I'm planning to try a tape driven x-axis on my small, but long, router cnc.

I've come across aerosol sprays to increase the friction of rubber belt drives,
but is there one for spraying onto steel tape, or the drive wheels, so that I increase the friction between the two ?

The set up will be to have a shaft coupled each end of the stepper motor driving two tapes directly each side of the gantry, and a pair of presser wheels on each side to maximise the area of tape in contact with the drive shaft. Probably another shaft across the other end with coupled wheels on that.

Another approach might be to change the material of the drive shaft/wheels to something with a high friction surface.
What would be suitable, any ideas ?

Many thanks
John

John, how are you going to garrentee that you don't lose position due to slippage? This style of motion would seem to be good for several applications, however I'm not convinced that it applies to well to CNC. Most people who are doing something similar are using drive belts which are "cogged" so as to not lose position.
If going to a long belt isn't apealing.. the next option is a rack & pinion setup.

My two cents..

Jerry

John, how are you going to garrentee that you don't lose position due to slippage?

This is the reason why I'm asking (note the use of the word "try" in my first line).
John

Hi,

Couple months ago I mentioned about the same idea here on forum, but at that time I was thinking about ruler tape for measuring and electromagnets. Later on I did some trials with metal reels in which I have inserted magnets from speakers and I preloaded the tape at the end and put it between two reels tangent to the drive reel. The last trial was with 40 mm diameter reel and stepper of 3 Nm. I could not believe the results. No slippage until I have obliged to stop as the MDF setup started to crack.

Try with magnets from speakers. I made a channel inside of reel where I inserted the magnet.
I definitely build a such a transmission in the future.

Zoltan

John;
You are probably, like me, old enough to remember reel to reel tape recorders. How about this for you steel tape drive: Have a reel at each end that the tape is firmly attached to with one reel full of a travel's worth of tape and the other with just a turn or two; i.e., you have just over two times the length of tape needed for the travel with the X carriage secured to the midpoint of the tape. When the axis drives the excess tape transfers from one reel to the other moving the carriage along but because everything is firmly attache you are not relying on friction. The distance moved by the tape per revolution will depend on the amount of tape on the reels but this is something that could be corrected by software with a lookup table. You would need a stepper at each end of the tape to keep it tight and the control of the steppers might require a bit of sophistication to make sure it stays tight. However, because there is no chance of tape slippage everything should be nicely predictable when it is all set up.

Hi Zoltan.
I remember your posting, and thanks for posting your results.
There's a posting of mine somewhere in the archives, where I, too, suggested using a measuring tape, but from a slightly different perspective.
We were discussing a cheap approach to encoders, and I tought that the tape could not only be used for the drive, but have a magnetic signal impressed on it as an encoder. I'm not sure if I, or someone else suggested one could read the printing on the tape as a simpler method if you only needed an accuracy of 1mm !
Now that I've got z- and y- axis finished, I'm looking at the problems of x-, and I agree with you that this should be one method worth trying.

Cheers, and I'll give it a go.
John

Geoff, that's interesting. I must get my head round the possibilities there.
Basically fixing the tape both ends, instead of a closed loop, with a tightener(stepper) each end....... mmmm ?

John

Zoltan - I can produce an improvement in "friction" between the tape and the drive wheel, but I only have small neodinium magnets. I'm not sure I understand where the "slots" are that you cut.
Can you post a sketch of your arrangement, or pm me if you prefer, to john at fanmaker dot co dot uk.

Many thanks
John

Check these guys out -- they have perforated belts and sprockets that fit in the holes. No slippage there. Their applications department is also very helpful, so they might be able to recommend a strategy for you if you give them a call:

www.belttechnologies.com

Ahren

Ahren - thanks, you've supplied the last piece of the jigsaw(see my signature !).
I've spotted an off the shelf source of perforated steel belt, but couldnt see how to produce the sprocket to drive it. These guys have shown me the method - ball bearings in holes around the face of the drive pulley.
Tape source ? The new digital retractable tapes have holes down the centre at 1cm intervals. Don't know the accuracy, but it will probably be fine for my use.

John

I recall the reel to reel tape recorders and also the grief on reel to reel memory tapes in the early computers. As you go to higher and higher speeds, you'll find that slipslide creep becomes harder and harder to avoid - especially at high speed direction changes in the data search on the computers.

Essentially, you're trying to improve the coefficient of friction. Until you go to mechanical interlocking (cogged belt, chain, rack and pinion) you'll only get better but you may never achieve the "nirvahnah" needed.

The challenge in ANY friction drive system is to get max traction without deteriouration in component life. CVT transmissions can be real oxymoronic in this nature as you MUST lubricate the device for it to live yet every lube used has to be used in a way so that it doesn't screw up/eliminate the friction so critical for it to transfer power from the drive to driven plate via the rolling wheel.

Mag clutches work well - 100% of the car A/C's I know of all have them.

The old Lionel trains had Magnetraction to help keep them on the track better and improve traction - dunno if it worked cause I could fly mine off at will.

The speaker magnet would help. So would pins and slots in the tape ala what they did in the old 8mm movie film. But that's essentially a chain and sprocket but it sure does keep the timing perfect.

Who'd have thunk that the nearly obsolete film product made so popular by Kodak would technology transfer to CNC so appropriately. Now, if we could only get the Stanley tape measure and the Kodak people interested in a joint venture....

Wanna bet neither would be interested???? Maybe Fuji Film and Kobe Steel. They've both done wonders with technologies that US competitors have abandoned over the years to the point that they rule the market in them, especially Kobe and the spring steel markets.

EDIT: in a final thought, check out what the bandsaw people do. They constantly have to deal with this exact same problem. Anybody who's walked a mile in their shoes would/should know the answer if anybody will. END EDIT

NC Cams - coincidence you should mention Kodak. There is a current forum of film technicians, I think, looking for steel film, that used to be made by Kodak. Use it for sprocket registration in film projectors , as far as I could make out.

But your mention of bandsaws prompts other ideas. Mine is driven by the 1/2" blade running on the outside of a v-belt . With a 6" wheel and only about 5" length of blade in contact with it, there's no way I can make it slip.
Interesting thought - thanks.
John

Check these guys out -- they have perforated belts and sprockets that fit in the holes. No slippage there.

Finally.. a solution that I can put some faith into. :D

I hear you all when you say "mine works.. it don't slip" and for the most part it probably doesn't, at least not completely, but I bet if you had a high speed camera and an indexed band running on an indexed wheel, you'd soon see that it is slipping a certain amount, its just the nature of the beast. This is all fine and dandy on a bandsaw or whatever, they are in continuous motion and if the speed isn't right you just crank it up a bit.. In the CNC world precision is what we're looking for. A "little" slippage is a little bit to much.. you have to know your machine is at its proper position or you'll never "know" your parts are coming out correctly, to their proper size/shape. A cogged tape will do this marvilously!

There is no harm in trying and we learn by trying. However, I could "try" to build a spindle w/ 0.0001" runout using a run of the mill "Ag" bearing and NC would tell me the same thing.. it will work.. but you'll never get close to 0.0001, and you'll never get the life you need etc etc.. My previous comments weren't to slam the idea, just to bring completely into the light, the area's where it could possibly fail or give you pains..

Jerry.. [I wish I could get this dude and the bongo drum's outta my head.. no wait.. maybe its all the cold medication's I've been taking..:D]

In the CNC world precision is what we're looking for. A "little" slippage is a little bit to much.. you have to know your machine is at its proper position or you'll never "know" your parts are coming out correctly, to their proper size/shape. A cogged tape will do this marvilously!

]

Oddly enough Jerry, we all seem to be looking for our own version of "precision", though it's not a word that I need to use for my use.

And the reason I started this thread was to ask for suggestions relating to an experiment I wish to conduct.

We agree in that sprocketed tape is a very possible solution for an accurate setup, and while I am following that up, it isn't a "one size fits all solution".

But I'm still looking for a way of using plain steel tape, and the bandsaw or the magnetic stiction ideas may fill that particular bill.

John

OK, so now, courtesy of ebay, I have 10metres of steel tape, punched every 1cm with a hole 2.5mm( or possibly 3/32", I can only check with drill bits as a gauge).

Can anyone link me to a supplier of a small quantity of steel balls, prefer UK if possible, unless you've got some in a bottle ? :D
I've tried googling variations on chrome steel balls etc but can't get past 'ballraces', if you see what I mean.
I intend to try the method that Ahren linked me to in his #9 posting.

John

I just Googled "precision steel balls" - only 4.4 MILLION hits.

You'd think that someplace there you should be able to get some "sampled"....

Sadly, the "precision steel balls UK" search netted only 984,000 hits.

How simple !
If only I'd thought to include the word 'precision'.
Ah well, must be time for a little something to get the brain cells going.
Thanks NC
John

Cycle Ballbearings

http://www.bicycle-mountain-bike-cycle.co.uk/index.php/product/ballbearings/bearingsandraceswithgrease/index.html?d2leisure=qhid84opppku0vt6rs49dtlsq1

Its a start.

RipperSoftware

Here is an UK Ebay list for 3/32" steel balls

http://cgi.ebay.co.uk/100off-2-381mm-3-32-AISI-420-Stainless-Steel-Balls_W0QQitemZ7624643870QQihZ017QQcategoryZ98641QQssPageNameZWD1VQQrdZ1QQcmdZViewItem

RipperSoftware

Thanks Rippersoft - spot on.

Glad to help. Please post pictures when you get this constructed. I am always interested in alternatives to the status quo (leadscrews and nuts).

RipperSoftware

Sure.
Without any decent skills or kit of my own, I may have to get a local precision engineer interested and pay to get the sprockets made to my spec.
However, I like nothing better than a challenge so I'm thinking about a diy approach to that as well.
I'll probably buy in both 2.5mm and 3/32in balls. If it's too small,there's the possibility of slip; if too big, the tape will ride too high, and not on the face of the wheel.
Though now I think about it, the oversize ball needs only to be sunk a few thou more into the surface for the positioning to be spot on.
I wonder if a few thou thick layer of rubber coating on the surface of the wheel would do the trick ?
This would be self adjusting to a small degree, and increase the grip as well, though not really necessary.

John

Maybe obvious, maybe not. You have to take into account that the tape goes over the sprocket over the "neutral line". The outside of the tape is stretched, the inside compressed. Because of this, the sprocket pitch is, although the difference is minimal, smaller than the tape pitch. The punishment for ignoring this, is error amplification by Pi.

Sorry, fka, but I'm not sure what your pointing out. While you are right about the stretch/compression of the two sides of the tape, in the context of my original posting ie this is for driving a long base router, this level of inaccuracy seems to be a little less than I am concerned with.
Unless, of course, I have misunderstood completely.
Could you put some figures alongside your thoughts ?
The tape is 0.15mm thick with 2.5mm holes at 10mm pitch, and will run round 25mm diameter sprocket wheels.
It will be driven by a 3:1 drive from a stepper motor with 400 steps per rev, which I calculate gives a single tep of 0.065mm.

Regards
John

I will try to explain. Your sprocket wheels are 25mm. 1 revolution gives a travelled distance of 25 * pi (pi simplified to 3.14) = 78.5 mm.

The sprocket + 2 * tape diameter = 25.3 mm.
The transporting diameter = (25.3 + 25) / 2 = 25.15 mm (neutral line). Transported tape at 1 revolution 25.15 * Pi = 78.971. This is 1.006% more than if you expect the tape to roll over the pulley.

I see now that you will have to rethink your sprocket, there is no multitude of the 10mm pitch. Let's assume it 8-ball. Neutral line transport = 80 mm. 80 / Pi = 25.48.
25.48 - 0.15 = 25.33 mm. Transport per step at 1200 step/rev: 0.066 mm

The same things you will see in timing belts, the neutral line outside of the sprocket. You will also see the need for the number of tooth in contact, to distribute power. This almost prohibits a flexible coating, because the more precise the pulley, the better the load is distributed over the tooth.

I hope you understand this explanation from a linguistic challenged person.

Quote
< Let's assume it 8-ball. Neutral line transport = 80 mm. 80 - 0.15 = 79.85.
<
Unquote

Is this correct??? You appear to have deducted the thickness of the tape from the length of 8 holes length of tape.


If we assume that I use 8 balls then the length of the neutral line is 80mm.
This would give a diameter of the neutral line of 80/pi = 25.46mm.

But if the tape is 0.15mm thick I must reduce the sprocket wheel diameter by 0.15mm, making it 25.31mm.

This is not a problem, nor is the new stepping distance that it would give.

Unless, of course, the balls are the wrong size anyway. :D

No, it is not correct, I deducted tape thickness from diameter * Pi ! Stupid, Stupid.
This was because I lined out the calculation and found your diameter did'nt match and had to go in reverse. I will also edit the original.

I think you use a more accurate value of pi, but now we agree on the diameter.
The size of the hole is 2.5mm so I can proceed.
Now I will see if I can come up with an assembly method which does not require the sprocket wheel to be turned to such an accuracy while I wait for the delivery of the balls !

John

Tonight I remembered I made such a thing more than 20 years ago to transport thermal paper. I used needles from a needle bearing because it's easy to make a press-fit for these. From my FAG catalogue the shortest 2.5 mm is 7.8 mm (NRO.B2,5x7,8G2).

"FAG catalogue" ? more info please fka.
John

FAG catalogue is nothing more than a standard bearing catalogue, in this case from FAG. Find out which brand your local supplier is associated with, get such a "bible" and order from the catalogue. Be amazed at what you can get. The needle bearing supplier of origin is INA.

Greybeard, try these people
http://www.chronos.ltd.uk/acatalog/index.html?http%3A//www.chronos.ltd.uk/acatalog/catalogbody.html&CatalogBody

http://www.hpceurope.com/index.html

Colin.

fka - many thanks. This aged person needs all the help he can get.

Thanks Colin - look very interesting and have bookmarked both.

As soon as the balls arrive, I'll start adding drawings/photos of progress.
John