History of the Telstra Research Laboratories


The Electrolytic Capacitor Life Test

In the 1970's, the PMG/Telecom carried out Research and Design at 4 levels:-
Level 1 - TRL - carrying out fundamental research (science) plus application work of advanced and/or complex nature.
Level 2 - EDC (variously called at times "Circuit Laboratories", "Electronic Design Centre", etc), who designed for specific applications, mostly to do with telephone exchange equipment.
Level 3 - Transmission Laboratories (otherwise known as "TMC", "Planning Laboratories", etc) who were dual role - their main role was transmission measurements on newly installed cables and pilot-run testing of cable installation practices, but they had a secondary role in design and development of electronics and practices related to transmission, and straightforward stuff not undertaken by EDC or TRL.
Level 4 - Special projects undertaken by any installation or maintenance area of PMG/Telecom.

In theory, all R&D projects where supposed to be on the national Development Register so that duplication did not occur, and funding did. However, most R&D only got on the register at a late stage, if at all, as once someone's pet project or pride and joy was listed, there was a significant risk of a higher level section seizing control, taking it over completely, or canning it.

There was an EDC and a TMC in each State. Each level was supposed to work under the engineering guidance/advice of the level above. In particular, each EDC had a 'Circuits Standards Engineer" whose directives on component choice and design rules we were all supposed to comply with. In practice everybody ignored him. We told him nothing, so he told us nothing.

Level 4 was not really supposed to happen, but occasionally some bright tech or liney would come up with something, and some of those "done-in-the-lunch-room projects" turned out to give a huge benefit to PMG/Telecom. The "Queenslander" Toyota ute mounted controlled tension cable hauler machine was one notable example.

It would be impossible to have that degree of in-house innovation in Telstra today. It's now all either purchased off the shelf, or it doesn't exist.

I worked in TMC WA, and ran or designed/built dozens of R&D projects - all manner of things.

The original intent with crossbar exchanges was that the electronic cards (e.g., MFC equipment) would be sent back to Ericsson for repair. However as the crossbar rollout progressed, and thus the volume of faulty electronic cards grew, the PMG/Telecom had problems with Ericsson. Their turn around was slow and cards often came back still faulty. Also, some cards, doing things required in our network that Ericsson Sweden had not anticipated, were designed by EDC, and I guess Ericsson didn't want to know about them.

Consequently, I and some colleagues were asked to set up a telephone exchange circuit board repair centre in Perth. Due to the volume of board types coming in for repair (let's face it, both Ericsson and EDC-designed boards were designed and built to the most abysmal quality standards), we did it on a semi-production line basis. We had jigs for each board type - plug in a board, and the jig's meters & lights would tell you what was wrong. I designed and built a lot of those jigs.

About the most common fault on Ericsson-manufactured circuit boards was faulty electrolytic capacitors. This was partly because they always used 63 VW capacitors even when they had only a few volts on them (electrolytic capacitors require a "working voltage" to maintain the dielectric layer), and partly because they used their own brand capacitor, which was the worst in the industry. We replaced them with Japanese Elna capacitors, or, sometimes, Philips capacitors. Those old Elna capacitors were very good. I'm still using stereo equipment and test gear I built at home in the 1970's using Elna electrolytics, and they've never gone faulty. The Philips capacitors were pretty damm good too. I have some I purchased for my electronics hobby when I was still at school (1960's), and they still test good. Ducon brand was disgusting.

Pretty soon, after our crossbar PCB repair facility had been established, we got a nasty memo from EDC, telling us we were bad boys for using cheap consumer grade capacitors instead of qualified/certified/approved industrial or military grade parts. Backed by our experience, we naturally replied "Prove there is a better option, otherwise get stuffed." Well, words to that effect anyway.

About a year went past. Then I got a personally delivered copy of a report done by EDC. They had, with the assistance of TRL, done an accelerated life test of various brands of electrolytic capacitors. They had temperature cycled them, applied voltage & current stress cycles, vibration tested them, pulled them apart, and examined them microscopically - a quite thorough job. They only tested certified industrial and military grade capacitors, and did not test any consumer grade parts. But they did show that the Ericsson brand was not the best. One of the very best in the tests was a particular type of Philips capacitor - not the "consumer grade" ones we had been using.

I thought, well, as the report shows which is best, I might as well be a good boy and do what I'm told by the TRL and EDC experts. So I rang up Elcoma (the local agents for Philips electronic parts) and enquired about price and availability of the relevant part numbers. The Elcoma chap said he had not heard of those part numbers, and they were not in the catalogue. I explained that these were high-reliability parts that TRL/EDC had tested, and had told us to use. He then said, "Ok I'll telex Holland and find out the score."

A few days later, the Elcoma man rang me back. He said "I have a telex reply from Holland. The capacitor you want was withdrawn from production three years ago. They were able to fill an order about a year ago for Telecom Research Melbourne from stock on hand. No stock is left. These capacitors were introduced some years ago for applications requiring higher reliability and longer life than the then standard production capacitors.
But they have been withdrawn, as improvements in standard production since then have meant that the standard capacitor now has reliability and life better than the former special production."