PDA

View Full Version : Cogged belts



QSIMDO
02-22-2011, 09:05 PM
What sort of tension is necessary on a cogged belt drive?

I have a motor drive with a cogged (tooth) belt and I'm thinking it's a tad loose...but it's now running so smooth and quietly I hate to screw with it again.

If I adjust the motor position to the limits of it's travel but with no extra force the belt is quite snug. Slacking it up a bit so that I can squeeze the two sides of the belt in about an inch each runs quite nicely but I don't want to jump the belt teeth either.

macona
02-22-2011, 09:12 PM
Really tight. They dont do any good if they are loose. We usually set ours so you get a "twang" when you pluck them. Of course this depends on the application too. If it is a tiny little motor or only bushings instead of bearings it might not be so tight.

PeteF
02-22-2011, 09:45 PM
Really tight. They dont do any good if they are loose. We usually set ours so you get a "twang" when you pluck them. Of course this depends on the application too. If it is a tiny little motor or only bushings instead of bearings it might not be so tight.

Wow, thanks for posting that, I had no idea they should be that tight! I had the motor off my surface grinder to convert it to delta windings and clearly haven't put the belt back on tight enough then (toothed belt).

Pete

Machtool
02-22-2011, 10:52 PM
(toothed belt).
Pete.
When you say toothed belt, do you mean a cogged vee belt or a timing belt? One has teeth on the pulleys, the other just plain vees in the pulleys.

Example of a cogged vee belt here.
http://www.bosch.com.au/content/language1/html/4662.htm

Regards Phil.

PeteF
02-22-2011, 11:15 PM
Phil, I mean a toothed belt, yes also called a timing belt, but that's just one application for them. I fiddled with the tension a few times and thought the tension I had it at provided the least amount of vibration, but I don't think I tried it as tight as you're suggesting. I would now like to experiment more with this, as you know any vibration is bad in this application so the smoother I can get it the better. It's a much smaller and wider version of one of these anyway.

Pete

http://img-europe.electrocomponents.com/largeimages/R474942-01.jpg

darryl
02-22-2011, 11:15 PM
Go out to your vehicle and check the tension on the serpentine belt. The tensioner should be putting an appropriate amount on the belt. I realize that's probably a v-ribbed belt, but for a similar width and thickness of carcass, the tension would probably be in the same ballpark for a toothed belt. I agree- it should not be loose at all. You should be able to hear the pitch rise as you pluck it while tightening, and you'd want to have it in that range at least. That's pretty vague I know, but it's something you can get used to. If it doesn't 'sing' a little at least, it's too loose.

Machtool
02-22-2011, 11:36 PM
I don't think I tried it as tight as you're suggesting.
Pete. It wasn’t me suggesting that, That was Macona.

This is on the drive from the motor to the rear of the grinding wheel spindle? Fairly small bearings in the front of the motor. I cant see the need to tension it until it twangs. Its not like its going to start jumping teeth. I wouldn’t be putting that kind of radial load on what should be a precision spindle.

On axis drives, going to ballscrews I might do that, it reduces tooth creep on the pulleys.

Regards Phil.

hardtail
02-23-2011, 01:28 AM
There are a lot of different things that come into it but span is an important one and I can't imagine it being anything over 16" here so I would not recommend any twang, as the span grows the tension can decrease abit but unless this was designed with this drive setup it can be hard on your bearings to run it really tight. From your description I would shoot for about twice the tension you currently have or max 1/2" deflection when squeezed together in the middle.

PeteF
02-23-2011, 03:55 AM
Pete. It wasn’t me suggesting that, That was Macona.

This is on the drive from the motor to the rear of the grinding wheel spindle? Fairly small bearings in the front of the motor. I cant see the need to tension it until it twangs. Its not like its going to start jumping teeth. I wouldn’t be putting that kind of radial load on what should be a precision spindle.

On axis drives, going to ballscrews I might do that, it reduces tooth creep on the pulleys.

Regards Phil.

Oh sorry Phil, I knew it started with an "M" :D

I'll take another look at the belt tension when I get a chance. I have to strip the machine slide and get all the grease out somebody kindly put in there, so will have a chance to experiment then. I basically figured that the teeth on the belt would prevent any slipping, so there was no real need for it to be super tight, unlike a V-belt which would slip. Having it a little looser would also help to isolate the motor from the spindle a little better too. When I was experimenting with the tension that certainly seemed to be what I was finding, but again, didn't try it super tight so not sure how that would go.

Pete

macona
02-23-2011, 04:51 AM
You would really have to have the belt super tight before you would come even close to the radial load on a spindle. You will probably break the belt first.

If you are running standard trapezoidal belts like MXL, XL, and L series belts know that they do run a little noisy. Thats one of the reasons for HTD belts and why they are used in spindle drives.

firbikrhd1
02-23-2011, 10:52 AM
I just finished installing a new timing belt of the "toothed" type on a Nissan Xterra yesterday. The belt has an automatic tensioner spring that tensions the belt before you lock the tensioner down. When finished I was curious as to how tight he belt was. I found that the tension was much like the same type of belt used these days as a final drive on Harley Davidson motorcycles. A good test, the same one I performed yesterday on the Xterra, is to attempt to twist the belt 90 degrees. You should feel high resistance when you get to the 90 degree point.
Technically, Harley uses a gauge to measure resistance to deflect the belt a specific distance to insure correct tension. I believe the numbers are 10 pounds and a deflection of 3/8", but am not certain.

QSIMDO
02-23-2011, 11:33 AM
You would really have to have the belt super tight before you would come even close to the radial load on a spindle. You will probably break the belt first.

If you are running standard trapezoidal belts like MXL, XL, and L series belts know that they do run a little noisy. Thats one of the reasons for HTD belts and why they are used in spindle drives.

So decently "snug" would be OK then?
It's the main drive belt on my lathe but the v-belts for the stock speed reduction slip long before that belt will suffer damage.

Thank you all.

Benta
02-23-2011, 11:37 AM
+1 for the 90 degree twist, that's the way I learned it and it works.

Cheers,

Benta.

Willy
02-23-2011, 11:51 AM
In order to give an accurate answer one should be aware of several factors that come into play when asking how tight any belt should be.

Some of these factors are, span between pulleys or sprockets, load to be transmitted, rpm range of drive and driven members, belt width. Is the belt dive to be used for registration as in a stepper motor drive?

Lastly, belt drives are somewhat forgiving in nature. As long as one is reasonably close and uses a bit of common sense one should not be overly concerned radial bearing load.

I've included a link to a pdf file that discusses some of the issues involved in timing belt tension.

http://www.sdp-si.com/D790/PDF/D790Tech.pdf

Remember this is not a definitive guide for everyone's particular situation, just a guide to illustrate some of the parameters involved.

Personally I've always used the "1/64" deflection per inch of span" rule combined with an educated feel with good results. I know, vague as the day is long.;)

PeteF
02-23-2011, 04:22 PM
A good test, the same one I performed yesterday on the Xterra, is to attempt to twist the belt 90 degrees. You should feel high resistance when you get to the 90 degree point.

Surely that will be highly dependant on the width and span of the belt? In my case it's a wide belt and the two pulleys are sitting right next to each other; there's no way I could twist the belt, even a small amount, unless the belt was stupidly loose. With something like a motorcycle final drive I can well imagine this would be a valid test, but basically unique to that particular application. Conversely I used to use these as precision drive belts in electronic equipment, and you'd be able to turn those little suckers several times yet their tension was quite high.

Pete

noah katz
02-23-2011, 05:36 PM
Go out to your vehicle and check the tension on the serpentine belt. The tensioner should be putting an appropriate amount on the belt. I realize that's probably a v-ribbed belt, but for a similar width and thickness of carcass, the tension would probably be in the same ballpark for a toothed belt.

I don't see why a belt with teeth would need to be nearly as tight as one relying on friction.

winchman
02-23-2011, 05:43 PM
The toothed belt on my garage door opener is tensioned by a spring on one end of the belt. It's not all that tight, as I can compress the spring with my hand, and the belt sags a little from it's own weight when it's not running.

macona
02-23-2011, 06:01 PM
I don't see why a belt with teeth would need to be nearly as tight as one relying on friction.

The reason for teeth is not necessarily for traction. It is also to keep the belt in time with the pulley. If the belt is loose then there is lost motion, especially under load. Any slack is lost motion,

gda
02-23-2011, 06:56 PM
All depends what you want the belt to do.

If you are just using it to drive something one direction, not critical I would not worry much about it. High accuracy servo position with high inertia loads and direction reversal without going slack? you need to do some calculations.

This is what I typically check my designs with:

http://www.brecoflex.com/download/B204.PDF

Gates also has some good online guides.

noah katz
02-23-2011, 08:14 PM
The reason for teeth is not necessarily for traction. It is also to keep the belt in time with the pulley. If the belt is loose then there is lost motion, especially under load. Any slack is lost motion,

We're not talking about slack, meaning no tension, but the amount of tension.

darryl
02-23-2011, 09:53 PM
The 'teeth' need to drop fully into each slot, and with the variations in loading, speed, etc, the 'action' can vary. You do not want the teeth to ride up the cogs at all, which lack of tension can allow.

The belts don't stretch much, but they do stretch to some degree. If it's highly loaded, that will mean that one section of the belt will be tight while the other side becomes a bit looser. You can end up with the belt doing a bit of a cyclic oscillation around the cogs, which will glaze and wear the teeth, and wear the cogs. A significant tension is required to keep this to a minimum. It should be tight enough that the loose side doesn't really become loose- it should not be able to sustain a vibration of more than say 1/2 inch in total between pulleys.

Now I'm wondering if there is a published spec relating deflection, side force, and distance between pulleys for various sizes and types of belts- would be good info to have.

hardtail
02-23-2011, 10:55 PM
Depends on the belt material.......some have no stretch at all due the outer layer material and construction........toothed belts are positive engagement as long as there is no jumping there is no lost motion BUT many factors come into play......tension and alignment are noteworthy, too slack can have premature wear issues.

Willy
02-23-2011, 11:13 PM
Now I'm wondering if there is a published spec relating deflection, side force, and distance between pulleys for various sizes and types of belts- would be good info to have.

Have a look at the 2 pdf links on the previous page.

firbikrhd1
02-24-2011, 12:49 AM
Surely that will be highly dependant on the width and span of the belt? In my case it's a wide belt and the two pulleys are sitting right next to each other; there's no way I could twist the belt, even a small amount, unless the belt was stupidly loose. With something like a motorcycle final drive I can well imagine this would be a valid test, but basically unique to that particular application. Conversely I used to use these as precision drive belts in electronic equipment, and you'd be able to turn those little suckers several times yet their tension was quite high.

Pete

Well Pete, there is no doubt that variables enter into the equation. My comments were to be considered as a guideline to proper tension. Common sense has to be applied to any given situation. My reference to "high resistance" is subjective as well. What seems to be high resistance to a 90 pound weakling might be nothing to a power lifter. All that said, in my recent experience the resistance I felt when I twisted the belt on the Xterra was about the same as it is on my Harley regardless of the difference in belt width, which is considerable (about 30% wider on the Harley), or distance between the belt sprockets, which is also considerably less on the Xterra than on the Harley. The reason I mentioned the Xterra as a reference was that there is a spring tensioner that applies the tension to the belt before the idler pulley is locked in place with the tension applied; after which the spring is merely ornamental as it no longer has any bearing on tension. The tension applied by the spring has been determined to be correct for that setup by the engineers at Nissan and belt life is supposed to be about 100k miles. Perhaps coincidentally it works out that the "90 degree twist" method of checking tension seems to be the same when applied to either situation mentioned above. Perhaps not, as well, for I have read in other manuals that the method described above is valid as a reference for tightness of this type of belt. It may not be as scientific nor as perfect as using a belt tension gauge but it is practical and it works.

For reference, the Harley belt is 1 1/2" wide and the Xterra belt about 1" wide. Distance between the sprockets on the Harley roughly 6" more than the sprockets on the Xterra.

One further thing I might mention is that in an attempt to be "scientific" in measuring belt tension on my Harley I have substituted a fish scale for a belt tension gauge with good result and verified the result with the aforementioned belt twist method. The Harley is a 1993 and has the original belt which continues to perform as well as it did the day I bought the bike in '93.

Willy
02-24-2011, 01:24 AM
Don't forget the drive belt/drive chain tension varies by a substantial margin on a motorcycle. As the bike's rear suspension goes through it's complete range of travel distance between drive and driven sprockets varies considerably.

Yet despite this variance of sprocket centerlines these belts do last a very long time. That's why I said in my previous post:



Lastly, belt drives are somewhat forgiving in nature. As long as one is reasonably close and uses a bit of common sense one should not be overly concerned about radial bearing load.



Let's not over analyze the subject...it's not our nature to do so.:D

Black_Moons
02-24-2011, 02:26 AM
Don't forget the drive belt/drive chain tension varies by a substantial margin on a motorcycle. As the bike's rear suspension goes through it's complete range of travel distance between drive and driven sprockets varies considerably.

Yet despite this variance of sprocket centerlines these belts do last a very long time. That's why I said in my previous post:

Let's not over analyze the subject...it's not our nature to do so.:D

I thought rear suspensions where allways designed to pivot at the same point as the drive sprocket centerline, So that the chain or belt length does not change.

hardtail
02-24-2011, 03:50 AM
It would be close but it will always be changing.......the manual actually I believe states that tension should be set with rider onboard......so the suspension load/tension changes right there.......now if you saw my buddy's OL you would know those belts are tough.......lets just say no buck five hardbelly...

We maintain probably over 350 fans at work with a variety of belt and drive differences....about 10 years ago we had a contract to convert about 15 of the bigger fans to VFD's with polychain belts, during the install we had a progress meeting that I had to sit in on and questioned the tensions, engineer and contractor all swore by the results claiming you wouldn't require as many belt changes. I remarked they were correct but I would still have to remove them and the sheaves to change out the bearings behind them......about 6 months later they claimed Gates had released new info and numbers and reduced the tension into a more realistic value.......

Willy
02-24-2011, 07:21 AM
I thought rear suspensions where always designed to pivot at the same point as the drive sprocket centerline, So that the chain or belt length does not change.

BM, have a look at the illustration below.
As you can see the variable here that changes the distance between drive and driven sprockets during suspension travel is the swing arm pivot.

In the the example below you can see that when all three elements in the equation are in alignment, the drive and driven sprockets are at their maximum distance.

But as soon as the rear sprocket moves up or down this relationship changes and the distance between drive and driven sprockets decrease.

M/C manufactures strive to keep the distance between drive sprocket and swing arm pivot as short as possible in order to lessen the degree of this effect. As you can imagine the further these two elements are away from one another the greater the distance variation between sprockets due to suspension movement.
Look at a motocross bike for example with 12-13 inches of rear suspension travel, you'll see that the drive sprocket's location is as close to the swing arm pivot as possible in order to negate this effect. Not so much of a problem on a bike with say, 3" of travel.

http://i76.photobucket.com/albums/j31/250willy/de797dda.jpg