Sporadic flicker — ST-20V field diagnosis

Why the diagnosis moved

The manual gives the station three states: solid green means the boundary wire is connected properly, flashing green means it's damaged or wrongly connected, off means no power. A coded blink is metronomic — same on-time, same off-time, forever, because a microcontroller is counting it out.

Irregular flicker is not a code. It's a symptom.

Which also rewrites the delay you noticed after plugging in. Not "boots, checks the loop, reports a fault." More likely: boots, sags, dies, boots again.

clearedThe cut in the boundary wire. Repaired, and a fresh wire behaves the same.
clearedThe splice. A brand-new test loop reproduces the fault exactly.
clearedThe loop as a whole. The LED does the same thing with the wire removed entirely.
now leadingThe power path. The adapter, its cable, and the DC socket on the station — which takes every bit of mechanical stress from that cable, and whose solder joints crack.

One rule: the SF-20 stays out of the wall whenever the station is open. Everything you probe on this side is low-voltage DC and harmless — just never put a probe on the 230 V side of the adapter.

Five tests

The first one needs no meter and might end this today. Tap what you actually observe — the verdict assembles as you go.

Test 00 · no meter needed

Make the flicker obey your hand

Station powered and flickering. Now go looking for the flicker with your fingers, watching the LED the whole time. If you can provoke it, you have already found the fault.

Hands · eyes wiggle, in three places

Where to press

1 · the plug at the station
2 · the cable, all along it
3 · the adapter end

Push the barrel plug sideways, up, down. Then run the cable through your fingers, squeezing — pay attention to the last 10 cm at each end and to anywhere the mower could have driven over it. It has form.

Free fix, best case. If the flicker tracks your hand at the plug, you can stop guessing: it's mechanical, it's inside, and it's almost certainly resolderable.

Test 01

Watch the voltage while it flickers

This is the test that settles it, and it has to be done live and under load — station running, cable plugged in, doing its job. Backprobe: slide the probe tips in alongside the barrel plug, or reach the terminals inside if you've already got it open.

V ⎓ · 200 at the station's DC input, powered

The question

Does the voltage dip
when the LED does?

Eyes on the meter and the LED together. You are not reading a number, you are looking for correlation. Nominal is 22 V ⎓ (the plate says 22 V, 2,5 A). If you read −22 V your probes are just swapped — harmless.

A steady reading is real information. It clears the adapter, the cable and the jack in one move, and sends you straight to the loop circuit at Test 03.

Test 02

Split the adapter from the socket

Only if Test 01 showed dips. Now you know the power path is bad — this tells you which end. Pull the barrel plug out of the station and measure the adapter on its own, still live in the wall, driving nothing.

V ⎓ · 200 the adapter alone, unplugged from the station

You should see

21.0 – 24.0 V, unwavering

Unloaded it can sit a little above 22 V. That's normal. What matters is that it holds absolutely still for a good half-minute.

Steady alone but sagging in place is the signature of a bad connection at the jack, not a bad adapter. Don't buy anything — go to Test 04.

Test 03

Is the station still shouting down the wire?

For the case where the power turned out to be clean. Station powered, both loop terminals empty. It should still be generating its pulsed signal whether or not anything is listening.

V ⎓ then V ~ across the two empty loop terminals

You should see

Something. Anything.

A live transmitter puts out pulses, so a cheap meter shows a small restless number — a volt or three, wandering. It won't be a clean figure. You are only looking for signs of life.

Don't over-read this. Cheap meters handle pulsed signals badly, so a low or erratic number proves nothing either way — only an absolute, unmoving zero on both ranges means anything. If your meter has Hz, use it: a working loop driver normally shows a steady frequency, hundreds of Hz up to a few kHz. Silence there is the strongest evidence you can get without opening the case.

Test 04

Open it. Go to the socket first.

Mains unplugged. Screws out of the underside, lift the shell. Everyone's instinct is to inspect the loop terminals — go to the DC input jack instead. It absorbs every tug and every trip over the cable, and its four little solder joints are where that stress lands.

Ω · continuity )))) jack pins → board, then terminals → board

What you're looking for

A dull grey ring
around a solder joint

That halo is a cracked joint — the classic failure on a DC socket. Then check continuity from the jack pin through to the board, and give it a gentle nudge with the probe while watching for the beep to cut out. Only then move on to the loop terminals: water staining, white or green crust, a terminal that won't beep through to its pad.

Reflow, don't just look. A cracked joint can still pass a static continuity test and then open up the moment the cable moves. If the jack joints look at all tired, reflow all four with fresh solder regardless — five minutes, and it's the likeliest repair on the table.

Verdict 0 / 5

No observations yet

Start with Test 00. It costs nothing and it is the single most likely one to end this.

Numbers you'll need

Parkside ST-20V

Input: 22 V ⎓ 2,5 A
Output: 21 V ⎓ 2,4 A
Adapter: SF-20
Protection: IPX4
Manufactured: 09 / 2023
IAN: 462730_2307
Art.-Nr.: 80001357
Service: Grizzly Tools

Power supply, EUIf the adapter itself is dying 92000047

Charging station, completeLast resort 92000066

Check the warranty before you buy anything. Parkside gives 3 years through Lidl. This station was built 09/2023 — if you bought it in 2024, you may still be inside it. A cut boundary wire is only an open circuit and cannot damage the station, so neither a failing adapter nor a dead loop driver is user damage. Quote the IAN and bring the receipt.