OA79 - Modern Replacement

[Pgs49sullit]

Hi, I got back into crystal radios a year or so ago and built, using my childhood Philips EE8 set, a fairly decent crystal radio. I have further improved my knowledge and skills such that I now have a small set which receives 8 stations very well. I can hear in the periphery, and see on the oscilloscope, another couple just out of audible reach. So it occurred to me that there may be a modern, more sensitive, equivalent to the 1960s Mullard OA79 germanium detector diode that I am using now. So there you have it, do you know of one? I've tried everything else, more sensitive earpiece, elongated aerial wire etc. It may very well be that what I have now is the best I'm gonna get but I thought perhaps a new diode might help. Can anyone help/guide me? Thanks.

['LIVEWIRE?']

The usual later replacements for the OA79, 80, 81, etc., are OA 90/91 or AA119, though (a) I haven't tried either in a Crystal Radio, and (b) these are now fairly old types. All,however should be available from one or other of the usual suppliers. If you get stuck, I may have a couple of one or other of the above, which, if you PM me your details, you can have for the cost of postage

[paulsherwin]

It's best to experiment with crystal set detectors. You could try one of the junctions of a germanium transistor, or a modern small signal Schottky diode. A scrap transistor radio will usually contain one or more detector diodes.

Ordinary silicon diodes don't work well because of their high voltage drop.

[Pgs49sullit]

Thanks for the advice both, I'll go for a AA119 in the first instance to see if it improves matters. I'm aware that the Schottky diodes are more sensitive but I've got some of those and they haven't improved reception, the reverse in fact is true. I'm sure there will be a Schottky out there that'll do the job better than anything else but I don't really know where to start, there's so many. I therefore rely on the experiences of others in my quest. Thanks so far.

[David G4EBT]

A crystal is a detector - not an amplifier, so any germanium crystal will perform in much he same way as any other. (But not a silicon diode such as the1N914). The modern equivalent generic germanium crystal is the 1N34, which features lower leakage current, high mechanical strength, is optimized for radio frequency response hence can be used in AM, FM and TV-IF applications,
such as AM/FM detectors, ratio detectors, FM discriminators, TV audio detectors, RF input probes and TV video detectors.

They're widely available from e-bay sellers. To comply with forum rules I won't provide an e-bay link but if you google: "10pcs 1N34A Germanium Point Contact Detector Diode FM Am TV Radio Detection" you should find a reputable UK business seller from whom you can buy ten for £1.92 post free. That said, I doubt they'll make any difference whatsoever. A crystal set needs as good an aerial as you can erect, which for most urban dwellings is perhaps 30 ft of wire at maybe 20 ft high. There are lots of little add on simple transistor amplifiers you can build and a number of web sites for enthusiasts, such as:

https://www.midnightscience.net/

http://www.crystalradio.net/

Within their limitations, crystal sets are quite sensitive - that is they're able to detect radio signals despite having no power source. However, their serious limitation is lack of selectivity. That is they're not good at separating one signal from another, so you may hear several stations all a the same time, with thee loudest one drowning out the others.

Back in 1959, there was a BBC design for a crystal and transistor radio intended as a 'lads and dads' project. It was more expensive to build than a typical one-valve TRF set of that era, but nothing like as sensitive, because all it was, was a crystal set with a 2 transistor amplifier tagged onto it.

In 2012, there was a discussion about the BBC Focus on this or another forum, so out of curiosity, I designed a printed circuit board (the original layout was very poor indeed), and wound a replica Repanco DRR2 coil to build one. As I suspected, it lacked selectivity and received several stations, all on top of each other, on a 60ft wire aerial. Back in 1959, I think it would have been an ill-advised project and a waste of pocket money with a disappointing outcome as compared to a typical one-valve radio of that era, but I guess that it was an attempt to provide a design which featured 'newfangled' transistors.

I've attached a couple of pictures for what it's worth, along with an alternative home-wound coil design to the long since obsolete DRR2 coil. I only used the original BBC 'Focus' circuit - the layout, PCB and front panel was my own version. The coil I wound was to the same design as the DRR2, with the same type of former and the same inductances, based on an original borrowed from a friend.

To try to improve selectivity, some constructors use medium wave frame aerials or other directional aerials to try to 'null out' unwanted stations, but what you've most likely to hear on medium wave are 'local radio' stations, which are intended to have a limited range to serve listeners in their area - not further afield.

I hope these notes are of interest and don't sound negative - many people derive huge enjoyment from experimenting with simple crystal sets and some do nothing else.

Every good wish in your endeavours.

[Philips210]

Hi

From memory, I always thought the OA47 gold bonded germanium diode was best suited to crystal set receivers. As David points out, the 1N34A is a good choice.

Regards
Symon.

[G6Tanuki]

Yes the OA47 gold-bonded types are good, but honestly you can use the B-E or C-E junction of any scrap Germanium transistor with pretty good success.

if using Schottky-type diodes remember to choose a *signal* diode not a power one: power diodes often have an effective parallel-capacitance of 100pF or more which as well as losing your precious RF can rather cock-up your tuning calculations!

[Al (astral highway)]

Nice illustration and wrote up, David.

It is helping me to recall with enthusiasm my crystal radio apprenticeship. As a teen I made the classic simple design and threw some wire into an old and huge copper beech tree in our garden, and then over the roof of the garage and into my bedroom. It was 150 feet long!

I was overwhelmed by the reception and even managed to isolate one station for minutes at a time. I then used an OC71 transistor in a simple follower stage. All was cobbled together using a now strange-seeming solder that came in little strips/ tabs from Tandy. I had a manual, brazing iron that had to be heated in a blow torch flame. It needed quite a lot of pAtience to solder a joint neatly with this set up in the garage and no supervision!

[retailer]

I built the crystal set from The Boys Book of Crystal Sets that had dual gang tuning and 2 OA79's - for an antenna I clipped onto the metal lamp shade of my desk lamp. I could leave the headphones on the edge of my desk facing my bed and listen to my favourite stations while lying in bed at night. Not sure what happened to it - I think it was tossed out after I left home. Years later I tried to replicate it for my children but could not come close to the performance of the original.

[ricard]

It would seem that the principal characteristic for germanium in this application is their low forward voltage drop. Would not a Schottky diode such as a BAT42 be a sensible modern component for this, as they also have low forward voltage drops especially at low current?

[Pgs49sullit]

Wow gentlemen, thank you for the inundation of information - particularly you David. I liked the idea of using the diode b/e junction of a germanium transistor so I tried it with an OC71 transistor I had in the box. What I found was that the signal was ever so slightly quieter than the OA79 but very much sharper and clearer. Having said that I've ordered some "10pcs 1N34A Germanium Point Contact Detector Diode FM Am TV Radio Detection" and some BAT42 Schottky diodes to play about with, got both for under a fiver delivered so I'll let you know. Thanks again one and all.

[Richard]

Hi

I built a few as a teen, so long ago now.......

Anyway, long wire aerial out of bedroom window, poor results, longer wire slightly better results, better earth and it came to life, the earth is as important as the aerial.

Richard

Experimenting with crystal sets as a young boy (diodes nicked from TVs on the dump, you get at least four from a dual standard set) I tried galena from a 'minerals set' instead, with a lot of finding the right spot I am sure it was better than any off the dump diode.

[Bazz4CQJ]

Attached is an article on diode selection by SV3ORA, which I think is quite interesting. I recommend you look at the the original URL's as I may have edited the version on file for my specific interests (RF voltmeters). I seem to recall that SV3ORA has a website dedicated to crystal sets.

Postscript; after posting , I checked the two URLs on the attachment and they no longer work, however, Googling SV3ORA gets quite few hits, so his stuff is still on the web.

B

[SiriusHardware]

Am I right in thinking that the use of a battery for biasing the rectifier almost to the point of forward conduction was one of the more advanced methods used for crystal sets?

The theory was essentially that if the rectifier has a VFD of 0.1V then the incoming signal needs to have an amplitude of +/- 100mV before the 'wanted' half cycle even starts to pass through the rectifier, so one way of improving that was to apply a standing DC bias voltage of 'almost VFD' to the rectifier, so that even the smallest (say positive half) signal added to that would be enough to tilt the rectifer into forward conduction.

I might be remembering this all wrong...

[G6Tanuki]

A smidgin of forward-bias can help a diode to get 'over the knee' - I recall various historic circuits which were intended to use rectified RF energy to provide this bit of bias.

Also, I've always wondered why crystal-sets inevitably use a simple single diode as a half-wave rectifier. Two diodes could work as a voltage-doubler so delivering significantly greater output.

[joebog1]

If you dont mind waiting for the mail from Australia, I have umteen dozen different geranium ( yep not a spello ) diodes I can send you for free.
They have very strange numbers though and I have no data for them. I do also have the OA series by the hundreds.

Joe

[Biggles]

I have used the OA85 successfully in crystal sets, simply because I had some around. I see J Birkett is selling these quite cheaply at the moment. The ones I have are in the larger glass package, but they don't have the nice black and red paint on them. I suppose you could paint them for a more authentic look.
Alan.

[mike g3zii]

I have done a bit with crystal sets, I found the best of the diodes was the gold bonded ge oa 47 also a ge transistor oc44 or similar with its collector and base strapped together as one terminal and the emitter as the other.
much better however is the ALD110900 this is a zero threshold dual gate mosfet
in an 8 pin package. details are on google and they are sold on the usual auction site. I use this with a British General vintage coil unit, which has ten tappings.
I use a 500pf cap across the coil and one in series with the aerial to adjust the loading. The aerial is a doublet with open wire feeders strapped together to form a 'T' aerial and a good earth underneath. with this set up I can separate out about 8 stations at good strength on MW and get R4 and RTE on LW without
crossover. this is listening on an ancient pair of ex gov' hi imp 'Phones.
this device is a bit pricey but it works well.

[MrBungle]

Schottky 1n5711's work pretty well. They have a sharper knee than the germanium diodes do and a very low threshold voltage. If you apply a little forward bias, this can improve performance substantially, however it somewhat detracts from the magic of a battery-less crystal set.