Belt tracking proble

[BOBINETTE]

Hi everyone,
I am building a CNC router and for the X axis I am using a belt drive(Gates HTD 5m15 2110) this is the second belt that I am trying, there is also a homemade binder with a spring,with two Nook acme rods 5/8-10 pitch and I have problem with the tracking of the belt. All my three pulleys are in line, the motor if perfectly perpendicular to the belt when I rotate the motor counterclocwise the tracking is ok but on the clockwise the belt keeps going inward or if you prefer moving away from the motor. Here is couple of pics of the set up. Any solution is welcome
Thank you for your help
P.S. Is there any harm can be done to the motor by rotating it manually or by a power drill turning slowly

[Mr.Chips]

I don't have any experience with belts but it seems that in one direction you are pushing against the sprung area and the other direction you are pulling against it. Think that has something to do with what is happening.

If the below crude drawing, the drive motor is mounted below and centered between the idler wheels, tension could be adj by moving the motor up or down. Or mount the motor on a pivot point and spring load it down keeping it close to the center if the two upper idler wheels.

Then again someone that is wise to belts and pulleys will respond pretty soon. Good Luck

[andy_ck87028]

Perhaps a bit more info at this stage would help. At which point does the problem occur? Is it only the idler that doesn't have a flange and so is it the idler wheel here the problem is?

I feel uncomfortable about the spring as it is putting a variable into the arrangement. I would achieve tensioning by positioning and clamping. Even though you say the pulleys are all in line, it doesn't take much for mis-alignment to reveal itself. If the pulleys / idler all have flanges, then mis-alignment will result in belt wear which might well be acceptable.

Assumedly the acme rods can be presumed to be parallel and therefore in line. Possibly the problem reveals itself because the free span distance from the idler to the acme rod / motor to the acme rod are not equi-distant.

As Mr Chips suggests, I would consider having a balanced arrangement with 2 idlers and the stepper close to the idlers. That means then that the motor doesn't have to be spot on and only the idlers will then need adjusting.

Andy

[CarveOne]

The belt drive on the Solsylva 25 x 37 machine looks just like this arrangement except that there is no spring tensioner mechanism. There is just a hand made ball bearing plastic roller in place of the spring assembly. The stepper mount has radial slots in two of the three mounting holes for tensioning the belt. The bolt that the roller is mounted on is warped with a wrench or by tapping it upward or downward as required for a properly running belt. A belt like this is really sensitive. Just because the measured alignment is perfect doesn't mean that the belt will naturally run centered in both directions. The belt itself is not perfect. The bottom line is to just misalign the mechanism to get the belt to run the best it can in both directions.

CarveOne

[BOBINETTE]

Thank you for the help,
M. andy_ck87028 it is only the idler(tensionner) that don't have any flange, the problem begin on the stepper gear and moves to the idler. Has for the belt ware, we can put this aside because this is the second one and with the same proble and both of them are new.

CarveOne, when you mention (There is just a hand made ball bearing plastic roller in place of the spring assembly) is this a idler bearing. the idea of bending the stepper bolts doens't make me to hot about it.

Mr.Chips the idea of placing two idler and the motor below it seems acceptable.

Using the MR.Chips idea, I was thinking on placing the stepper on a plate with the sides chamfer and hadding one plate on each side with the opposites chamfer to hold ithe plate +stepper in place. The question I am having about this kind of set up is that the stepper would't have any spring once the belt would be in tension, can this create an abuse on the stepper. What are your thought about this, will it resolve this bad tracking situation

[nwrepair1]

Mr Chips is right. You can't use a spring tensioner on a belt that travels in both directions. An idler on a plate with a slot in it to allow adjustment can be used but only if it is placed in the middle of the length of travel. Also, get rid of the HTD belt. It is designed for power transmission not linear positioning. Use an AT5 or AT10 belt.These metric cog belts have better position accuracies and less backlash. How big is this router. Are you planning to use 2 steppers and 2 belts for this axis? If not, how are you going to control racking (the stepper moter side will move foward faster than the non motor side)? Also, wrapping the belt around the adjustment idler is a bad idea. The idler should be just pushing the belt inward (smooth idler) or pushing the belt outward (cogged pulley). Again, at the middle of travel...

[Mr.Chips]

Mr Chips is right. You can't use a spring tensioner on a belt that travels in both directions. An idler on a plate with a slot in it to allow adjustment can be used but only if it is placed in the middle of the length of travel. ...

With the two smooth idlers and stepper motor centered, how about moving the stepper motor vertically up and down for adjustment?

In the below photo I have shown two methods of tensioning. The one on the left keeps the stepper exactly centered while the one on the right because of the arc will be a little off center. The one on the right would also need a shoulder bolt on the left side of the motor shaft point to keep that pivot plate flat. What do you think?

[CarveOne]

BOBINETTE,

The idler is fixed. The stepper mounting bolts are not being bent to provide alignment adjustment. Since the x axis stepper motor doesn't attach to a lead screw I can use the stepper motor mounting plate nuts to make minor alignment changes.

The idler is just a gray plastic electrical tubing coupler that has a 7/8" skate bearing pressed into each end. I trued up the surface of my idler on my lathe but a lot of the builders probably don't even do that. The idler is mounted on a carriage bolt and has no adjustment other than to tweak the bolt by slightly bending it. If the idler bolt hole is drilled at 90 degrees on a drill press the bolt won't need much bending at all, if any. Rotating the stepper motor within the radiused slots of the mounting plate is used to tension the belt. Maybe this photo will explain things a little better.

CarveOne

[BOBINETTE]

Hi everyone,
I will begin by answering M.nwrepair1 questions.
The router Porter Cable 2 1/4 hp variable speed, I will try to make it with one stepper for both screws like the Solsylva 25 x 37 one.

For the racking, I can't see why it would rack, one belt, one stepper and both scews has the same pitch and NOOK claims that their screws are .0003 in/in repeatable, maybe I'm missing something here. .

(the stepper moter side will move foward faster than the non motor side)? I don't quite understand.

Also, wrapping the belt around the adjustment idler is a bad idea can you explain a bit more

Mr.Chips idea seems to be the easyest one, of what I can see the most critical part is placing the stepper exactly in the middle of the two screws and also placing the both idlers (TENSIONNER) at the same distance from the srews, is that wright.

M.Carveone, thank you for the pics, I will give an eye on yours also, as I don't have many precisions tools it might be hard machinig the arc on the plate for the stepper

Thank you all for your help

[nwrepair1]

Ok. The stepper motor moves clockwise (looking at the motor end) the belt will be tight on the right side. The belt will be looser on the left side. the effect will be that the right side will move faster than the left side. I have a commercial router made in Canada that uses belts as a rack and pinion. I am currently redesigning the system with the help of Brecoflex. There is a design section that will help you. The belt will only stretch a certain amount according to the load put on it . Once this happens, you make one final adjustment. The idea is to have the same tension on both side of the stepper. The deflection should be equal. That's how a commercial router does it. To do that, the easiest way is to use the trolley system. Two idlers above the stepper pulley that's pushing the belt straight up on the back side of belt. The site is Brecoflex.com. Sorry, I couldn't hot link it. Can someone tell me where I can find the info to do that

[BOBINETTE]

Ok. The stepper motor moves clockwise (looking at the motor end) the belt will be tight on the right side. The belt will be looser on the left side. the effect will be that the right side will move faster than the left side. I have a commercial router made in Canada that uses belts as a rack and pinion. I am currently redesigning the system with the help of Brecoflex. There is a design section that will help you. The belt will only stretch a certain amount according to the load put on it . Once this happens, you make one final adjustment. The idea is to have the same tension on both side of the stepper. The deflection should be equal. That's how a commercial router does it. To do that, the easiest way is to use the trolley system. Two idlers above the stepper pulley that's pushing the belt straight up on the back side of belt. The site is Brecoflex.com. Sorry, I couldn't hot link it. Can someone tell me where I can find the info to do that

Thank you for the info and I will give it a look, For the hot link you only have to click on the top of this quare where it says Insert link icon( a globe with eye glass)http://brecoflex.com/, a window will show up, you enter the URL of the site

[CarveOne]

The arcs in the stepper plate are not hard to make with a hand drill and a round file. If your arc holes aren't so perfectly shaped just use a flat washer to bridge the hole and it will work fine. I could have done it on my mill with a rotary table but didn't bother to set it all up.

I think that Mr.Chips method should work better than the Solsylva method but I haven't seen any posts by Solsylva builders discussing problems with excessive belt wandering with forward and reverse rotation. The plastic coupler idler described in the plans will eventually split due to the stress of the pressed in bearings. That can be avoided by warming the coupler before pressing in the bearings. A better idler can be made from two or three 1" (or 1-1/8") OD bearings stacked together (and using no plastic coupler).

I can feel about one stepper detent worth of backlash between the belt being tight in one direction versus the belt being tight in the other direction while seeing no movement of the lead screw pulleys. My belt has not been tightened enough to eliminate this yet. Doing that puts a lot of side load on the acme shafts and bearings. For reference, I have all-aluminum pulleys with two flanges at the stepper shaft and at both lead screws, so my belt will not wander very much anyway.

CarveOne