Test Equipment for Test Equipment

I started out to write a piece about repairing a cute little Hewlett-Packard bench power-supply. However, the deeper I dug (looking for information), the more I was convinced that I should really talk about test equipment.

It came to me the other day that as companies, there were similarities between Racal and Hewlett-Packard. Both became giants in their own specific field; Communications and Test Equipment respectively. Both were founded by two men. The name Racal is derived from the names of two former Plessey engineers, Ray Brown and Calder Cunningham, who founded Racal in 1950. Hewlett-Packard, rather less subtly, was founded by Bill Hewlett and Dave Packard who met whilst at Stanford University in the 1930s and on occasion worked closely with Fred Terman, who is widely credited (together with William Shockley) as being the father of Silicon Valley. I have a First Edition (albeit 16th impression) copy of Terman’s book, ‘Fundamentals of Radio’.

Both companies excelled in their (initial) field at a time when advances in electronics were thick and fast. In the 50s and 60s Racal produced a seemingly endless range of adapters and enhancements for the RA17 series of communications receivers. Whilst right from its incorporation in 1947, HP was responsible for an ever-growing range of test equipment who’s build quality was legendary.

As well as my collection of Racal HF receivers, I also have a growing collection of Test Equipment. Someone once joked that if you have test equipment, you need more test equipment to maintain it. At first it sounds funny, but in reality its actually very true.

Last week a friend informed me that he was having a clear-out and asked if I was interested in some test equipment. The list included a Sweep Oscillator, a couple of Voltmeters, a small bench Power Supply and a couple of Pulse Generators. The Sweep Oscillator, Voltmeters and Power Supply were all Hewlett-Packard ... no model numbers given ... but its HP test kit! ... and you can't have too much of that, so I said yes please!

Today I came to the conclusion that I can even divide my collection of test equipment up by manufacturer ... Hewlett-Packard, Racal-Dana and Tektronix ... and a few others. My Hewlett-Packard collection was about to get larger.

The Sweep Oscillator, I recognised immediately; an HP8601A. The college where I worked when I left school had one, although in three years, I never saw it being used ... perhaps because at 110MHz, its upper frequency was rather low. As a new Radio Ham in 1976, I myself was somewhat disappointed since back then as a newly licenced Class-B operator, the lowest frequency that I was allowed to operate on was 144MHz. So the 8601A was dismissed as of little use ... at the time.

I powered it up and found that it worked, albeit a bit ‘noisy’ on account of a couple of dirty switches. This issue was resolved by several carefully aimed squirts of De-Oxit. At first I thought the internal FM modulator was inoperative until I connected it to my Racal-Dana 9008 Modulation Meter. It was indeed working. On the other hand, massive levels of Frequency Modulation can be achieved when using an External modulation source. In retrospect, this is very obvious when looking at the schematic. The vernier knob on the output attenuator was damaged, so I was going to have to find a replacement .

The two Voltmeters turned out to be what HP called System Voltmeters. The HP3437A was clearly designed to be part of a larger network of equipment and isn’t really suitable for bench use. It was very likely designed as a means of making and recording precisely timed high-speed voltage measurements.

Whilst one of the 3437s had suffered physical damage to the front panel, internally it was in better condition than the other. The latter had also had the front panel Triax connector (sometimes called a Twinax connector) replaced with a standard BNC connector. I swapped the front panels and replaced the BNC with a Triax connector. I haven’t tested the trigger functions yet, but as a voltmeter it works.

The small bench power-supply turned out to be an HP723A (0-40V @ 500mA) with a smashed ‘Course’ Output-Voltage knob. It was also faulty in that the output wouldn’t go above 27V. Interestingly this is one piece of HP kit for which a manual is not freely available to download. As a result there are several reputable document suppliers who are happy to sell you a copy. I also found at least two decidedly dodgy websites keen to cash in on the fact that the manual is not out there in the public domain.

This little power supply and a slightly larger one (HP6963A) that I was given some time ago are apparently originally by Harrison Labs, who HP went on to acquire. My 6963A is shockingly dirty inside, looking like it was sprayed with dirty water at some point. For this reason, I have NO intention of ever switching it on! Fortunately both PSUs use the same knobs, so I was able to replace the smashed knob on the 723A with one from the 6963A.

After working out most of the 723A’s schematic, I was able to compare it with that of an HP721A (also by Harrison Labs). This then enabled me to re-draw what I had drawn to some degree and then figure out which of the germanium PNP transistors had failed ... a 2N1300 in the voltage reference circuit.





Left: The side panel can be removed to give access to the PCB. The faulty transistor is the one with the red heat-sink. I replaced it with a TO-39 2N2905A. Silicon, not Germanium, but it does the job.




I was curious as to what was going on with the duff 2N1300. I wasn't even sure that it was the original transistor since it was not 'company-coded'. Hewlett-Packard have a long history of identifying their semiconductors with their own HP-specific part-numbers, and this one wasn't. I checked it with the diode feature on my multimeter and indeed the Base-Emitter and Base-Collector junctions appeared OK. But with the meter between Emitter and Collector it was a low-value resistor!

Quite fortuitously I had just assembled a low-cost Chinese-made Transistor Tester, so I popped the 2N1300 into the Zif-Socket and pressed the button. Sure enough, it was detected as a pair of diodes and a low value resistor. Just to prove that the tester 'knew' what it was doing I also tested an unused OC44 and a truly vintage semiconductor, an XB102. The results speak for themselves. Exactly what has happened to the 2N1300, I cannot say, other than it no longer functions as a transistor.
The little HP723A now works beautifully. It pre-dates the adoption of the now standardised IEC Mains Connector and thus sports an oval 3-pin chassis-plug (as on my Systron-Donner Spectrum Analyser) sometimes referred to as a NEMA-Connector or Belden-Connector. I came up with a simple solution for this.


To get around the fact that I didn't have a NEMA-fitted mains-lead, I made my own , sort of. I took three round steel pillars which fitted nicely over each of the three pins. I then took some thin copper tape and cut it into strips about 5mm longer than a pillar. I then rolled each strip once around a small screwdriver whose shaft just fitted through the pillar. I then carefully pushed each rolled-up copper tape into each pillar and tested them for snugness on the connector pins. It worked! The copper tape gives just the right amount of friction and contact. I then soldered appropriate wires of a mains cord into exposed copper solder-spils then finally shrouded each pillar/socket in heat-shrink sleeving.

My Hewlett-Packard test equipment

I also have another oscilloscope, an HP1740A, however the traces have noise on them. One day I may get it out of the attic and attempt to fix it. I do like the 1745A. It is very easy to use and the four rubber feet at the rear make it ideal for 'off the bench' work where it sits nicely on the floor, provided you have a right-angled IEC lead. One unfortunate thing though is that the rubber that these feet are made from deteriorates over time. They become brittle and break apart. I have salvaged the rear feet from my 1740A to replace those on the 1745A, and these too are now breaking apart. For bench work, I use a Textronix 7600 series 'scope.