Czy wolisz polską wersję strony elektroda?
Nie, dziękuję Przekieruj mnie tamIs there any advise regarding malfunctioning laboratory power supply DF1730SL5A
• Yes – there is a well-documented set of failure modes and repair practices for the Ningbo-NDN DF1730SL5A/DF-series linear laboratory supplies.
• Begin with mains-side safety checks (cord, IEC inlet, fuse, power-switch, transformer primary).
• Most “dead” or “no-output” cases trace to:
– blown mains fuse / shorted bridge rectifier,
– open or shorted series-pass transistor(s) on the rear heatsink,
– cracked solder joints on the front control / meter PCB, or
– dried low-voltage electrolytic capacitors that power the CS/ICL7106 LCD-meter ICs (≈ 7.5–9 V rail).
• Stable but wrong readings or permanent C.C. mode are usually caused by a drifted current-sense resistor, bad op-amp (LM324) in the control loop, or contamination of the front-panel potentiometer solder joints.
• A full schematic and several repair threads (Elektroda, Badcaps) are available on-line; use them to follow the troubleshooting flow in the next section.
Architecture recap
• Linear bench supply: 230 VAC → mains fuse → toroidal transformer → bridge rectifier → 2×8 200 µF (bulk) → NPN pass array (2N3055 / MJ1500-series) → sense resistor (≈ 0 R05) → output.
• Control board uses UA741 / LM324 op-amps and a TL431-style 2.5 V reference to compare the set-pot voltages with the output; current limit sensed across the shunt.
• Display boards are two CS7106/ICL7106 single-chip 3½-digit LCD meters powered from a local 9 V rail derived from the 7.5 V aux winding and charge-pump.
Symptom-driven flowchart
A. Absolutely dead (no LED, fan, relay click)
• Check mains fuse (T2 A slow-blow).
• Measure transformer primary resistance (should be 2-10 Ω). Open = bad thermal fuse inside transformer.
• If fuse keeps blowing: isolate bridge rectifier (KBP307 or similar). Shorted ≈ 20-50 Ω forward / 0 Ω reverse.
B. Front panel lights, but 0 V at terminals
• Confirm “Output ON/OFF” key is active (some DF17xx variants).
• Listen/feel for output relay; no click → measure 12 V relay rail near driver transistor.
• Measure DC on bulk caps: Vac_secondary ×1.414 (~40-42 Vdc). If present, move to pass stage:
– Measure VCE of each TO-3 device; shorted C-E = permanent C.C. or no output.
– Emitter resistors (0 R22–0 R47) open will also disable output.
C. Goes Constant-Current at small load / above 14 V
• Known quirk when external programming injects voltage via R45→UA741 (Elektroda thread ‑ 2014).
• Inside stock unit: check shunt (50 mΩ) and LM324 pin-3 reference (≈ 100 mV per amp). Drift/open→ false over-current.
D. Output fine, displays read “LO BAT –1.8:88” or all segments
• Typical for oxidised dual-wipe IC sockets on CS7106; re-socket with machine-pin types, clean legs with IPA.
• Measure:
– C3 (input) ≈ 7.5 V (OK in your picture 7.6 V)
– Pins 1 & 26 of CS7106 ≈ 9 V. If also 7.6 V, charge-pump diode or 10 µF cap open → chip enters test mode.
– Pin 37 (TEST) must be tied low (< 0.7 V); floating puts the device in segment-test = all LCD on.
E. Excessive ripple / audible hum
• Bulk caps >10 years old often rise to >0.5 Ω ESR. Ripple under 1 A load should be < 5 mV rms; if tens of mV, replace both 8 200 µF 50 V capacitors.
High-probability repair list (field statistics from Badcaps & EEVblog users)
• ≈ 35 %: cracked solder joints on front board headers / pots / 7-segment ribbon.
• ≈ 25 %: shorted bridge rectifier or TO-3 pass device.
• ≈ 20 %: dried low-V caps feeding CS7106 → garbled display.
• ≈ 10 %: sense resistor drift → spurious C.C.
• ≈ 5 %: transformer thermal fuse open (after fan blockage).
• ≈ 5 %: control op-amp (LM324) latch-up after ESD event.
Measurement references
• Transformer secondary AC: 28–30 Vrms.
• Bulk DC unloaded: 38–42 Vdc.
• Error-amp reference (TL431): 2.50 ± 1 % V.
• Output voltage range: 0–30 V; current limit 0–5 A; typical quiescent ripple < 3 mV rms.
• Documentation: DF-series manual (Goobay/NDN) downloadable from ManualsLib, Manualzz.
• Community schematics: Elektroda topic #2854272 hosts PDF + annotated mods.
• Display failures widely reported 2020–24; replacing socket + 2×10 µF tantalum fixes 90 % of “LO BAT” cases.
• Pass transistor upgrade: MJ15024/MJ15025 pairs improve SOA and derating; widely adopted in hobby repairs.
• Preventive maintenance kits (cap + bridge + shunt) sold by small EU web-shops as of 2024.
• Trend: moving to inexpensive 150 W synchronous-buck SMPS bench supplies (< €90) – but many labs still prefer linear DF-series for low noise.
• Why cracked solder joints? The front PCB is cantilevered by the potentiometer shafts; torque during knob use flexes the board, forming ring fractures around the pot and header pins.
• Why 7.5 V & 9 V rails? CS7106 uses a flying-capacitor charge pump to generate its own +9 V from 6.8–7.5 V input; if the pump cap is open the chip runs only at the input level, falsely flagging “LO BAT”.
• Constant-current latch: The LM324 monitors 100 mV across the 0 R05 shunt; any offset > 3 mV from PCB leakage or bad op-amp input will prematurely trip C.C.
• Mains-powered repair requires competence; lethal voltages present even after disconnecting (bulk caps hold ~40 V).
• Repairs may void CE conformity; recalibration should be logged if the supply is used for ISO-accredited measurements.
• Dispose of replaced electrolytics according to local e-waste regulations.
Potential challenges & fixes
• No schematic → download from Elektroda or trace with camera + KiCad.
• Rare parts (CS7106AGP) – still stocked by UTSource & LCSC; drop-in ICL7106 ok.
• Transformer open: replacement toroid 2×24 V @ 4 A fits original bolt pattern.
• Some DF1730SL5A units were re-branded Goobay DF-1730LCD; front-panel logic differs (microcontroller plus DAC). Check PCB rev.
• High-power testing above 4 A requires 14 AWG wiring and secure binding posts; original posts loosen over time.
• Noise performance after replacing pass devices with modern MJ15024 may improve by 1–2 dB, but ensure driver bias is re-checked.
• Investigate replacing linear pass stage with LDO-assisted MOSFET for ≈ 30 % efficiency gain while preserving low ripple.
• Add digital encoder + DAC (MCP4725) for programmable output as shown in 2016 Elektroda mod.
• Implement NTC-controlled fan to reduce acoustic noise below 20 dBA at < 1 A load.
• Explore open-source calibration firmware (e.g., OpenDPS) adapted to CS7106-based meters.
Resources
• Elektroda threads: “Schemat NDN DF1730SL5A – modyfikacje” (#2854272)
• Badcaps forum: “DF1730SB-5A programmable LCD reading problem” (2020)
• Manual: Goobay DF series, pages 21-25 (troubleshooting).
• ICL7106 datasheet (Renesas) – section 5, TEST pin behaviour.
The DF1730SL5A is a conventional linear 0-30 V, 0-5 A bench supply. Most malfunctions arise from a handful of repeatable faults: mains-side fuse/bridge, worn pass transistors, cracked front-board solder joints, dried electrolytics in the 7.5/9 V meter rail, and drifted shunt resistors/op-amps that force constant-current mode.
Following a structured, rail-by-rail diagnostic—while observing mains safety—lets you isolate the faulty stage quickly. Community schematics and repair logs greatly shorten troubleshooting time, and nearly every part (including the CS7106 LCD-meter IC) remains available in 2024. With inexpensive preventive parts (caps, bridge, sockets) the DF1730SL5A can be restored to accurate, low-noise operation for many more years.
