Hall Effect Sensors From eBay: Supply Current

As a follow-up to those surprising (in an un-surprising way) magnetic field measurements, I measured each sensor’s current from a +5 V supply with no magnetic field applied:

Seq 49E 231NB AH49E
1 7.12 4.08
2 6.98 4.11
3 6.85 4.18
4 6.93 4.04
5 6.81 3.95
6 7.00 4.02
7 7.02 4.05
8 7.00 4.00
9 3.65
10 4.15
11 3.97
12 4.05

The sequence numbers do not match those in the field measurements, because the sensors spent the night in their respective bags and I didn’t want to re-measure everything from scratch. I may wind up doing it with a DC field, but not right now.

The first column averages 7.0 mA, the second 4.0 mA. It turns out that the Honeywell specs distinguish between SS49 and SS49E sensors, with the “E” suffix denoting the “economy” line:

  • SS49: 4 mA typical, no max given
  • SS49E: 6 mA typical, 10 mA maximum

The sensors in the first column have supply currents that are close enough for SS49E sensors, albeit with out-of-spec 1.8 mV/G sensitivity.

However, I’d say the sensors in the second column should be marked 49 rather than 49E and, if that’s true, then the plot thickens. In-spec SS49 sensors have 0.60 to 1.25 mV/G sensitivity, which neatly brackets the average 0.875 mV/G I measured: it’s faintly possible those have incorrect markings, rather than being manufacturing rejects.

But I wouldn’t bet on that

I should pick up some genuine sensors from a reputable supplier and measure those, just for completeness.

About these ads
  1. #1 by Red County Pete on 2013-01-25 - 13:02

    I wouldn’t bet against mismark. If parts are marked after final test, there’s lots of opportunities for foulups. In the couple of years I spent at National, we had several issues that arose from marking/mismarking and reworking parts at mark. Picture high-tech meeting Gutenberg.

    I liked it when we had parts that were either good or trash. We could mark before test. Kept (most of) the grubby processes in one facility, a long way away from the wafer fab. It gets complicated when you can get several product numbers from one lot of material.