Belt tension Allen key?

[greenmonster]

Brad`s take on 1kDS:

[greenmonster]

So, at last, l caved in and did it the sonic way.
Gates gave me 100 Hz (occasionally...) and
Allen & feeler gave 4mm VE & 1mm HZ.
Bit better throttle respons than old belts
setting (they were a bit off, 4mm on HZ f e).
Old dog finally overcame the 900 oldie way.

[koko64]

 

[DuciD03]

Brad`s take on 1kDS:

Nice to see an expert at work; thanks for posting that;

Brad makes it look soooo easy; see how he's having troubles picking up the sound on the tuner app .... an he doesn't like the 5 mil hex method ... interesting.

Loose side of snug when hot would be interesting to hear what brad would say to that ...

[koko64]

I think checking with a hot motor is insurance to ensure the belts aren't dangerously tight due to human error. Human error can occur with using or interpreting the reading of any tool.
It reminds me of setting drive chain tension on a side stand then checking for sufficient play with a rider, just to be sure.

[MonsterHPD]

I´ve tried some frequency apps but never managed to get consistant results, only managed to get sore fingers from all the plucking, really. After watching Brads video (thanks, Gerth :-)) I tried the Gates app, which seems to be slightly better than the others I´ve tried, the results still vary rather wildly. It may be just me, Brad seems to get consistent results, but that does not help me.
 
Also, trying to adjust whatever tension you have is very difficult to do consistently, it´s often too little or too much, and I also don´t like the Allen key method, it´s also very difficult to do consistantly. Since I like mechanical solutions (as opposed to electronics ....), I figured a repeatable mechanical solution would be nice. So I´ve machined a collar with 4 mm wall thickness to fit over the tensioner roller, sort of a "consistent 4mm Allen key":     

2021-05-04 20.12.04 by torbjörn bergström, on Flickr

With this over the roller, it´s easy to pinch the rollers towards oneanother, and since the belt is very stiff longitudinally, I don´t think it´s too sensitive how hard the pinch is.

Anyway, with the belt tightened against the collar and the tensioner torqued, the app gave me reasonably consistent readings (with collar removed, of course), but still varying between some 80Hz to 110 Hz.
The belt is slightly lose to my taste, so I´ll make another one with maybe 2.5 mm wall thickness, and see where that takes me.

Of course, I don´t think this is the holy grail of belt tensioning, but I do think it will at least make it easier for me to get the belt tension consistent. 

 

[DuciD03]

1- I´ve tried some frequency apps but never managed to get consistant results, only managed to get sore fingers from all the plucking, really.....

2 - ..... very difficult to do consistently, it´s often too little or too much, and I also don´t like the Allen key method, it´s also very difficult to do consistently. .... a repeatable mechanical solution would be nice. So I´ve machined a collar with 4 mm wall thickness to fit over the tensioner roller, sort of a "consistent 4mm Allen key":    

.... but still varying between some 80Hz to 110 Hz.
The belt is slightly lose to my taste, so I´ll make another one with maybe 2.5 mm wall thickness, and see where that takes me.
 

1 - ex-zactly! ditto its inconsistent ... and ya all get sore fingers too.

2- Collar spacer is a great idea; just remember different engines & cams on the same engine have different hex spacers to get the right tension;
so I guess you'd need 2 for a 1k ds; Vert 3mm & Hor 1mm? to dial that into spec.

... thinking it'd be better without the inner lip; not too tight fit over the roller; something that's easy to slip on and off when the rollers snugged in place.

[MonsterHPD]

I have not delved into the different specs, but the HYM1100 is 100Hz on both cylinders (but 140 on new belts ...).  I have no idea how well this might work in the end, but I will do some different collars and see where that takes me.

The lip is there to help pull out the collar from under the belt, but it was easier than I had expected. Might come in handy for the vertical cylinder since that roller is pretty well hiddden behind a fram tube. Easy to remove, should that prove the better option :-)
   

[koko64]

Really cool tool. You could sell thousands.$$$$$

[MonsterHPD]

Really cool tool. You could sell thousands.$$$$$

 

Free for anyone so inclined to make their own and experiment with frequencies and sore fingers 

Not much of a businessman, I´m afraid ....   

[MonsterHPD]

Well,
 even a potentially ggod idea can have it´s flaws .....

2021-05-05 16.23.13 by torbjörn bergström, on Flickr

The sleeve will need some more maching time before it will fit the vertical cylinder, the belt cover is very tight here. But I think even with enough machined away on the sleeve it will still be sufficiently well guided on the fixed tensioner roller to make sense.   

[greenmonster]

1 - ex-zactly! ditto its inconsistent ... and ya all get sore fingers too.   

+10.

On the 900, l put a 5mm Allen 45 degrees across pulley,
then tighten. Pretty consistent.

[DuciD03]

Well,
 even a potentially ggod idea can have it´s flaws .....

The sleeve will need some more maching time before it will fit the vertical cylinder, the belt cover is very tight here. But I think even with enough machined away on the sleeve it will still be sufficiently well guided on the fixed tensioner roller to make sense.   

youve com up with a neat simple solution; flaws can always revised; yep looks like very tight spacing of roller to that flange; maybe; no lips and just cut er in 1/2!

[MonsterHPD]

For now, I cheated and used a ca 2mm stack of feeler gauges.
I´ll try to machine the sleeve so there´s a little more than half left, soit will stay on the roller without me holding it. Let´s see if that works :-)

[Armor]

I would have a Ducati dealer tighten the belts withe their harmonic tool and them make your tool to match.  You may surprised how tight the belts are.  My opinion is that the motor runs better with tight belts.  You can feel the difference.