Small Signal Germanium Diode

[Rectangular]

Hey All

I'm working on some synthesizer-based circuits from the early 70s, and the schematics contain a diode called a "DK10", which is apparently a small-signal germanium. I'm not familiar with germaniums,  I was curious if anyone had heard of this DK10 before, or perhaps knew where I could obtain some small signal germanium  replacements. when a circit asks for germanium, does that usually imply that its using some special electrical characteristic of Germ. is being exploited ,   and that silicon wouldn't likely be appropriate for the same application ?  is each germanium diode type very different, or would any old small signal likely do ?

any assistance or enlightenment on these matters would be appreciated  -Rec

[John Lyons]

It's hard to say without seeing the circuit but a "germaniun diode" should be able to work with a 1N34a type as smallbear and aron sells here in the store. If you can post or link the schematic I'm sure someone here can tell you if it's going to work correctly in the specific schematic

[billings]

I may be wrong here, but I think that germanium diodes used to be used in applications that required a low breakdown voltage.  Schottky diodes are now often used as substitutes.  The knee is sharper than germanium, though, so they're not a good substitution where they're being used for the way their signal clipping sounds.

[Rectangular]

I can't upload the schematics,  but here's an example of a circuit that uses four DK10's in various places.  maybe this will give you an idea of what niche they fill

http://www.hinton-instruments.co.uk/ems/vcs3env.gif

[TheBigMan]

Germanium diodes have a lower threshold than silicons.  As billings said Schottkys are sometimes used as replacements, but 1N34A, OA90 and OA91 Ge diodes are still pretty easy to get hold of so I'd use one of them.  Aron sells 1N34As in batches of ten IIRC.

That circuit shows 1N4148s in other places so if they would do the job I'd assume they'd be used.  In other words a silicon signal diode probably won't do in place of the Ges.

[Paul Perry (Frostwave)]

I think that the two in series with pots are chosen for low voltage drop, and the one across the fet is limiting the input signal swing. I could be wrong, though...
in any case, I'd agree that any Ge diode would likely do.
I think anyone rebuilding VCS3 circuits is a glutton for punishment, though

[Rectangular]

talking to someone who's developed this circuit in the past,  I've gathered that I apparently require Germanium diode(s) with a "low leakage" characteristic. I'm familiar with the "forward voltage drop"  matching of diodes, that's important in certain ring modulator circuits (like the Lovetone Ringstinger).. however I have no experience with this "low leakage" thing. How would I go about testing for this characteristic , and can someone de-mystify this for me ?

[R.G.]

A perfect semiconductor junction would let no current at all pass when it's reverse biased. Real semiconductors generate electron/hole pairs inside the reverse biased depletion region of the junction because of thermal agitation. These pairs are torn apart by the voltage gradient inside the reverse biased area before they can recombine and are swept to the ends of the depletion region. This constitutes leakage. If it happens in a transistor collector-base region, it acts just like base current and is amplified. In a diode, it just leaks.

Germanium leaks much worse than silicon, about a factor of 1000. And poorly made germanium leaks even worse. In the crude days of germanium processing, it was HARD to get low leakage germanium anything.

You test for leakage by putting a voltage across the diode and somehow measuring the current that flows through it. This is only hard because the current is in the microampere range.

One way to do this is to use an opamp to amplify the resulting current into a voltage.

You connect the opamp to a bipolar (+/-) power supply. Connect a resistor from output to - input. Connect the + input to ground in the middle of the power supplies. Connect the anode of the diode under test to the - input and the cathode of the diode under test to a negative voltage. The negative power supply will do fine.

The output of the opamp will be at positive voltage equal to the diode leakage times the feedback resistor. So if the diode leaks one microamp, a 1M resistor will make the output be +1V.

For low leakage, you either test and select, or use silicon. If the diodes there are being used as clamps, I think that using silicon Schottky diodes would work. Perusing the schemo, it looks like all four diodes might be OK replaced by Shottky diodes. Otherwise, set up a test rig and try to get under 1uA leakage on germanium junctions. You can use 1N34a diodes, but you also may get good results using the collector-base junction of germanium transistors which are too low gain to be used for other circuit.

I'm guessing at the uses, of course. The diodes on the 1M On and Off pots are polarity separators, used to keep one pot from affecting the other's operation. Those probably matter more than D27 and D29. I think these last two are clamps and could be replaced by Schottky easily.

[Rectangular]

hey RG

thank you very much for that little explanation, it was exceedingly helpful. the way people would mention it,  I kept thinking that leakage was desired attribute, not a flaw... but if  it's not desired, I'd best try out higher grade components. concerning those two dk10's on the pots, the polarity separators,  what would be the ideal moden silicon diode to use there ? a 1n4004 ?  or would a 1n34a germanium suffice ? I'm still not quite sure if I should try and preserve the germanium, or just move on to modern schottky and silicon

I require a bunch of 1n34a's for the lovetone Ring Stinger clone anyway, apparently enough to get 4 with matched voltage drops, so I should have enough left over to play around with. are there any other germaniums still being manufactured that might be of use here ?

-rec

[Rectangular]

Hey R.G.

so I rigged up your little diode test circuit,  but the only dual voltage opamp I had on hand was an MC1709. are you sure I should be connecting the anode to the inverting input, and the cathode to the negative power supply voltage, and not the other way around ? I was using a +15/-15v power supply I use for modular synths ,   but the output of the opamp was giving me a voltage, even when the diode wasn't rigged up (!). when testing the diodes I was getting -15v output (well, -14.XX,  which I assume was the voltage drop) and when I reversed the diode, I got 0.0X volts. I'm using a batch of 1n34a's I bought in bulk for my ringstinger.  the feedback resistor was 1M

I'll grab some 741s and give it another try, but any follow up on this would be extremely helpful. in particular, what schottky's would be used as clamps,  I know nothing of schottkys, some part names would be great 

-Rec

[Rectangular]

bump