What happened to my awesome Ge transistors 1 year later???

[vanessa]

Well as you may have guessed (recent posts) lately I've been going nuts figuring out what happened to my CV7112's (OC140) that I bought off eBay last year.

A very nice forumite turned us all on to this deal off eBay and they were snatched up ASAP by us all. I was lucky to pick up about 50 before they were yesterdays news.

I originally tested 25 or so and I can tell you these were fantastic. I had maybe two or three throwaways and the rest were all in perfect hfe ranges for fuzz faces etc. A few days ago I went back to test the rest and all the hfe's are through the roof. Where as before the average was between 60 and 140 hfe they now read between 150 and 300 hfe. I doubled, tripled, quadruple checked my test equipment. I changed out the battery on my DMM. I even re-tested them with another DMM just to be sure and they still came up with the same results.

They have been stored in a moderate temperature environment (@ approx. room temp). These transistors are 30 something years old. I find it hard to believe that they could change hfe without some sort of factor involved.

I thought maybe the cabin pressure during their air transport (from U.K. to U.S.A.) might have something to do with it??? I guess a test would be for others who purchased these back then would be to compare hfe values to mine. Maybe others have experienced the same occurrence?

Ugh!

[Mark Hammer]

They change measured hfe with temperature, even though they haven't "changed" per se.  Possible you handled them differently at each time? or that the heat was on one time but not the other?  or something that might have changed the temperature of the trannies?

[rockgardenlove]

IMO cabin temperature is more likely the cause than cabin pressure.  In a cargo plain it isn't very warm in the cargo bay.

[Peter Snowberg]

Ge transistors are just fussy things that vary wildly with temperature. I'll bet it's just the room temperature.

[phaeton]

And this is precisely why they don't make stuff out of Ge anymore.  Delicacy and consistency. 

[cd]

Don't tell me you just stuck them into the transistor socket in your DMM to test them.

http://geofex.com/Article_Folders/ffselect.htm

[vanessa]

Don't tell me you just stuck them into the transistor socket in your DMM to test them.

http://geofex.com/Article_Folders/ffselect.htm

Come-on CD... cut me some slack here. No I used the tried and tested GEO method as my "test equipment". 

[spudulike]

The CV7112 was possibly the best Ge ever made. Phillips made them using a silicon alloy technique, and they were designed to be extremely temperature stable. Given that they were for computer use (yes, really) they were designed as bi-directional switches and so were extremely linear. This linearity is why they were highly prized for audio use as well. It caused quite a stir when Phillips stopped production quite suddenly in, oh, years ago  .

The CV7112 IS NOT an OC140. The CV is the military spec equivalent , tighter tolerances and batch tested for consistent parameters for military applications. I have around 200 in my spares box, having bought 500 in a job lot for peanuts. All of my CVs test fine with an hfe range so far of 70 - 80, the odd one rising to 110. I borrow a locally owned (electronics club) Mullard transistor tester from the late 70s to test them. And BTW the red spot is the collector ... reversing the C&E for better audio use has been known for some time and I sometimes use the red spot lead as emitter for this reason.

I find it hard to believe that ALL your CVs will have drifted so badly (and all at once), I suspect your test rig.

(And to answer the next question, yes, I am an old git )

<Edit>   Some common pinouts ...

[nelson]

The CV7112 was possibly the best Ge ever made. Phillips made them using a silicon alloy technique, and they were designed to be extremely temperature stable. Given that they were for computer use (yes, really) they were designed as bi-directional switches and so were extremely linear. This linearity is why they were highly prized for audio use as well. It caused quite a stir when Phillips stopped production quite suddenly in, oh, years ago  .

The CV7112 IS NOT an OC140. The CV is the military spec equivalent , tighter tolerances and batch tested for consistent parameters for military applications. I have around 200 in my spares box, having bought 500 in a job lot for peanuts. All of my CVs test fine with an hfe range so far of 70 - 80, the odd one rising to 110. I borrow a locally owned (electronics club) Mullard transistor tester from the late 70s to test them. And BTW the red spot is the collector ... reversing the C&E for better audio use has been known for some time and I sometimes use the red spot lead as emitter for this reason.

I find it hard to believe that ALL your CVs will have drifted so badly (and all at once), I suspect your test rig.

(And to answer the next question, yes, I am an old git )

<Edit>   Some common pinouts ...

This info is golden!

I am glad I have a small stash of them. Around 10 left. I have seen the actual HFE as high as 160 on my small batch.

[trevize]

let me ask a stupid question.

i know that ge trannies are temperature sensitive.
my ge circuits, especially the only buyed one (fulltone soul-bender)
are tenperature unstable (it needs also some minutes of use before stailization, yes the sound changes
in the first minutes)

for what i know the switching transistors are the modern ones. now most transistors seems designed for switching purposes.

i think that linearity is a nonsense when you talk about devices that are just on or off, ie for switching purposes. am i right?
consider that i really don't know much about the question, I'm just curious.

so the best audio transistors should be the non-switching ones.

sorry for the english but explaining this was difficult in italian, in english is even worse!

[vanessa]

"The CV7112 was possibly the best Ge ever made. Phillips made them using a silicon alloy technique, and they were designed to be extremely temperature stable. Given that they were for computer use (yes, really) they were designed as bi-directional switches and so were extremely linear. This linearity is why they were highly prized for audio use as well. It caused quite a stir when Phillips stopped production quite suddenly in, oh, years ago  icon_redface .

The CV7112 IS NOT an OC140. The CV is the military spec equivalent , tighter tolerances and batch tested for consistent parameters for military applications. I have around 200 in my spares box, having bought 500 in a job lot for peanuts. All of my CVs test fine with an hfe range so far of 70 - 80, the odd one rising to 110. I borrow a locally owned (electronics club) Mullard transistor tester from the late 70s to test them. And BTW the red spot is the collector ... reversing the C&E for better audio use has been known for some time and I sometimes use the red spot lead as emitter for this reason.

I find it hard to believe that ALL your CVs will have drifted so badly (and all at once), I suspect your test rig.

(And to answer the next question, yes, I am an old git Smiley )"


Not my test rig...

Why did you not post all this info a year ago?

[mac]

When I test germaniuns I also write down the day of the test. If possibly I repeat the test on summer and winter to have and idea of  hfe & leakage range.

mac

[vanessa]

I live in southern California, USA. The weather here is the same year round (for the most part). It pretty much stays at 70 degrees +/- 10 year round. Not a significant change in temperature say that one would have in the north east USA.

I even retested some of the ones from the original tests (originally several tests over a few weeks, which leaves me stumped) about a year ago and their hfe's are way up from what was originally tested.

I just ask anyone who bought these from 6K8 and you are in the USA if you could retest maybe a few just to see before you all write me off as being crazy. 

[David]

You're not crazy, Vanessa.  The point being made is to get you to eliminate all possible variables from your tests so they're consistent.  Some physical factor is obviously skewing your results.

Examples:  Is the room temperature ALWAYS the same for each test?  Do you adjust it to make it the same?

Is it possible cold (or warm) air is blowing over your test rig?  Further, is it possible that the sun has been shining on your rig for some tests and not for others?
Are you performing your tests with the exact same rig you used in the original tests?  Are you doing the tests at EXACTLY the same voltage?

[Peter Snowberg]

As everybody who has tested Ge transistors has noticed, simply touching the case for a couple of seconds is enough to skew the results by a large amount.

A temperature stabilized room and a regulated power supply are really mandatory for arriving at consistent test results.

[Mark Hammer]

Here's a question whose answer would make lots of folks happy.

Assuming one clamps a heatsink of some sort on the case of the transistor under test, and that the room itself is not any sort of extreme temperature, to what extent will this likely provide consistency of transistor temperature across multiple measurements over time?

In other words, say I have a bin or Ge trannies.  I go to build a pedal and I fish around looking for the one that has the desired characteristics.  Six months later, I go fishing in the same bin for another one or two to build something else.  In both instances, I clamp on a heatsink of some type, give the device a couple of seconds to have its temperature altered by the sinking device, and then measure it.  Am I going to see, or be more likely to see, some stability or consistency in measurements from the same device on multiple occasions?

Alternatively,is this simply a case of having some cotton goves hanging around to pick the trannies up with before measuring and that's enough?

[RDV]

Just finding out that the red dot is the collector is a revelation to me. I thought it was the emitter. No wonder I couldn't get one to sound all that great in my RM test rig. My RM that I built with the smallbear asian GE kicked major booty on those CV7112s with me assuming the red dot was emitter. I'll plugging some trannies in 2nite!

RDV

[MartyMart]

ER .... what happened to my post with the hfe/comparison list !!!
Did I say something bad .. ??

MM   

[vanhansen]

Just finding out that the red dot is the collector is a revelation to me. I thought it was the emitter. No wonder I couldn't get one to sound all that great in my RM test rig. My RM that I built with the smallbear asian GE kicked major booty on those CV7112s with me assuming the red dot was emitter. I'll plugging some trannies in 2nite!

RDV

Same here.  I've got some OC140's and although he said they are not the same as CV7112's, they have the red spot and the bag is labelled "red dot = emitter".

[spudulike]

If the ambient temperature in the test area is stable, and you are not touching the device then temp drift shouldnt be a problem. Connect up the test rig and give the tranny a few seconds (~10 to 20) to stabilise the junctions, then test. How is your test rig built ? Breadboard ? PCB ? Vero ? I would check your test rig internal connections (actually I'd rebuild the rig).

The biggest threat is moisture/humidity in storing the trannies. Early generation Ges had "whisker" type junction leadouts that were prone to vibration and poor manufacture. Later generations had whiskers applied mechanically and were more stable. Ever wondered what the gel is for inside the case ? Vibration damping.

More of a problem with glass case Ge's (OC70/71/81/140 etc) is if the transistors are stored in damp or humid conditions. Over time, the pinout wires oxidise where they enter the glass envelope, and can give slightly false test readings. Eventually, the corrosion around the glass envelope allows contamination to reach the junction & substrate. This leads to some whack results on testing, and kills the tranny. All military Ges were supplied in paper packets - the paper was treated to inhibit corrosion, and the packets were replaced on stored devices after regular inspections.

A major developing problem is "whisker growth" inside the device. This doesnt just affect Ge trannies, it affects all devices containing tin, and has got NASA stumped. Check out this website if you dont believe me. http://nepp.nasa.gov/whisker/index.html

The shuttle is full of old tin-cased devices ... we have already had batches of pacemakers and other critical gear being recalled. It seems that the move to lead-free soldering etc could cause further problems as more tin is used in electronic assemblies.