Continuity Test

Continuity test is a basic but essential diagnostic that checks whether an unbroken electrical path exists between two points. In continuity mode a multimeter passes a small current through the probes and measures resistance; if the path is connected (very low resistance) the meter beeps, and if it is open it stays silent. It is the fastest way to answer the repair technician's most common question: is this connected to that, or not? Performed with the board fully powered off and discharged, it is a static test, safe to run at the very start of a diagnosis before anything riskier.

What it reveals

Continuity testing finds broken traces, lifted or cracked solder joints, blown fuses, and open connectors — and the opposite problem of unintended paths: a short circuit or a solder bridge between nets that should be isolated. By probing from a known point to a component pad you can trace a signal through a board and pinpoint exactly where the path breaks. A related static measurement, resistance-to-ground, extends the same idea: measuring across the main power input of a healthy hashboard should read several ohms, and a near-zero reading means a short on the power bus — typically a failed MOSFET, boost converter, or ceramic capacitor — that must be found before power is ever applied.

Using it on hashboards

On a Bitcoin miner, continuity work is a core step in hashboard repair. The ASICs on a hashboard pass data chip-to-chip in a serial chain, and a single open joint anywhere in that chain stops enumeration for every chip downstream. The standard method is to probe the signal test points — CLK, CI, RO, and RST — at each domain boundary and compare readings against a known-good board: if a signal is present on one side of a boundary and absent on the other, the fault is a boundary component such as a level shifter, coupling capacitor, or series resistor. One caution repeated in factory repair guides: keep the probe away from the heatsinks while measuring, because bridging a probe from a live pad to a grounded radiator can create the very short you were hunting.

Method and limits

Good continuity practice is comparative. Record impedance values from a working board and diff them against the patient, domain by domain, rather than trusting absolute numbers — expected values vary by model and revision. Remember also that the beeper has a threshold (often 30–50 ohms), so a degraded joint can pass a continuity check while still failing under load; when the beep is ambiguous, switch to resistance mode and read the actual number. And because continuity only proves a DC path exists, it says nothing about whether a signal arrives clean at speed — that is oscilloscope territory. Continuity narrows the fault to a link; powered tests on a test fixture confirm the repair actually restored function.

Continuity thinking also extends beyond the board itself. The 18-pin ribbon cable between control board and hashboard is a classic silent failure: a single fractured conductor in the reset line or data pair produces a board that powers up but never enumerates, and pin-to-pin continuity through the cable settles the question in a minute. The same applies to fan headers, sense lines, and the bolted power path — an interface that measures open explains a fault that no amount of chip-level probing will ever find. Cheap, fast, decisive checks belong first in the diagnostic sequence, and continuity is the cheapest of them all.

The tool for the job is any decent multimeter in continuity or diode mode, paired with visual inspection for cold joints and bridges under magnification. For a home miner learning board-level repair, mastering this one test — where does the path break, and is anything connected that shouldn't be — resolves a large share of "dead board" mysteries without any exotic equipment, and it is the skill every other bench technique builds on.