Tube Amps / Music Electronics
|For current discussions, please visit Music Electronics Forum.||New: view Recent Searches.
New: visit Schematic Hell!
The sunn still shines online!
|Listen to great tunes streaming live right now!|
|3/27/2000 2:39 PM|
||How to identify "good" Ge transistors?|
I have a stack of NOS Japanese germanium transistors of various kinds, purchased as variety bubble-packs from a local place that closed. Many of these LOOK like they have desirable characteristics, but not all have met with success when I tried to use them.
I know the following about Ge transistors:
1) the pins are laid out differently
2) it is unlikely to get one that has a gain of >200 or so and is functioning appropriately
However, I am curious as to what other tell-tale signs will let me identify useful ones with your average $50 DMM (i.e., restricting one's self to hfe and diode measurements).
I will point out that Mike Irwin and I were testing some a few weeks ago and noticed that when I had to hold the device in place (the fit in the meter's transistor test socket isn't quite as snug as what it used to be), the gain would not stay steady. Sometimes it would rise and sometimes it would descend. Since I tend to use rechargeable Ni-Cad 9v's in my meter (which only charge up to 8.5v and discharge quicker than carbon-zincs), I wasn't sure if the change was stemming from a change in battery conditions or a change in the device under test. Mike pointed out that Ge devices (especially given the metal can) are temperature sensitive, and sure enough, when I released my grip, or thermally insulated it, the gain stayed steady in most instances.
The question that emerges out of all of this is: "What kind of hfe readings should be taken as a sign of an unstable or unusable device?".
This includes ascending vs descending changes, rate of change, absolute value of change.
|3/28/2000 5:47 AM|
Mark, are you saying that there are some Ge transistors that don't come in metal cases? Could you give me a part number or two?
As far as your question goes, when I was building my Austin Treble Booster (big grin) I followed R.G.'s pdf file. I also came across a Ge with an hfe reading of >1000 (which was great luck -- I'm saving it to build a guitar amp with (another big grin)).
When I was researching Ge Xistors months ago, I talked with people who worked with Ge "back in the day." Invariably, they all talked about how heat sensitive Ge was/is and how they were so greatful when they didn't have to use Ge any more (and what the hell was I trying to build that I needed Ge for, and on like that...).
I also seem to remember R.G. and I having a conversation about Ge one time that involved having to actually reverse-bias the Ge base-emitter junction to turn the transistor off (as opposed to just dipping below 0.7V on the Si base-emitter junction). I can try and dig that up if you're interested.
To sum up, I've never been 100% happy with my Austin Treble Booster because of (what I think is) thermal noise from the Ge Xistor. I still use it constantly and love the overall sound. So, I'd be real interested to try a couple of other Xistors...
|3/28/2000 4:25 PM|
"Mark, are you saying that there are some Ge transistors that don't come in metal cases? Could you give me a part number or two?"
No. I just mentioned it because one tends to overlook the obvious: that when you hold a metal can betwixt thumb and index finger so it sits well in the test socket, you can be sure that body heat is factored into what you measure!
|3/29/2000 5:04 AM|
I didn't think so. Every one I've seen has been in a metal can. I wonder if there's some physical reason? It would seem to me that the metal would tend to make the things more sensitive to heat, not less. Perhaps the metal dissipates the heat produced by the transistor better...
But you're right - I've seen the body heat effect on hfe. One of the guys I talked to said that every time the AC would turn on, his Ge circuit would change - seems obvious now.
Anyway, thanks for the thread. You've got me all jazzed up to try and fix this Austin Treble Booster (grin). I'll be sure to look at the Japanese Xistors, too.
|3/28/2000 1:24 PM|
Just a small heads up on two points: Ge transistors are typically leaky therefore the rollercoaster readings that you witnessed are quite typical. They will however, eventually stabilize within several counts. This is a good Sunday morning with coffee exercise. Also, it is very likely that the metal body of the devices you were testing is connected internally to the collector. This is common, and would explain why touching them caused the readings to jump. As far as the meter socket becoming loose, I have a breadboard set up and have connected short jumpers from the meter socket to the board. With this method, the only risk is wearing out the board.
|3/28/2000 4:36 PM|
I don't get a "jump" in readings, merely something that keeps going up and up. The fluctuation-then-stabilization you describe is VERY familiar. Glad to hear it is a normal part of their behaviour. Indeed, this probably answers my original concerns. I was fairly sure that the unit which started out at an hfe of 115 and ended up at 180 before I took it out of the test socket was probably a bad device, but initially, I wasn't sure if the one which hovered from 112 to 118 and settled at 117 was exhibitting something expected of a normal or bad device. I suspect the heat from my fingers, while trying seat the device right probably exacerbated the phenomenon, further blurring the boundary between good and bad.
Using a breadboard for the test socket is sound practice and a great idea. I think I'll start doing it. Thanks for setting me straight and allaying my fears. I suspect I have a MUCH bigger assortment of usable GE devices now than I thought.
|3/28/2000 5:14 PM|
Probably the easiest way to test these unknown transistors would to put together a working version of the circuit in question with a socket (or several sockets for different pin configurations) to permit plugging in different transistors. Hook up to 9 V and use your ears as a guide. Measuring the collector current with the multimeter and a 9 V battery (with the transistor base unconnected) would help to weed out the really leaky ones. Maybe have a 100 ohm resistor in series with the 9V battery to limit the current to protect the multimeter if the transistor is shorted. Use the diode test function on the meter to check the emitter-base and base-collector junctions, then do an Hfe test on the ones which aren't excessively leaky. Mike I.
|Page 1 of 2||Next>||Last Page>>|