Sunday, January 09, 2011

A New Twist on Lace Fliers

I came to spinning not as a spinner, but as a person interested in yarns that I could not get from commercial sources. As a good student, I asked, “What is the best way to spin fine, high twist singles?” The conventional wisdom that I received was to use a “Lace Flier”.

Lace fliers are Scotch tension flier and bobbin assemblies ostensible designed for the rapid production of lace singles. They tend to be “balanced”, and have low friction bearings to allow them to spin at very high speed. They are flier lead and have small whorls resulting in “high ratios” so that they can turn very fast with a limited treadle rate. Or, at least that is the conventional wisdom.

In fact, at lower speeds (400 – 800 rpm) this is how lace fliers work. As a beginning spinner starts wanting to spin faster, a lace flier helps them insert more twist into their fine singles, and for the beginning spinner, it seems like a lace flier is the ultimate answer to rapidly spinning fine singles.

At slightly higher speeds, things start to fall apart. At low speeds, it takes very little power to drive the flier and bobbin. Windage is minimal. Energy to accelerate the flier is minimal. Energy going into twisting the yarn is minimal. And, (if you oiled your wheel) friction is minimal. However, windage is proportional to the cube of the speed and acceleration is proportional to the square of the speed, with the power dissipated as friction and going into twist being directly proportional to the speed. Together these represent a power consumption function. All the power to the flier is supplied to the flier and bobbin by the drive belt.

Thus, if “windage” consumes 1 watt at 400 rpm, it will consume 8 watts at 800 rpm, 27 watts at 1,200 rpm, 64 watts at 1,600 rpm, and 125 watts at 2000 rpm. This power must be delivered by the drive band. A piece of kitchen string tied with a square knot can easily transfer 8 watts, even if it is sweeping a small whorl. To get that kitchen string to transfer 125 watts (i.e., 2,000 rpm) you are going to have to ply it up into a cable and douse it in a high-friction belt compound, otherwise that belt is going to slip or break. You will be treadling like like crazy, and you wheel will be making all the noises that say it is going fast, but drive belt will be slipping against that little lace flier whorl, so that the flier and bobbin are not going as fast as you think they are going.

Wheel wrights have put great effort into reducing the windage of their lace fliers, but that cubed function for windage is against them. And, in a Scotch tension system, the bobbin cannot go any faster than the flier. In fact, the bobbin speed (and hence twist) will be between 10 % and 30 % less than the flier speed.

The truth is that a high-speed double drive system will provide higher bobbin speed and hence insert more twist into the yarn, allowing faster spinning than a lace flier. There are several reasons for this. Double drive wheels have more than twice as much contact surface with the drive band to transfer more than twice as much power. In a double drive system the bobbin goes 10 % to 30 % faster than the flier. Thus, if you have a Scotch tension wheel and a double drive wheel with the fliers going at the same speed, the double drive wheel will be putting 20 % to 60% more twist into the yarn. I did not know this when I looked at the ratios of various fliers as I prepared to buy my first wheel.

Last summer, I spun 10 miles of 5,600 ypp singles with a lace flier on my Ashford Traddy, and I guessed much of the above. In December, I got a Neiko Digital Tachometer (NDT). That confirmed my worst fears. I was wasting effort. I was treadling and my drive belt was slipping like crazy. (Despite that fact that I have long experience with drive belts in the power range of 0.5 to 800 hp, and I know how to minimize belt slip.) The bottom line is that the small swept area of the lace flier whorl is not capable of handling the power required to drive the flier at much over 2,000 rpm.

The NDT tells me that if I treadle diligently, I can insert a lot more twist into my yarn using the high-speed double drive (ratio stated as 17:1) than I can using the lace flier that Ashford tells me has a ratio of 40:1. For example, the NDT tells me that the Ashford Turbo dive band starts slipping at ~2,000 rpm. It does not matter how much tension I put on it, the band does not drive the Ashford Lace Flier faster, no matter how fast I treadle. From my treadle pace, I think my flier is flying, but the NDT is there to keep me honest.

The NDT tells me that much of what is tossed around about speed of spinning is not accurate. And, it tells me that much of the conventional wisdom about spinning is wrong.

handmade high speed DD set

A faster DD bobbin/whorl set for my Traddy that I made by hand.

Moreover, the industry has a odd way of calculating the ratios for DD spinning wheels.  The actual drive wheel/ bobbin ratio affecting spin insertion for the high-speed Ashford DD bobbin/whorl set is closer to 22:1.  The drive wheel/ flier ratio is ~ the stated 17:1, but in practice the bobbin turns faster than the flier, and thus twist insertion is greater than would be expected by comparing the ratios of the DD kit with the ratios of the lace flier kit.

DD is the smart spinner's way to fast twist insertion.


deserttyrtle on ravelry said...

So, if you were figuring out dd ratios, it would be more accurate to count bobbin turns than flyer turns?

Anonymous said...

Interesting post. Love the technical aspect you bring to knitting, especially your gansey construction.
Have you thought about a Canadian Production Wheel for a high twist singles? A wonderful DD wheel. As the owner of 3 I'd suggest trying one.


Aaron said...

Yes! At low speeds it does not make much difference, but at high speed, it is bobbin rotations that count!

Aaron said...


The Canadian Production Wheels are one of the things that got me started thinking about this. They know how to get a lot of spinning done, and they used moderate ratio DD wheels. They knew their spinning.

I would like a CPW, but am not willing to pay extra for one. Mechanically, my Traddy is similar, i.e., DT, large heavy drive wheel, DD, & high ratio. And, my Traddy is modern, so I do not feel bad, doing radical modifications to it. For example, it has holes in it for mounting test equipment. I would feel bad drilling holes in an old CPW, just so I could mount a a tachometer on it.

At this point, I have a multitude of flier whorls and different sized drive groves on my spinning bobbins yielding many different spinning ratios for different grists/kinds of yarn. Today, on my wheel, I could spin you a hank of (worsted) 11,200 ypp single or ply you up a hank of 800 ypp 10 ply. That is enough range for me at this time. If I need to produce other grists, I can easily make other bobbins to give me other ratios. I like my Traddy.

In short, my wheel allows me to spin anything I need to spin without taking up too much space. As parts wear out, replacement is easy and inexpensive.

Julie H. Rose said...

I'm staring at the photo & my wheel, and wondering if it's possible to turn a Lendrum upright wheel into a DD wheel. It appears that one would have to turn the flyer around. Was that true for the Ashford?

Aaron said...

Julie, the Ashford has 2 mother of alls, one for DD and one for ST. I happened to have both, or I would not have gotten here.

I have never worked with a Ledrum and have not idea. I would go ask the people on Ravelry.


RMD said...

Hi Aaron -
I'm getting ready to kit out my 2ndhand Traddy for some thin spinning and am poring over AA's book and your blog. If I understand correctly, the ST setup is limited to about 2000 RPM - so a 40:1 lacewt flyer can effectively only be run at about 1 treadle/sec. This is hardly an improvement over the standard flyer at 2 treadles/sec (which is about the max I can sustain for an hour or so). So this is not going to get the yarn I want any faster - just with less treadling. Bah.

Looks like a better route is a high-speed DD kit (I think I can modify my ST MOA with a few extra holes, no?). What kind of drive band do you recommend? Looks like yours isn't kitchen twine :)


Aaron said...

Ashford flyers as they come from the factory require some slip between the flyer assembly and the drive band. Given the need for that slip, and the aerodynamic drag that increases as the cube of rpm, you are not going to get an Ashford flyer to go faster than about 1,000 or perhaps 1,400 rpm.

I run my flier at 2,000 to 2,600 rpm by 1) using a specific differential rotation speed between my flier whorl and my bobbin whorl. And 2), using a small aerodynamic flier.

DRS is a major pain the ass to the point that Alden Amos and Henry Clems both say it is not practical. I say that it works for spinning very fine threads when both the whorls have a "U" shaped profile, both whorls are very precisely turned, and the bobbin is wound off often enough that the effective diameter of the bobbin remains constant. I am willing to put up with those limits to get real speed for high twist singles.

The high speed DD kit can deliver as much actual spinning rpm as the lace flyer.

The stock Ashford flyers need a lot of power. For them, I made up Alden's bumpless drive bands and I used his drive band dressing. This is a lot of work, - caused by the high aerodynamic drag of stock Ashford flyers.

Now I use very small, high speed custom fliers that Alden made for me with DRS whorls that I make. These rigs require much less power than any of the Ashford flyers. It also runs smoother with much less vibration. (It turns out that vibration can affect power transfer by drive bands.) for these small fliers with DRS, I use #3 crochet thread, held together with a short splice and drive band dressing. I also use a spring system to tension the drive band and damp vibration. The drive bands last for ~80 hours of spinning and take 5 minutes to splice together.

RMD said...

Thanks. Lots of information to digest. I'll be getting a high-speed DD kit to learn the rudiments of double-drive operation.
Oh, and a tool idea I've found useful: I mounted a small bike computer (cost $15) on my swift (4-arm type, not umbrella) and have adjusted it to read rotations. Very useful as a yarn-mileage counter!