moves. However, I will make one observation: All of the street/dual purpose motorcycles that I have seen don't use an active chain tensioner but instead use a replaceable plastic rubbing rail along a portion of the upper swing arm to keep it from being damaged as the chain rubs against it. Now that is either an indication that you don't need an active tensioning device to control chain slap - or that the manufacturers selected the cheapest method of dealing with the issue and that they don't think chain slap is a safety problem and just want to protect the swing arm from damage. :confused:

Stevo
06 October 2018, 0806
I agree with you Richard... and on the prior video with inactive top tensioner, it appears to be ok to be run without it. This company went bankrupt in 2003-4 and getting consumable parts is near impossible now. Brake pads is not an issue as Husqvarna and KTM uses the same, but the swingarm slider is unique. The bottom rear chain guide will be an issue when it wears out. I may have to fab one up out of teflon, or modify something from a different make/model. It's really hard to find fork seals for it...last time I only found 'em in Europe and had to have 'em shipped. All said, it's one of the main reasons I converted it to ev. I'm also a glutton for punishment :rolleyes:

Richard230
06 October 2018, 1741
Stick with it, Stevo and good luck keeping your bike on the road. :)

Michael Moore
07 October 2018, 0937
Stevo, teflon "cold flows" under pressure. A strip of UHMW-PE would probably be a better choice and it is easy to find.


Extremely tough abrasion resistant, low cost plastic, used for a wide range of wear applications. UHMW (ultra high molecular weight polyethylene) is an extremely tough plastic with high abrasion and wear resistance.

cheers,
Michael

Stevo
10 October 2018, 0856
Thanks Michael, for your expert opinions and suggestions. I wasn't sure if "pressure" and "friction" are the same. I thought for chain guides, friction was the main abuser. I have a small, 4"x6" 1/8" thick piece of teflon in the garage somewhere that I was going to try a test with when the time comes, but I like your suggestion better. Thanks

Michael Moore
10 October 2018, 1740
If you've got some Teflon on hand, by all means give it a try and monitor it. With intermittent contact between it and the chain cold flow may not be an issue.

https://en.wikipedia.org/wiki/Creep_(deformation)

There are soooooooooo many different plastics available (I know a chemist who specializes in plastics) so garage tinkerers like us have our work cut out to try and pick an appropriate materials. Some of them absorb water (and change size while doing it), others cold flow, some are a PITA to machine (you may need to freeze them first). Garden variety acetal may not have the same properties as a name brand version (Delrin) but does it make a difference in our applications? UHMW-PE is what was recommended to me for one-off rub strips by that road-racing chemist but it wouldn't surprise me if there was something even better that I've never heard of.

I think we are all well advised to make our best selection of a material and then keep an eye on how things are holding up for "just in case" issues. PTFE may be great stuff on cookware, but that doesn't mean it is ideal for using on a motorcycle (maybe if you use bacon grease to lube the chain . . .).

cheers,
Michael