Old Time Radio Engineers

[Mark Hennessy]

O.K. Mark, I just read your comments on voltage and series resistors - I'll probably print out the page here at the library and then stick it in my textbook for reference. I hope you understand I'm not being rude when I seem to delay responding directly to explanations because sometimes I have to do this later on. Let's say I get about 4 or 5 posts but only have an hour and bit, plus e-mails to sort through.... Anyway, thanks for your help and patience.

Hi NM,

No problem at all - I understand the time pressure at the library and everything.

It reminds me of the days before having broadband at home, where you'd set aside a couple of hours for the computer. First, dial up to download the messages and grab all the web pages you wanted to read. Then, once off-line, respond to all the emails and read all the web pages that you got earlier. Then dial up again to get the messages sent and follow any new thoughts. It was very different.

I held off getting broadband for as long as possible because I knew it would take over. And it certainly does. But of course, you can't be without it!

I had a pretty late night trying to advance further because now there is some similar information on testing resistors. This bit was tricky. The fundamental point that gave me pause for thought is that, back in the early Forties, the radio engineers weren't as spoiled as we are. Checking if a resistor was O.K. wasn't simply a case of sticking your DMM on, say, 200 K, and probing a resistor. Times were hard and testing equipment was costly so pocket Milliameters were vital. 

Yes, as you said in your post, current limitations were an issue. The book states (from memory):

1000 R (up to 30 Milliamps on tester)
10000 R (up to 10 Milliamps on tester)
100,000 R (up to 5 Milliamps).

So, for the sheer hell of it, this is what I did in various imaginary scenarios:

With the aid of a calculator, I imagined I needed to test a 20,000 R resistor. The voltage according to the book was 100 volts. According to my calculation, the meter ought to deflect to 0.005 or 5 Milliamps because 100 divided by 20,000 is 0.005.

Absolutely right; 100V will cause 5mA to flow in a 20k resistor.

A movement that has a 5mA full-scale deflection is pretty insensitive, and as you can see, to measure high resistors, you'd need a big set of batteries!

In the famous AVO 8, the FSD is 50 micro-amps. Pretty sure that's right - mine are up in the attic. At least, the most sensitive current range is 50µA, which is pretty handy occasionally.

So now, to measure 100k, we only need around 5V. There is a 15V battery in an AVO 8 for the high-ohms range...

However, here is the punch line: Suppose the resistor is knackered? Suppose it's only able to offer a resistance of a mere 10,000 R? Or suppose the resistance is so low you could get 500 milliamps kicked out. Well, apparently the done thing was to start off with a much smaller voltage and work out the whole scenario through proportion. I won't risk giving any examples in case I make a typo or mistake and jar a few nerves but basically I plotted out a few expected current readings at, say, 30 volts. Then imagined the resistor was damaged and how that could be detected by proportion.

This is a good "thought experiment" to go through

Luckily, resistors tend to fail high.

But even so, the principle is a good one: with a moving-coil meter, which is at constant risk of being damaged via overload*, then yes, you start on the least-sensitive range, and work your way up.

* OK, some meters have overload protection built in, but it's best not to stress them needlessly.

The scenario you're describing suggests that you have a variable voltage source, and by knowing the voltage, and being able to read the current from the meter, then you're able to apply Ohm's law to determine the resistance. That's good - it works. Of course, it depends on how accurately you can determine the voltage. In effect, you have to make 2 measurements in order to determine resistance, so you've got double the opportunity for errors to creep in... Also, there is that risk of damage.

As you might know, a "real" multimeter doesn't work that way. With that, you short the probes together, and the needle goes over to the right hand end of the scale. You can adjust this with a pre-set resistor on the front of the meter. Having done that, there's nothing a resistor can do - whatever the value - to cause the needle to go beyond the right-hand end stop.

The down-side of this approach is that the scale is non-linear. But it's a scheme that served us well for many decades.

A better scheme might be to use current source rather than a voltage source. Then, we pass a known current through the unknown resistance. If we choose the currents sensibly, then we get sensible voltage readings. Imagine the current is 1mA - so if the resistor was 1kΩ, then we'd see 1V.

But the trouble with a current source is that it requires some electronics. A couple of transistors, perhaps. No problem in a digital meter, where there's already plenty of electronics (and that's how most DMMs work), but it's hardly ever seen with analogue meters...

Of course, why bother? All I can say is maybe it gives a kind of perspective on how different things were back then. Not only did most people wear a hat on the street but a fair few seemingly relied upon a pocket milliameter. Even worse, no calculator so they probably did it all on paper!

To be honest, I've no problem with thinking back in time. It helps to give us an understanding of how we got to where we are today, and it's humbling to think how much was achieved with so little. Slide rules rather than calculators - and much more reliance on mental arithmetic (a skill I really admire). I always try to estimate the answer before hitting "enter" on the calculator, just so I know that I didn't make a typo when entering it into the calculator. Try persuading a kid today to do that!

Keep at it!

Mark