On April 2, 2021, a Badcaps member opened a thread titled "HP Z600 PSU Delta DPS-725AB Dead": an HP Z600 workstation that worked fine one day and would not power up the next. The green LED on the Delta DPS-725AB (PCB DPS-725AP) would not light. No burned components, nothing visibly wrong. The repair that followed is one of the cleanest examples on the forum of why a dead power switch chip is very often only half the story.
The first measurement pointed straight at IC902, a Power Integrations TNY279PN in the standby supply. Drain-to-source resistance read 2–3 ohms. For comparison, a brand-new TNY279PN measured 5.5 megohms. That is a textbook shorted MOSFET, so the owner replaced the IC. The PSU still would not start. He also replaced the 100 nF capacitor on the BP/M line, on the theory that the bypass capacitor had gone with the chip. Still nothing.
The numbers he posted next are the interesting part. Main filter caps: about 322 V. EN/UV pin: 1.89 V. BP/M pin: 1.80 V. Drain-to-source: 3.90 V. The BP/M reading was the tell — the TNY274-280 datasheet describes BP/M as the connection point for the externally-bypassed, internally generated 5.85 V supply. A BP/M pin sitting at 1.80 V means the chip's internal rail is not being fed. The chip was new and the bypass capacitor was new, so the feed itself had to be suspect.
Forum member R_J told him to check R901. R901 is a 10 ohm fusible resistor marked "10ohm F1J". Desoldered and tested, it read completely open — not 10 ohms. The replacement chosen was a Royal Ohm FRN01WK0100A10, a 1 W fusible metal-film part of the same 10 ohm value. R_J also suggested replacing C905 and C907: small electrolytics in the standby circuit that runs 24 hours a day, every day, and whose electrolyte dries out over years of that duty cycle.
With IC902, R901, C905 and C907 all replaced, the PSU came back. The owner stress-tested it and cycled it on and off repeatedly. The thread then quietly accumulated data points for three years. A second member confirmed the TNY280PN works as a direct substitute for the TNY279PN in the DPS-725AB ("i just repair my dps-725ab by this chip, and work well now"). A third tried a TNY276PN because no 279 was in stock locally and hit a BIOS "Non-OEM hardware" warning — reported in the thread, with no confirmed root cause. A fourth, a biomedical engineer with several of these workstations, fixed machines that forced a shutdown after less than three minutes by fitting a TNY280 and replacing the 100 µF/50 V capacitors.
The output-power table in the TNY274-280 datasheet makes the substitute hierarchy visible: TNY279 is 18 W in a sealed adapter or 32 W peak/open-frame, TNY280 is 20 W or 36.5 W, TNY276 is 10 W or 19 W. Two further details explain why this failure mode plays out the way it does. The BP/M pin is more than a supply rail: the capacitor on it selects the current limit. 0.1 µF gives the standard limit, 1 µF drops the limit to the next smaller device, 10 µF raises it for TNY275-280. And if the current into the bypass pin exceeds the shutdown threshold, the part latches off until the BP/M voltage falls below 4.9 V on power-down. The standby switcher has exactly one vulnerable supply node — BP/M — and a BP/M pin far below its 5.85 V target tells you the chip is not being fed. On this board the culprit was R901, a 10 ohm fusible resistor that tested open, found alongside the shorted chip. A brand-new chip on a dead feed is indistinguishable from no chip at all.
When you find a shorted switcher, expect its series fusible resistor to have opened, check the small electrolytics in the always-on standby circuit, and use the BP/M pin voltage as your verdict — about 5.85 V when the feed is alive, and nothing close to it when it is not. The full thread, with the owner's measurements and photos, is on Badcaps: HP Z600 PSU Delta DPS-725AB Dead (thread 84038, April 2021 through April 2024).
ICMASS stocks the TNY279PN and the TNY280PN used in these repairs. The datasheet referenced here is the Power Integrations TNY274-280 TinySwitch-III family document, January 2009.
Sources: Badcaps forum thread 84038 "HP Z600 PSU Delta DPS-725AB Dead"; Power Integrations TNY274-280 datasheet (BP/M pin description p.4, Output Power Table p.1, 700 V MOSFET p.1).





