Cakes and Twist

Some folks on the Internet are saying that "center pull' cakes of yarn should be unwound from the outside (with the cake rotating) to preserve twist.  This finding is the result of poor lab technique.

My Royal Jumbo winder inserts (or removes) between 1 twist in 8 inches and 1 twist in 24 inches depending on whether it is winding the inside or the out side of a cake.  When I pull the yarn out of the center, the process is reversed and the original twist is restored.

This can be easily checked by taking a low twist yarn (e.g., worsted 2-ply) winding a cake, and then winding cake after cake of the same yarn  in the same direction - each time pulling from the center of the previous cake. After a few dozen rewindings, measure the twist.

Then, repeat the process, only this time rewind from the outside of the cake as a side delivery package where the yarn cake rotates to unwind the yarn.  If you wind the same direction each time, the twist will be noticeably different.  However, the change in twist will vary along the length of the yarn so it will not be possible to restore the twist by just running it though a spinning wheel.  Rather the spinner will have to pay real attention to the twist and insert different amounts of along the length of the yarn.

edited to change:  It has come to my attention that I have a Royal Jumbo Cake winder, which is one of the very few winders designed to produce true center pull cakes.  Most home winders, wind side delivery and unwinding by center pull will change the twist.

However, it is perfectly possible to wind side delivery and knit from that package end delivery.  yes, it does change the twist a bit, but not enough to bother any but the very best craftsmen - and there are not many master knitters around any more!

As demonstration, I am knitting from a side wound pirn in a shuttle - that is an end delivery yarn package. It works just fine!  Yes winding and unwinding the pirn changes twist as much as 1 tpi, but that is not really a problem for a competent knitter.

Twist and grist

I said, that for a hand spinner, inserted twist defines grist. A test is reported below showing that hand spinners can control grist to plus or minus 5%.  Can you control grist better?

This summer I have been spinning "10s" - worsted 5,600 ypp singles.

I wound a dozen hanks of 560 yards and weighed them.  Each should weigh ~45.4 grams.

I used a particular set of flyer bobbin whorls, spun 7 or 8 grams of single, then wound it off onto a storage bobbin. Then, I wound the storage bobbin on to a skeiner, blocked the hank with steam, and weighed it.

Mostly the hanks weighed right in the range of 43 to 48 grams.  Thus, by using DRS to control twist, it is possible for hand spinners to control grist to within plus or minus 5% over samples of 560 yards and maintain a specific grist over projects of many hanks.

Smaller samples tell me that this can be done with 20s or 40s.  I did samples of finer singles in the middle of the above test.  It is possible to spin 10s, then spin much finer 20s and 40s, and then go to back to spinning 10s at the same grist as before.

The 8 lb of  5s loom warp that I did last year with the smaller and less precise whorls, runs about plus or minus 12%.  (At the time, I thought the yarn was reasonably consistent for hand spun, and was very happy to have hanks anywhere in the range of 80 to 100 grams.)

By planning twist, one can control grist.  This is deep inside the world of  "the intentional spinner." The only path here, that I know, is DRS.  Without DRS I do not think you are going to be able to control grist to within 5%.  Without DRS, it is going to take you much, much longer to spin 7,000 yard of 5,600 ypp,  2,240 yd of 11,400 ypp and 1,120 yd of 22,400 ypp singles. Really, how many tries would it take you to spin a 560 yd hank of single that weighs between 10.5 and 12.5 grams? With DRS, I can be reasonably confident of spinning such a 40s on the first try.

It was not always like that.  Better tools brought better skills.  Better skills let me design and make better tools, and so forth.

Today, I find the omission of DRS by Judith Mackenzie to be grievous.  When my variation on grist  with DRS was 12%, and SG told me she could maintain grist within 10%,  I could give a pass to intuitive spinners who did not discuss DRS.  However, now that I know more of the control that DRS can provide, then failure to discuss DRS is a serious error and omission.  Alden Amos at least brings the topic up and discusses the math, even if he does not get into the details of hand movements and mechanics.

Twist and Grist

I have been accused of getting my twist numbers wrong.

That is unlikely.  I may have made a typo here or there, but over all, I am the one hand spinner that always knows his twist.   I use differential rotation speed (DRS) without slip to control twist insertion.  Once I install the whorls on the wheel, I know how much twist will be inserted.  If I want a thicker yarn, I do a more woolen preparation.  If I want a thinner yarn, I do a more worsted preparation, and thus I can get a range of yarns from particular whorl combination that inserts a specific twist.  However, I can always look at the whorls on my wheel and know how much twist is being inserted.

Thus, I have to know the twist for every yarn that I plan to make.  Once the whorls are (made and) installed -- the twist is set.  How fast I treadle does not affect inserted twist.  How fast I draft does not affect inserted twist.  Yarn build up on the bobbin does affect the effective diameter of the bobbins and hence rate of take up and hence twist.  However, this is a known factor that is predictable.

I know that when the whorls that produce 9 tpi are on the wheel, and I am spinning worsted I get 10s (5,600 ypp).  I have measured, and validated this hundreds of times.  I know that when the whorls that produce 17 tpi are on the wheel, I get 40s.  I have measured and validated this many times. I also have whorl combinations for 5s, 20s, 30s, 50s, 60s, 70s, and 80s that have been have measured and validated. Each allows me to produce a specific grist of worsted and a different, specific grist of woolen and intermediate grists of semi-woolen and semi-worsted.

If I want to sit down and just play with a fiber, then I am likely to set the wheel up as single-drive, bobbin lead (Irish Tension).  Most of my plying of hanks of  5-ply  gansey yarn or 10-ply Aran weight are done using the Ashford ST Jumbo flyer. Thus, every week, I do work with ST and IT fliers. With IT or ST my twist is as fallible as anyone's. However, most of my spinning is done with DRS because it is more productive and my twist more precise.    Spending a lot of time working with precise twist makes me a better spinner when I am using ST or IT.  Sock weight yarns with large numbers of plys are plied using DRS control of  the AA #1 flier. This holds twist to within a few percent. That limits my sock yarns to skeins of about 10 grams or about 80 yards.  Small skeins are my penalty for demanding precise twist in my 6-ply sock yarns.

On the other hand, my WPI measurments tend not to be consistent with several other authors.  I worked with yarns of known grist and practiced doing WPI until my measured WPI for 5,600 ypp (10s) was a consistent 75,  my measured WPI for 11,200 ypp  was just over a hundred, my WPI for shirting or hosiery yarns was 150, and etc. This seems different from other authors, but most do not write about spinning this fine. In every case, my WPI tests were done with worsted spun rather than woolen yarns., and this may explain some of the difference. However,  I always pack to refusal, because it provides more consistent results. At this time, I do not trust my WPI for woolen spun yarns.  It is likely off by 10 % when it should be accurate to 6%.  Thus, to get the grist of woolen yarns I weigh a measured length.

I think some authors do not pack to refusal, so their WPI and hence estimated grist and required twist are different from my numbers.  My numbers work - consistently.