Column: Embedded design
Figure 1: Chaos
distinctly sub-standard power-on transient output. 4. Working on a new prototype which has 30V and 5V rails, and some fairly large reservoir capacitors on the high voltage rail. I am re-positioning the board on the bench when my ‘scope probe ground clip springs off and hits a 30V connector, generating a terrifying bang and spot welds itself firmly in place. Embarrassing! So, the next time I
Figure 2: Order
resolve to discharge the 30V rail before moving the prototype, which I gingerly do with a short length of hook-up wire. However, what I thought was a ground as I jabbed my wire in was actually a via connected to the 5V rail, which was instantly overvolted, killing off a half dozen op-amps, a tantalum capacitor and a microcontroller. 5. I need to measure the (under -60dBm) local oscillator leakage on a transceiver capable of about 5W of TX output. So I connect it directly to my spectrum analyser because “the unit is safely in receive mode, and can’t possibly go into transmit”. Several measurements later a mis-
accidents – that I think I am qualified to make these comments. Here are a few examples of my mishaps:
1. Te de-mating force of an SMB connector is actually about the same as the shear force of the solder joints holding it onto the PCB. Over-torque an SMC connector and the little retainer nut comes straight off. You can guess how I found those “happy” facts out. 2. A relatively simple bias controller circuit functioned perfectly in use but always manifested faults during testing …
because I had placed the (bare) PCB on an anti-static mat, with sufficiently low bulk resistivity to upset the voltage dividers in the circuits. 3. A new radio PLL chip works perfectly at the start of its development, but weeks into the task the prototypes start failing. Cue days of exacting diagnostics, before the culprit (an over voltage transient at start-up killing the rather fragile chip) appeared … But, why? Because I had changed my bench power supply for a new model, which had a
placed probe tip (of course!) shorts the n_TXE control signal to ground, activates the transmitter for a few seconds, and teaches me just how expensive getting a spectrum analyser input circuit can be. Hopefully a few of my experiences will
avoid you having to go through similar ones. But, at the end of the day, while it is never possible to limit a risk to zero, a little prudence and care can significantly reduce it.
Tis column on embedded design prepared by Myk Dormer continues in the next edition of Electronics World
www.electronicsworld.co.uk July/August 2026 11
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44
