pswallace":3bgok3ty said:
'Just tested mine (TREND) against the incra 1000se mitre gauge ,picked out A few random angles and it was spot on.
Sometimes mobile phones,micro waves . .so on can affect digital readouts. I have A small set of digital scales that go haywire if near the phone.
That's really interesting.
Thanks to everyone who checked theirs.
If I had to sum mine up in a word, it would be "inconsistent." Not being blessed with an Incra mitre gauge (wrong mitre slots, etc!), I printed out[1] a range of angles onto A3 paper and checked against that.
Observed errors
- It doesn't increment and decrement the readout sequentially. If you move the arms, it will stick and then jump as much as two degrees in one go.
- I can get a repeatable error at 60deg., which I'd guess to be of the order of two degrees.
- The errors can appear to be quasi-random, for example as bad as 15deg off at 60deg (evidently not counting properly), then bang-on, then off by a couple of degrees.
- Sometimes there is a consistent error through the quadrant, sometimes it starts incorrectly (at 10 and 15), then self-corrects to be right at, say, 30 onwards. I can't understand the reason for this, unless there is some form of parity checking or encoding of actual angle (rather than just counting pulses).
Mechanical issues
The two arms of the thing are not identically sized. The upper one is narrower than the bottom one: 34.7mm compared to 35.1mm (+/- 0.05mm). At the hinge, the difference is split between the two sides - asymmetrically! Thus neither side can be zeroed against a flat surface.
I haven't tested it with obtuse angles, partly because of printing difficulties, partly because I'm unlikely to use it that way.
I've tried four, different, nominally good batteries in it with similar results[2].
So I think I'm giving up on this unit. It's evidently a cheap copy of something like a GemRed one. I'm still undecided about replacing it, but I'll look at the build quality of Trend and Gem Red, and possibly take my piece of A3 to the shop when I go!
Thanks for everyone's inputs.
Regards,
E.
[1] This isn't as rough as it seems. The lines were created in Corel Draw, which is a vector drawing package with a nominal angular accuracy of 1/100 degree, and printed on a laser at up to 1200dpi - far higher resolution than I can see with the naked eye these days! For 'ease of use' I made the lines 2pt thickness (it will do much thinner, but they get very hard to see except in really bright light). The image above is a rendered and scaled version of the original. To print it, I sent it directly as vectors from Corel Draw - the laser rasterized it, thus minimizing any floating point/rounding errors.
[2] The brand new battery was better than the lowest voltage one (still within spec). This may be due to the RF interference PSW reports (above). I've yet to try screening the thing (a bit tricky, but probably possible) and repeating the experiment. It has a flat metal surface in proximity to the circuitry, which may alter its susceptibility (could go either way, depending on the frequency involved and the direction it's coming from.)
[edited after a second coffee, to improve legibility, etc.]
