In short. Start from the symptom, find the likely cause in measurements, then choose the technology that addresses that cause. Each device has a defined job and a defined limit; using the wrong one hides the problem rather than solving it.
Selection table
| Symptom | Likely cause | Technology | Will not fix |
|---|---|---|---|
| Motors overheat or trip at low voltage | Sustained undervoltage | Servo stabiliser | Outages |
| Resets or lost data during dips | Sags, short interruptions | Online UPS | Long outages |
| Failed electronics after storms | Surges | SPDs | Sustained overvoltage |
| Hot transformers and neutrals | Harmonics | Filters, reactors | Voltage drift |
| Motor heating with uneven phases | Imbalance, phase loss | Phase protection, independent phase | Load imbalance itself |
| Measurement noise and earth issues | Common-mode noise, earth loops | Isolation | Voltage level |
| High bills, penalties | Low power factor | APFC | Harmonics |
| Long outages | Supply failure | Generator with ATS | Short breaks alone |
Method
Reading → baseline → trend → deviation → diagnosis → intervention → verification.
- Record the supply and loads. See monitoring.
- Identify the dominant problem.
- Pick the device with the matching job.
- Check interactions on generator supply.
- Verify with measurements.
Combinations
Many sites need several. See stabiliser and UPS, UPS and DG and integrated architecture.
Illustrative example
A workshop with nuisance trips records 360 to 440 V and daily dips. It installs a servo stabiliser for drift and a small UPS for controllers. Trips fall. Harmonic heating in the main transformer remains and is handled with filters, a separate problem.
A diagnostic path from symptom to cause
- Define the symptom precisely. What failed, when, how often, what else was running?
- Check the basics. Earthing, loose connections, undersized cables, overloaded circuits.
- Measure. Voltage, current, harmonics, events, at supply and load.
- Classify the event. Duration and depth. Use the table above.
- Choose the technology. The one that acts on the identified cause.
- Verify. Measure again and compare.
Combined problems and layered answers
| Combination | Layered answer |
|---|---|
| Drift plus outages | Stabiliser, UPS and generator |
| Harmonics plus resonance risk | Detuned APFC with filters |
| Noise plus drift | Stabiliser and isolation transformer |
| Surges plus outages | Surge protection, UPS, earthing improvement |
Cost-benefit thinking
Estimate the annual cost of the problem: lost production, damaged equipment, service calls. Compare with the lifetime cost of each solution. Do not buy the most expensive option by default, and do not buy the cheapest without checking that it addresses the cause.
Common mistakes
- Treating symptoms: replacing failed equipment without finding the cause.
- Mixing up devices: using a stabiliser for outages or a UPS for motors.
- Skipping earthing and wiring checks.
- No verification after installation.
Getting help
If the cause is unclear, send us your recordings, single-line diagram and a description of failures. We will help you identify the likely cause and a sensible sequence of fixes. Use the application details form above, or call 098852 01608.
The master engineering question
Which problem is costing us most, and does the chosen device actually act on its cause?
What to send us
Send symptoms, recordings and equipment lists. Send your application details.
Questions engineers ask
Can one device solve everything?
Rarely. Use the right device for each cause.
What if I have no data?
Record first; guessing is costly.
Do I need a study?
For complex sites, yes. See harmonic studies.
How do I confirm success?
Measure again under similar conditions.
Values are illustrative. Selection needs site data.