A nightmare on the bench, part 2

My "nightmare" criticisms of this instrument focused primarily on its poorly designed accessibility for servicing.

By Brad Thompson, Contributing Technical Editor -- Test & Measurement World, 2/1/2011 12:00:00 AM

Email

By the time this column appears in print, the Hewlett-Packard 6827A bipolar power supply/amplifier that I described in my previous column ( "A nightmare on the bench," Ref. 1) will be on its way to the recycling center. The troubleshooting effort that I expended over many evenings uncovered multiple problems, including a showstopper.

Readers may recall that my "nightmare" criticisms of this instrument focused primarily on its poorly designed accessibility for servicing. Wiring bundles for the large iron-core power transformer and heat-sink assembly are too short to allow easy repositioning of these heavy components.

SIDEBAR Toss it or test it? Read past "Test Voices" columns. Read more from the February 2011 issue.

After obtaining a copy of the HP6827A's service manual, I found that all raw DC outputs from the supply measured approximately 25% above the voltages shown on the schematic. Several areas on the main circuit board and a plug-in board showed discoloration from overheating.

Noting that the AC-input voltage selector switch allowed for operation for 100-, 120-, 220-, and 240-V power lines, I verified that the switch was set for 120 V. Next, I traced the transformer's primary-side wiring, confirming that 120 V reached to the 120-V primary-winding tap. Finally, I bypassed the input-selector circuitry and connected a variable-voltage AC source to the transformer. With around 90 VAC applied, the raw DC outputs are within specifications.

Also, at 90 VAC, a faint buzzing sound that I had absent-mindedly noticed disappeared. On a hunch, I measured the transformer's excitation current—at 120 VAC, some 75 mA are going somewhere, likely into acoustic energy and core losses. The transformer's primary has thus apparently partially short-circuited, effectively altering the turns ratio and raising the DC output voltages. Finding a replacement for this custom-manufactured transformer is highly unlikely.

At that point, I should have declared victory and moved on, but further probing has uncovered several additional component failures, none of which repair the power supply's malfunctions. Its transformer, circuit boards, and aluminum parts—everything except its meters—are headed for the scrap bins!

---

REFERENCE