In the present, we have sewing machine reference cards that sew themselves
My Singer 20u Professional is a very capable sewing machine. The operating latitude is much wider. I trivially punch through eight layers of cordura and denim and make tidy seams, and I’ve also sewn 1/8″ hems on gossamer-thin silk.
The trade-off? There is no single tension setting that will work under all those circumstances, especially when I am changing thread types from cotton embroidery thread to heavy duty quilting-thread. Sewing under ‘home conditions’ normally I can just adjust top-thread tension and get the stitch I want. However, after lots of flailing, I realize I do have to adjust bobbin thread tension, at least sometimes.
In wandering around the 2-dimensional(*) adjustment matrix, I was confusing myself, mis-interpreting problems, and really wasting a lot of time, thread, and material. So today I decided to make myself a reference card set of the front/back of stitches as I moved around the top and bottom tension axes.
I used my Gutterman cotton/cotton quilting thread (red color on top, black on bottom) and two layers of canvas duck. I found a decent stitch setting, ran a row, and then methodically tightened the tension dial, running new stitches. Once I had gone as tight as made sense, I tried to go back to zero. Who knows how close I got, so I called my new location 0′ and then loosened the tension until it got silly.


I was suprised how much of a tolerable range of tension the machine had. The scans aren’t ideal, but from the fabric itself, I can see a lot of detail, and with the markings, it’s in context, so I can interpret it much better.
I did the same procedure for the bobbin thread tension. The adjustments here are much finer and more precise and give less latitude. It’s also easier to see exactly where I was set, so I had no need for the Zero Prime marker.


In looking at stitches, I wish I had something better than my naked eye and a really optically-lame magnifying glass on a cheap solder station. Having a stereoscope would be tremendous, especially when I am not using bobbin and top threads of different color.
(*) I’m also not sure that this tensioning is a 2d matrix. I suspect there is interaction with the fabrics themselves, and also vaguely suspicious that remaining thread on the bobbin makes a difference. I know I had a bad wind once (my own fault) and it seemed to give me an unusable bobbin. I ended up hacking the thread away.