Previous Note Previous Note

DI-145 Data Acquisition Starter Kit Developer's Diary

Next Note Next Note

3/17/2011 Note #21 A Glitch! (Caused by a lousy resistor? Really?)


Before we describe our most recent, trivial problem, our thoughts and prayers have been and continue to be with and for the Japanese people. Moreover, we salute the Heroic, Stoic, and Selfless Fifty as we ask a timeless but still unanswered question: From what hidden bastion of valor do such people hail?


Perhaps you've heard about the wave of counterfeit electronic parts that has assailed the industry over the past year or so (see here). It seems that supply disruptions caused by the Great Recession in tandem with central banks' Great Reflation have understandably confounded parts suppliers' best forecasting efforts to result in a general dearth of some parts. Some lead times are as long as 40 weeks, leading the surprised customer to wonder if they should even bother. Of course, crooks (located mostly in the Far East) have taken notice and offered to resolve the supply vacuum with "equivalent" products, sometimes even at premium prices. These alternatives look exactly like the part they portend to replace (the same manufacturer logo and part number), but it's just a ruse. They end up operating with degraded specs, if they work at all.

DATAQ Instruments has not been immune to this supply chain disruption and we've had to make adjustments accordingly. One problem that bit us during the DI-145's parts buying process was a general shortage of 1MΩ, 0.1% resistors. Each DI-145 needs 8 (2 per differential analog channel). Having learned from past mistakes we made certain to buy only from approved manufacturers and resellers, and after some effort we were able to locate enough from two for the first production run. We have a long history with both, and they are well regarded in their industry. Problem solved, right?

With our first wave of production units in our possession we immediately began calibrating individual units. Most passed, but an alarming number did not. We randomly tested around fifty units and had about six failures for a dropout rate of 12%...horrible by any standard. Digging into the problem we found that all the failures were caused by defective 1MΩ resistors that actually measured anywhere from 1.4MΩ to over 20MΩ. Replace the defective resistors and the DI-145 works like a champ. In nearly 27 years in business we have never, and we mean NEVER seen such consistent failures caused by a bad resistor. We smell a rat with counterfeit stripes, but we purchased these things directly from a top-tier vendor. Could the counterfeit scam have worked this far up the supply chain?

Fortunately, all the production boards used resistors from a single vendor and there were around 200 left on the reel after the run. We're working with the manufacturer (who shall remain nameless) to return both the depleted reel and some defective boards and resistors for analysis. In the mean time our calibration process has slowed compared to the planned production rate, but we're still making progress and do not expect this glitch to delay our delivery promise of the first quarter (that's by the end of this month).

We'll share the results of the rank resistor review when more information is available.

DI-145 Product Page: