On Mon, 17 Dec 2018 14:08:18 -0800, mike <
ham...@netzero.net> wrote:
>I misunderstood. Thought you were discussing the handle and how it
>didn't hold the bits well.
No. It's the way the bit is made that's the problem. Torx drivers
and screw heads have a torque limit:
<
https://www.wihatools.com/torx-sizes>
Exceed the torque limit and bad things happen. However, this table
assumes that the bit and screw head are perfectly mated fit, with no
air gap anywhere on surfaces perpendicular to the direction of applied
force. If there's an air gap, the force (torque) will be applied to a
much smaller area than with a perfect fit. It doesn't take much of an
air gap, especially small screw heads which don't have much mating
area anyway. That's why the tapered sides of the new and
not-so-improved Husky bits are a problem. There is quite a reduction
in contact area near the too narrow tip, and quite an air gap due to
the tapered fit.
>If you're not using an impact driver, there ain't no such thing as
>excessive torque.
I beg to differ. Look at the max torque specs at the above URL and
test your ability to apply torque with some kind of torsion torque
tester. This one should cover T2 to T10 screws:
<
https://www.ebay.com/itm/392178731130>
I just tried it with a small torque wrench and found that I could do
about 8 ft-lbs or about 11 newton-meters. That's sufficient to hit
maximum for a T20 screw head far more than necessary to strip out the
tiny screws found in cell phones.
>If the bit rounds off or breaks before the screw
>turns, it's not an acceptable tool.
Maybe. If the bit was a good mating fit for the screw head, then yes,
it's probably junk. However, if it was a sloppy fit and/or had air
gaps, I could also blame the screw or bit depending on which one was
poorly made. There's also a question of metallurgy. If the bit were
properly heat treated, only the outside surface would be hardened and
annealed, while the core would be softer steel. That would give the
maximum hardness to protect the surface from gouging, and a somewhat
flexible core to give the bit strength. However, as the bit becomes
smaller and smaller, more of the material becomes hardened and
annealed steel, until the tip becomes very hard and very brittle. Do
anything wrong while driving this type of brittle bit, and it's going
to break.
>If it damages the screw in the process, you've got problems
>well in excess of a $2 screwdriver bit.
I have a collection of screw extractors and center drilling jigs. I
rarely need them.
>Since I don't have any of the tiny bits, my experience
>is with T7 to T10. I found the shiny bits
>round off very easily. The dull grey or black
>bits hold up better.
Most of the bits are electroplated with some kind of chromium
anti-rust compound. In the past, screwdriver tips were hardened
giving them a darkish color. This was the symbol of quality for
Proto, SK, and other quality screwdrivers. Now, every manufacturer
does this, except there's no hardening and the darkish color is either
an acid etch or paint.
>I have shattered bits with the impact driver, but
>that's not the issue when taking apart a phone.
Hint: If you feel the need to use an impact driver to remove a screw,
it's probably a left handed screw thread. (e.g. early Motorola
DynaTac and MicroTac).
Suggestions:
1. Don't use driver bits. There's nothing really wrong with them
except that the handles are too large for smaller size Torx screws.
2. I have a few of these "precision" Wiha Torx tools:
<
https://www.wihatools.com/torx-tools/precision-screwdrivers>
which I use for working on smartphones. They work very nicely.
3. Test the fit of whatever driver you purchase against the screw
head before applying main force. There are some really weird screw
heads out there.