This past week was a ballbuster. I didn’t find time to shave once, so the 26 mm 2CH3K4 I’ve mainly focused on testing since 11/15 hasn’t seen use since Saturday of Thanksgiving weekend (i.e., 12/02). The last two times I lathered with that knot (which is set in the Butterscotch Moe handle shown above), it didn’t give up a single hair. That – along with assurances from the supplier and time we spent rinsing, combing, palm-brushing, and vacuuming every single 2CH3K4 knot we’ve set in a handle so far – satisfied me that we could responsibly release 2CH3K4 brushes notwithstanding some initial concerns as described in the Knot Notes I posted on 11/24.
But yesterday I caught sight of the Moe, took it in hand, and lightly palm-brushed the knot as I thought about whether I might be able to squeeze in a shave. Then I noticed a couple loose hairs pop up. I continued brushing. More hairs emerged. That was it. I showed Pam (my wife) and then called Cody. This is what we decided to do.
1) We’re suspending sales of brushes with 2CH3K4 knots until we can determine and communicate with high confidence a causal explanation for the shedding we’ve observed.
2) We’re going to issue a $20 refund for every brush with a 2CH3K4 knot we’ve sold to date (that’s 29 brushes in total). We didn’t intend for our customers to serve as Guinea pigs to trial these knots.
3) Although we encourage those who have already purchased one of these brushes to keep and use it long enough to see whether it settles out satisfactorily, we will accept returns from original purchasers and grant full refunds inclusive of customary shipping costs upon request. In any such case where the brush was forwarded directly to a buyer outside the US, we’ll work to achieve a fair resolution.
For what it’s worth, I remain optimistic that these knots will settle out and fulfill our expectations in terms positive performance. But we certainly won’t order any more of them unless and until we sort out with the supplier what accounts for the loose hairs and come up with a reliable solution. I sincerely hope that happens. The hair used to make these knots uniformly (and that’s a key word) exhibits characteristics we want to offer as a select type with reliable consistency.
With all of that said, here’s what I think is going on with these knots. And I still think it with pretty high confidence, just not quite as much as I need in light of the Moe’s little back-slide yesterday.
As mentioned in the Knot Notes entry for the 2CH3K4 code, a sample knot I used several times before placing a substantial order behaved in much the same way as the test knot set in the Moe. I related that, along with photos, to the supplier and was given strong assurances in response. But I’ve had my share of experience with hollow assurances, and when the 2CH3K4 knots were delivered on November 15, I randomly selected a 26 mm semi-bulb and headed straight for the shave den. Déjà vu awaited. I shared this with the supplier once again and got back much the same response as before, this time accompanied by photos of techniques commonly used to remove loose hairs that I must understand inevitably occur with badger-hair knots. Of course, I already knew that. I also knew that I was encountering more loose hairs with these knots than I ever had before.
Absent some traumatic event, a single hair sheds from a knot for one of three reasons: 1) it was insecurely bound by glue and separated from the base to which it had been attached, 2) the length that shed broke off from a lower segment that remained bound to the base, or 3) it somehow came to be bound in the knot among other hairs secured by glue but did not itself come into any effective contact with glue. Hairs that become separated from glue without breaking typically exhibit some residue. Hairs that have broken off above a knot’s glue base should be expected to display a somewhat uneven appearance at their broken end. It seems reasonable to assume that a shed hair which appears to have a very cleanly cut and residue-free base-end most probably was bundled and tied (perhaps very tightly) in the knot but never came into any effective contact with glue when the knot was cemented.
Immediately after first use of the test knot, I pulled another 26 mm knot from the box and cut it in half with a bandsaw. (That’s something I usually do anyway when new orders arrive.) On this occasion the one I cut in half appeared to be well constructed with general penetration by hair into the glue base measuring in the range of 4-5 mm.
We have a large collection of hairs that were liberated from these knots by combing, palm-brushing, vacuum suction, lathering, and some other techniques. We’ll be sending a bunch of them back to the supplier for its examination. I’ve paid closest attention to hairs that shed from the Moe’s test knot during ordinary handling and use as shown in the photo below.
Here is a higher resolution inset from the photo above. Note the even ends and absence of apparent residue .
Almost all of the loose hairs I’ve collected from the Moe and measured were close to 46 mm in length compared to free loft at the center of about 48 mm.
Below are two microscopic photos (approx. x140) of the same hair’s base end. This hair shed from the Moe during use. The end appears to me to have been cut.
Next is a pair of photos, the first taken at the base end and the second at the tip of another hair collected from the Moe during use.
Finally, below are two photos showing a 2CH3K4 hair that I cut near both ends with surgical scissors. Although I’d guess the German-made, precision instrument I used might be a little superior in quality to whatever tools are used by knot-makers in China to cut the base-ends of badger hair, just the same, I’d say these look a lot like the ends of shed hairs in the previous photos.
None of this stands as proof of cause, but I believe it supports the hypothesis that the shedding we’ve observed with 2CH3K4 knots is attributable to unglued hairs dislodging rather than glue failure or breakage. In any case, we’ll take this up again with the supplier and almost certainly learn more about something.
This obviously isn’t a situation we’d choose to be in, but we’ll make the best of it. Hopefully, if nothing else, our customers and others who might be paying attention will see that we’re highly focused on problem solving and continuous improvement toward making the best brushes we possibly can.
We’ll share more about this as we have more to share.