this page describes my encounter and restoration of the
Hewlett-Packard HP6282A variable DC laboratory power supply
this treatise was prepared for the monash university esolutions Hack Day program
October 2020 by ralph klimek VK3ZZC
copyright declaration: my material is copyleft. It is freely available for all honorable uses.
The HP trademarks remain the property of HP and its successors.
This power supply was biult and marketed by Hewlett Packard about 1966. It is a very early model low volage DC power supply aimed at laboratory use. Its target market would have been the first generation of engineers who were using transistors and the early integrated circuits for the first time.
It produces 0 to 12V DC, with a settable current limit. It can be used in both constant voltage or constant current mode.
The unit was designed to be rack mountable with additional hardware.
Like all equipment made by HP in these years, the design was a highly advanced and complex for the mid sixties.
There are no integrated circuits, op amps are made from discrete components. The only integrated components
used is an exotic , single die, dual transistor used for differential amplifiers in the main circuit.
These days, for a project requirement like this we would simply use a three terminal regulator, like an LM317.
A a few additional elements will give us variable voltage output, variable current limiting with perhaps only
15-20 additional components.
But , HP did not (and still dont) do it this way.
The circuit and manual which can still be downloaded from Agilent today. Agilent and Keysight are the Test and Measurement spinoffs from HP. Agilent, very generously , make availalbe service manuals for HP equipment for older and obsolete equipment as a public service. You can get the functional equivalent of an electrical engineering degree just be reading these manuals. Much of their classical age equipment can be restored and maintained by a determined amatuer in this day and age.
The HP design includes remote voltage sensing, remote commanding, remote current sensing, and full galvanic isolation with assorted wire strappable options.
The design does include a switch mode element, long before the technique became popular or affordable. The power semiconductors used here would have been considered "alien" technology back in 1965. Here in Australia we played with little germanium point contact transistors that burned out at milliwatt levels. A 10 ampere transistor was the most outrageous science fiction...and Silicon, what on earth was that ?
The heatsink for the TO3 packaged SCRS is very small and internal. The linear, low noise, fine control, regulator is very inefficient as evidenced by the large, generous cast aluminum heatsink on the rear.
The circuit has multiple feedback loops to achieve fine voltage and current control and the feedback loops ensured that the output is remarkably free of noise. This is a very valuable feature to me as an Amatuer Radio Operater and justified my time in restoring this equipment.
Finding the unit.
On my daily campus walk I make it a point to inspect the assorted dumpsters . Here I inspect them for "non-compliant" items. A "failure to comply" will result in me confiscating the item so that " a full report can be made". This is the story I tell those in minor officialdom that challenge me. Here we look down at dumpster divers. Today's non-compliant item was the carcass of a HP-6282A power supply. It was in the metal recycling skip and had been expelled , in disgrace, from the Engineering Faculty.
I dithered a while over this. I do not require even more junk. It was in extremely poor condition. The covers were missing, the chassis was bent at a funny angle, the front panel meter had been smashed. It was under other heavy metal scrap and been out in the rain for a few days as well. The metal work showed signs of deliberate physical abuse. Someone had tried to repair this unit, given up, and then showed it who was boss with a chunk of angle iron.
Its prospects of success were poor. Yet, it was a piece of HP's golden age. It still was adorned with its authentic HP badge. The main circuit board was intact, despite sustaining an obvious impact injury. What to do ?
It followed me home. (all by itself ! ) It is the HP way. No further explanation will be tendered for your approval.
The main circuit board sustained significant damage when it had been hurled into the metal recycling skip.
Nothing that could not be easily replaced was missing or damaged. Horrors ! Then I spotted it. An unobtainable dual transistor , two single die transistors inna tin, was smashed beyond rescue. Of all the plentifully abundant components, only the rarest had been smashed. In the most critical part of the circuit ! The main filter capacitor was missing. It must have exploded. I will tell you why in a minute. The orginal HP front panel moving coil meter, having sustained deliberate smashing, was still working ! The meter movement had escaped ! Gentlemen, we have the technology, we can rebuild him !
A large pcb circuit board trace, carrying the main transformer secondary current had burned out. There were other zaps in this area . Clearly, a prior engineer has attempted a repair and given up. I then begin a methodical check of EVERY component. A simple multimeter will reveal semiconductors and capacitors that have become short circuits and resistors that have aquired impossible resistances and also obvious short circuits where they are clearly not required. At least 80 % of electronic faults can be diagnosed and repaired this way, even without possession of the circuit or even having to understand the design.
I must say, I am easily impressed !
Apart from the minor physical damage, nothing had measured wrong. I replaced the broken double transistor with a matched pair from my collection. I sorted through a batch of npn silicon transistors with a beta tester untill I found a pair with the same current gain. This would make a functional, but imperfect replacement . A new filter capacitor, fuse, and new power cord and then I fixed the burned out PCB trace with a jumper link and power the thing up , slowly ramping up the voltage through a variac. At 120 volts input, the little neon mains indicator comes on. No output yet. "Igor", I shout into the dank and fetid laboratory air, "increase the power" ! At 200 volts input the mains fuse blows. I put in a bigger fuse. It blows at 200 Volts. Now I monitor the mains input current with an ac ammeter. At 200 volts, it is drawing something very wrong. Four amperes. That is 800VA going in, nothing coming out with the exception of the smell of distressed power transformer.
Lemon and Violin merangue.
There was something wrong with the big power transformer. Gotterdammerung ! I yell at Igor and command him to consign this corpse to the mortuary. Ach und Vie ! What ahh waste of time. Alas, I am undone! Local peasantry gather at the castle gate brandishing pitchforks and torches.
You may notice from the photographs that this unit once belonged, or still belongs to the University of Maryland. We do not have one of those things in Australia. But the Americans do ! What did the USA have that Australia missed out on ? The Bomb ? the G.O.P ? Hollywood ? Woodstock ?
Bzzzztt. Its Mr Edison, his electric light and his awfull 115Volts DC.
Now Igor dials the input variac down to 115 Volts. Nothing bad happens. Nothing bad keeps happening ! The HP6282 has an internal auxilliary power supply the powers the control circuits and provides bias for the big SCRS. All the bias voltages are correct. Still no output. I had set the strapping options on the rear teminal posts to the default jumpering as per the manual. Nothing. Now we look again the the circuit that Agilent has kindly provided. The manual was wrong ! Now we set the jumpers according to what the schematic and circuit theory requires.
My Monster, its aliiiyvvvvve !
It works , my abomination lives . It nows has a proper big filter capacitor. I manufacture the top and bottom covers from 1 mm powder coat steel plate. The pictures show how I recover the front panel meter and make a new support bracket and housing for the smashed meter movement. I wanted to retain the original HP dial. I create and add carrying handles on the front panel. These should feature on all electronic equipment. Their primary purpose is to create a crash barrier to protect the front panel controls and instrumentation from mechanical accident. They permit easy lifting. Some savages think that this is their primary purpose.
The original top and bottom covers were made from perforated aluminium sheet. I abhore, detest and abjure ventillation holes on the top of equipment. It has a good intent, however, in the real world it only serves to permit the ingress of dust, metal swarf, moisture, drool... Worse still, it is inevitable that a small rodent will use your valuable equipment as a heated commode. To compensate for the lack of free air flow, i will fit small muffin fans in assorted locations. You can see in the picture, the series pass transistor bias resistors gets very hot in normal operation.
Why had the primary filter capacitor exploded ?, just before it become a short circuit that vaporised that heavy 10A PCB trace? A spectacular and unsuccessfull attempt had been made to power it from Australias 240 Volt mains. This unit is stricktly 115V. After nearly 60 years, all the transistors are still good. The small electrolytics and other passives are still within specification ! This unit has seen sustained use through its working life and severe abuse at the end. I should ask the question, but I regret that you and I know today's most regretable answere.
This HP power supply will occupy an honorable place on my bench. The University of Maryland may require it back, but they will have to pry it from my cold dead hands.
I wonder how it made its way across the known world to Australia and for what unfathomable purpose?
Bonus material. Hint for restorers of old equipment. Mark the position of trimpots as found. Then turn them through their full range and return them to their marked settings. Old trimpots can go open circuit at the wiper position. Simple exercise will restore them.
To be perfectly reasonable and fair to the late Mr. Edison, it would have been extremely difficult to make a light bulb filament with the technology avaiable to him that could sustain 240 Volts. It would appear that 115 Volts was a necessary compromise between transmission efficiency and manufacturability. This, given that the sole commercial purpose of the electric light power company was to power the electric light. Affordable artificial light is without doubt the great enabler of this civilization.
|Fri Oct 2 12:19:57 EST 2020|