Westector characteristics?

[ronbryan]

Does anyone know what I should expect when testing a Westector copper-oxide rectifier, used as an AM detector in a Wartime Civilian Receiver.

The one I have here measures 120k in the forward direction and 5M in the reverse direction, when using an AVO 8 on the ohms x100 range (15V battery).

I have read that the design of the radio ensures that there is a DC bias on the rectifier to ensure it is biased into conduction, but the forward resistance I have measured is comparable with the detector load resistance, which can't be a good thing.

Thanks

Ron

[David G4EBT]

To quote from Chas Miller's 'Vintage Radio & Audio' book:

8-<

"The forward resistance should be a fraction of the backward resistance. If the two should happen to be much the same, it is is almost certainly faulty".

"A small germanium diode can be substituted".

8-<

Source: https://books.google.co.uk/books?id=...tector&f=false

The large difference between forward and reverse measurements on your Westector suggests that it may be working Ron.

Chas mentions that the Westector (a term now used for an over-current device incidentally), was used briefly in a few superhets in the early 1930s before double-diode-triodes came along, and the only instance of it being widely used is in the Wartime Civilian Receiver in 1944.

Be interesting to compare the characteristics with a germanium diode - OA91 or whatever. If need be, plenty of space in the glass body of the Westector to fit one.

[trh01uk]

The Westector diode was famously used in the military WS38 manpack, which was produced in large numbers (over 100,000 I believe).

The radiomuseum site has extensive data on the WX6 diode - see http://www.radiomuseum.org/tubes/tube_wx6.html

By modern standards, the diode characteristic is a poor one. And that's even when they are working to spec. You can't necessarily judge your diode by its forward resistance as measured by an ohmmeter - as the value will depend heavily on the voltage used by the meter. A modern DVM uses very low voltages typically, and the forward resistance will measure artificially high. The only real way to do it is to put the diode in series with a current source (e.g. a psu with a resistor to limit the current), and measure its I-V characteristic - both forward and reverse).

I've found that when they meet the spec, they are fine. When they don't - and most of the ones I have measured don't - their reverse leakage current is unacceptably high, and not far different from the forward current. In other words the "diode" has transformer into a non-linear resistor with little forward-reverse difference.

If you are very careful and you have a supply of spare westectors you can repair them. Inside, they are made up of a number of very tiny metal rectifier "pills" which are in series. You can identify the faulty one (or ones) and replace them with good ones from another sample. Fiddly - but possible.

Richard

[lesmw0sec]

Originally Posted by trh01uk

You can't necessarily judge your diode by its forward resistance as measured by an ohmmeter - as the value will depend heavily on the voltage used by the meter. A modern DVM uses very low voltages typically

Some of the digital testers are not even satisfactory for checking a silicon diode, due to the low applied voltage! I always rely on my trusty AVO 8 for diode checking!

Les.

[FIXITNOW]

Westector Metal Rectifiers

The history of commercially produced metal rectifiers officially began in 1928 and continued until the more familiar semiconductor rectifiers ousted them sometime in the 1960s. They were used in communications equipments for detection, automatic volume control applications, noise suppression, and clipping, and in power units for radio and television receivers and battery chargers.

The first Westectors appeared in 1933. At a little over an inch long and rated at 25 piv, 0.1mA and 5MHz, the W6 was used in equipments manufactured during WWII. This is one of four kindly donated by Barry Smith, G4IAT. What were they used in, was it an 18 Set I can't remember?

There were five types of Westector by 1938 as follows:-
W4 half wave, 24v peak to peak 0.25mA
W6 half wave, 36v peak to peak 0.28mA
WX6 half wave, 36v peak to peak 0.12mA
WM24 full wave, 24v each side at 0.5mA
WM26 full wave, 36v each side at 0.5mA

Metal rectifiers encountered in the immediate post war era, say around 1950, were selenium, small signal copper oxide and larger types for battery charging. Competing with metal rectifiers after this date were early semiconductor types. For example, germanium types were soon available for relatively low voltage low current rectifier applications (see below). However other semiconductor varieties had already far outstripped metal varieties and had no competition from them. Some, silicon-based types known as "crystal valves", were used for example, in waveguide systems where they were good for over 30GHz. These types had emerged, as a matter of dire need (defence radar), from the discovery of the PN junction in 1940. Metal rectifiers however held their ground in EHT applications, up to over 17kV, and in HT applications in half wave, full wave, voltage doubler and bridge configurations and by 1960 the range of these rectifiers had increased to its peak and then went into sharp decline as silicon semiconductor devices improved and became increasingly cheaper and more efficient. Although semiconductor devices were certainly feasible much earlier, their development for mass markets was marking time whilst refinement techniques of silicon and germanium to acceptable purities was being perfected.
Strangely, although the metal rectifier was available before 1930, it never really became popular enough to oust valves which were still used in most HT and detection applications to the end of the valve TV era. Although some manufacturers used the odd metal rectifier, for example Lotus, Eddystone and Grundig, their most common application remained in battery chargers and battery eliminators and in multimeters for providing AC ranges although there was a small surge in popularity of the flat, aluminium-cased, selenium rectifier in the last AC only rather than AC/DC, valve radios and tape recorders. In AC/DC sets I suppose it was better to use a rectifier valve with its heater requirement than to waste the power (if the valve was substituted by a metal rectifier) in a ballast resistor.

[G6Tanuki]

One thing worthy of note about Westectors is that - unlike Germanium point-contact diodes - the Westector has a rather high parallel equivalent-capacitance, making their efficiency fall off rapidly with increasing frequency.

Fine as a detector at low IFs (the aforementioned WS38 used a 285KHz IF) but diminishingly-good as frequency rises: you certainly wouldn't want to use a pair of them in a 10.7MHz FM discriminator!

I have in the past taken the end-caps off, removed the 'pills' and inserted a gold-bonded diode [OA47] in there as a replacement.

[trh01uk]

Here's a characteristic curve I found on the radiomuseum site. It shows a forward resistance of about 6500 ohms at 3V forward voltage.

Reverse resistance varies about I make it about 2Mohms at about 60V reverse.

That's a reasonable diode provided your load resistance is a lot higher than the forward resistance.

Richard

[WME_bill]

Westectors. June 1956.
I found the attached Westinghouse brochure, giving the specifications and suggested uses of these small signal metal rectifiers. The method of testing them may be of interest, and it also gives the forward and reverse characteristics. wme_bill

[ronbryan]

Thank you all, for your suggestions and contributions and to Richard and Bill for the characteristic curves and the Westector data sheet.

I've plotted the characteristic curve for my WX6 Westector and as you can see the forward voltage at the 0.3mA point is just over 23V, which for a six element device is well over the 2V per element maximum described in the Testing section of WME_Bill's data sheet and far more than the voltage shown in the Radio Museum WX6 characteristic curve.

So even with the forward bias provided by the Wartime Civilian Receiver detector circuit, it's not a very good diode now, compared to the Germanium OA70 point contact type I have tacked into circuit.

I've just replaced the WX6 in the set, expecting a very poor performance and to be truthful, it does not seem all that bad! Maybe there is more IF gain than I thought and the AGC is masking the diode losses. The diode load resistor is around 330k.

Ron