Ge Diodes

[rocket88]

i figured i would ask everyone that has Ge diodes what their experiences have been. the reason being that that i've been trying to collected a bunch of the ones i use and i've heard great sound clips of. now my understanding is that ge diodes on average have a lower forward voltage then the silicon and led counterparts, somewhere around the .30V range, and they should be glass tubes in D-07 packages not D-35.

so i started testing the ones for forward voltage using my DCA55 and the hand heat test, increase of temp alters forward voltage, similar to Ge transistors. i have and some which include 3 different types of 1n34a (1 D-07 and 2 D-35's), 1n270 (2 D-07's), 1n277, 1n60 (D-35), and OA91/CV7130 (D-07). my results we not what i expected and now i'm not sure some of what i've got are actually Ge.


1n34a - D-07 clear glass 2 black bands: Vf .65V               Vf drop with hand heat        
1n34a - D-35 gold 1 black band: Vf .37V                         Vf no change with hand heat
1n34a - D-35 gold 1 green band: Vf .23V                         Vf drop with hand heat

1n270 - D-07 clear glass 2 black bands: Vf .66V                Vf drop with hand heat        
1n270 - D-07 clear glass 2 black bands: Vf .68V                Vf drop with hand heat      

1n277 - D-07 clear glass 1 black band: Vf .37V                 Vf no change with hand heat          

1n60 - D-35 gold 1 green band: Vf .33V                            Vf drop with hand heat

OA91/CV7130 - D-07 1 purple, 1 tan, 1 orange, 1 black band: Vf .78V                         Vf drop with hand heat

the OA91/CV7130 i know for a fact are the real deal as they were bought from a reliable store that were sealed and marked May 10, 1966. with those a side most of the diodes are in the .60V range, which would indicate that they are either silicon or schottky workalikes, but if they were they would no pass the hand head test. i can tell you that if most of them are swapped into a circuit that had silicon in them, there is a drop in voltage and the expected warm, smooth Ge clipping sound.

i did notice that the DCA55 used a different test current for each diode, so for example the 1n270 test current was 4.77mA while when i tested a 1n4148 the test current was 5.14mA. could the test current cause the higher then expected Vf readings on the diodes? does anyone else have similar results? what are others experiences?

[dbat69]

HI Rocket - I have quite a few different types of what should be Ge diodes - this might be something I check out over the weekend.  Like you I have some genuine ones bought from a reliable source and others which are from China etc - some that I know are workalikes as they are clearly schottkeys (but sold as Ge) but I kinda guessed that when I bought them.

If I get the time, I will add to this.  I need to rebuild my dumbass computer (just hope it doesn't end up like Heath's Little B Chorus) - could give me something to do while I install all the software, and wait for Windows to take forever to sort itself out - mmm wonder what taking a Dremmel to a computer would sound like - sure to sound like some fuzz pedal

[dbat69]

Windows updating is sooooo sloooow - got bored

OK  done a few tests - nothing too much - finger warming test only for 20secs or so - tested with DCA55

Pass = reduced with finger heat,  Fail = no change

Results
OA7        0.33vf   black glass body    Pass
SFD112   0.67vf   glass with green & orange bands     Pass
OA91      0.71vf   glass with black / orange / purple bands     Pass
1N277     0.35vf   blue glass     Pass
1N34A    0.37vf    glass with 2 black bands     Fail
OA85      0.65vf   v large glass with red band     Fail
AA137    0.65vf    glass     Pass
???         0.69vf    unknown russian diode, glass with 2 yellow bands    Fail
OA90     0.67vf    glass    Pass
OA70     0.55vf    glass    Pass
1N276    0.35vf    black glass    Pass
1N270    0.86vf    glass with 2 black bands    Pass
1N60P    0.35vf    DO35 from Tayda     Pass   - I though this is a schottky!
1N34A    0.51vf    glass with black band from Tayda     Pass
1N60      0.51vf    glass DO7     Fail    -  This was from a good source
AA143    0.33vf    glass with 2 black bands     Pass
OA1160  0.57vf    glass     Fail    -  This was from a good source
1N542    0.61vf    glass with green / yellow / red bands     Fail    -  This was from a good source
OA91     0.67vf    glass    Fail    -  This was from a good source

Some surprising results, but a non-scientific test
With the exception of the 1N60P (that passed) they all visibly "look" like Ge diodes
Where I have mentioned only 'glass', these usually have just one black band for the orientation so I've not mentioned it
I only tested 1 diode of each type as I went through my diode boxes, so there could be some variance (well there should be variance for Ge diodes)

Hope this keeps the water muddy

[inefficiency]

Did the random yellow russian one come from the fella we got all the transistors from on ebay? mp38s and the like?

I got a freebie of a few unknown ones with yellow bands from him.

[rocket88]

see this is what drives me nuts, because with all the Ge diodes when you check the datasheet for forward voltage they all say the same thing < 1.00V, but within the community there are accepted values, for instance on almost all the forums (madbeans, DIYSB, FSB) it's accepted that on average Ge diodes have a Vf around .3V and if it's higher then it's not Ge.

i did some searching and found from the AMZ blog an interesting post that Aaron bought "Ge" diodes that according to his results were actually schottky. more interesting is that he has similar results to us.



here's what Aaron said about determining if the diode is Ge or not - "You’ll see from the chart that the fakes had a quite low forward voltage – much less than any real germaniums. Also, the genuine Ge diodes had more variability in the voltages while the fakes were very consistent. Lastly, the fakes are physically too small. They look to be DO-35 package style while real germaniums are all the larger DO-7 package."

this makes me question everything i understood about diodes, forward voltage, and clipping. maybe its that Ge diodes clip at half their Vf as opposed to Si that clip at their full Vf?  its the only think i can think of since there is a known volume drop with Ge diodes when you swap them in place of Si.

dbat - how do each of the diodes sound? i'm always looking for good clipping diodes. in fact i just ordered a bunch more, mostly some different russian d9's.

on another note, inefficency, i forgot to test the super secret special diodes, SFD-112, from our new favorite seller in Bulgaria. i'll post them soon to see if they're the same as dbat

[inefficiency]

Speaking of our Bulgarian friend, I just tried to look at his items for sale and there's nothing there.

Either he's stopped or he's reached his selling limit and had to open a shop/business account or something.

[IvIark]

Those 1N34As with the black bands that have a reading in the .60s with the Peak, read between 0.3 and 0.35 with two other multimeters I have, and I've read some people mention false readings with certain germanium diodes with the DCA55.  Whether the DCA55 or my other meters are giving the incorrect reading is yet to be established, but it seems coincidental that the other meters have them bang in the area you would expect them to be.  If you have another meter just measure the diodes with that and see if you get a different reading too.

[iggy]

i was following your tests as well. and did the same test as Mark. My DVM seems to be pretty accurate on reading for germ diodes between 264 mv - 387 mv. and silicone i tested at 624 mv.

but in the Atlass the germs ran .39v on one but the others were all .54 v- .92 volt. and the silicon was very close to the DVM reading at .68vlt. which would make me think the atlas reading are off.

i checked a bunch of different types from different vendors. and the specs were very similar on all. on the DVM all in the range we'd expect, but very high on the Atlass.

wish there were .65 & .7 fv

[rocket88]

i just had some time to check mine with my DMM rather than my DCA55, and mark is spot on. the DCA55 ends up providing results for Vf that are higher then they should be for Ge diodes. i made the note earlier that i noticed the DCA55 runs a higher test current through the Ge diodes rather then the Si and schottky diodes. i'm not sure why, and i'm thinking about sending Peak an email asking them about it.

i think that besides using your ears to tell if you have Ge diodes or not there are 2 things to check, especially if you're using you're DCA55. number one, do the temperature test, and warm the diode with your hands to see if there is a change in Vf, if there is then it's definitely Ge. number two, do the size test, in that 9 times out of 10 if it's the larger D-07 package then it should be Ge, if its D-35 then it's more likely schottky or Si.

i'm still curious as to what others get, and think it might be a good idea to start compiling a list of voltages with the DCA55 and DMM for the Ge's we have something to compare to when purchasing them online so we don't get taken. also, it would be interesting if we start making a list describing the tone we get from different diodes, just like we were doing with Ge transistors, and might be help in picking clipping diodes.

[iggy]

there's a nice article you might want to take a look at. at this link.


http://music.codydeschenes.com/?p=1615

seems someone with good ears did just that testing with a load of different GE diodes. and wrote a good summary of the different qualities. Very good bit of information. let me know if the link doesn't work and I'll send see you the address.

[rocket88]

thanks man, that's a great article. i randomly found that which is why i was thinking about creating a list. i think midwayfair did an awesome job, but there are a lot of other diodes that are not on the list, i mean look at dbat's stash. i've never heard of some of them. lol

[jubal81]

FWIW, I've come around to just using schottkeys - Bat41, Bat46, Bat85, etc. I match the expected FV and find they sound just as good to my ears, plus the added benefit of being consistent and easy to source.

EDIT: Wanted to add a favorite of mine, the LM385BZ-1.2. These come in a to-92 package and have a FV of about .67. Very smooth and lively and you can get them at Mouser.

[dbat69]

inefficiency wrote

Did the random yellow russian one come from the fella we got all the transistors from on ebay? mp38s and the like?

I got a freebie of a few unknown ones with yellow bands from him.

Hi Inefficiency - yes they were from our friend but I've not been able to find out what they are yet - I haven't seen any with that marking, but I will keep looking.

Rocket, I haven't tried them all yet to see what they sound like.  I'm going to build a nice diode clipping pedal as a test bed - but at the moment I'm just trying to build some of the lovely pedals Mark and Miro keep putting up when I have the time.  I have a large backlog of builds   I think I need a Ge time warp machine (maybe the next build after the 'Deathstar')

It is my intention once I've done the test pedal and got a few more different ones, to add to the info created by 'midwayfair' mentioned in the link above.  I found that too when looking for info on a particular diode - being a relative noob, I wasn't aware of the different types of the same diode and getting different sounds from them .

[iggy]

you use those as a clipping diode? isn't that a voltage regulator or some such thing. how do you attach it. which legs.

[rocket88]

 dbat, i'm sure you can those yellow ones as clipping diodes. check to find the anode and cathode using the DCA55, and put them into the circuit. what's the worst that happens??? they sound bad and you replace them.

but seriously, i've never even heard of some of the diodes you have like AA143, and didn't know there were so many different OA model diodes.

[dbat69]

I just notice ones mentioned in various threads, and get some to have in to try.  

I was intending to have a base pedal by now, but just not got round to it .  I hope to have lots to try and see which one I like best, including schottkys and Si types - some will be good and some not.  

[iggy]

what circuit do you guys use for testing diodes? I built a dist+ or 250 , cant remember which. that i hooked up a switchable diode selector to. works pretty well for testing different combinations. i think i has ten diodes each direction and a switchable one for a asymmetrical layout. Is there a better circuit to use for this type of testing?

and yeah dbat has a bunch i've never heard of too.

so what was the verdict on those ones from the GE tranny guy. ASD112? i missed those, but someone else has other ones in that line for sale. asd110 or something. are they worth ordering?



thought i'd offer this up if there was any interest.

any one interested in some 1s1588's i bought a bunch and could swap or share some if you guys have any interest. those are the ones from the original OD-1 and early TS808's i believe. i do know that analog man and providence use them also.

[jubal81]

Yeah, they're an IC meant for a reference voltage, but it works as a diode. You can see in the data sheet that one of the 3 legs is NC and which is the anode/cathode. They're killer in Klones.

[dbat69]

OK I did a similar test with DMM and the results are different as suggested and predicted, so here goes:

Results    Old       New        Voltage drop test (finger heat)
OA7        0.33vf   0.21vf     Pass
SFD112   0.67vf   0.32vf     Pass
OA91      0.71vf   0.36vf     Pass
1N277     0.35vf   0.22vf     Pass
1N34A    0.37vf    0.23vf     Pass was an F
OA85      0.65vf   0.33vf     Pass was an F
AA137    0.65vf    0.32vf     Pass
???         0.69vf    0.32vf     Pass was an F
OA90     0.67vf    0.33vf     Pass
OA70     0.55vf    0.32vf     Pass
1N276    0.35vf    0.22vf     Pass
1N270    0.86vf    0.54vf     Pass
  "             -         0.30vf     Pass  -  look the same but from Tayda
1N60P    0.35vf    0.24vf     Pass   - I though this DO35 is a schottky!
1N34A    0.51vf    0.27vf     Pass
1N60      0.51vf    0.25vf     Pass was an F
AA143    0.33vf    0.20vf     Pass
OA1160  0.57vf    0.29vf     Pass was an F
1N542    0.61vf    0.31vf     Pass was an F
OA91     0.67vf    0.33vf     Pass was an F
AA116      -         0.32vf     Pass

As the result for what I thought was a schottky looked dubious, I tested some BATs and some SI diodes and all showed a voltage drop with finger heat - suggesting this simple test is flawed

Still helped pass the time while Microsoft installs six billion updates on my other 'puter

[iggy]

so even the silicon ones were affected by heat. Think i might have to do some testing on silicon transistors to see if that proves true there as well.

1n270 .54fv on the DVM? how do those sound? and do they have less volume drop from your experience?  Thats the highest i've heard of for a true GE diode. interesting. Is Ebay open?  shopping time.