[sdiy] Transistor help

[Bob Weigel]

Transistor checkers on meters are good for measuring the response at a 
fairly low signal level.  I've found with some power transistors that 
they measure super low but when you get them in a circuit they actually 
function...for a while anyway.  Then are fail prone :-).  I think it's 
because I'm measuring some kind of impurity that got into the process.  
Some companies get floor scrapings and sell them I'm afraid. Stay away 
from Dalbani who I documented selling remarked transistors and they said 
"oohhh but it out of warrranty.  We can do nothing...good bye!".   
Everything I bought from them save a Kenwood scope was junk in one way 
or another.  Soldering Iron that shorted out.  Fuse holders that were 
some bizarre non-standard design...
      Hehe..ok anyway so all that to say they do have their purposes, 
but they don't give you the entire curve like a curve tracer.  And to 
really be able to see if a transistor is going to work as absolutely 
'matched' you have to be able to see how it responds in some cases over 
a wide range of BOTH base currents and collector currents!  But 
certainly it's better to match than not to match at all with a meter.
      I'm..not familiar with using transistors as a noise source 
really.  When a transistor is noisy it usually means it is defective and 
it's characteristics are likely to be unstable :-).  All devices have 
some noise and it's usually not the most sonically desireable noise 
either I don't think but maybe I'm missing the boat there.  Anyone?
    THe Hfe or "Beta" of a transistor is just a rating of the current 
amplification.  Eg. if you have a base current of 1mA say it allows a 
collector current of 100mA in a given case lets say....until you drive 
the voltage up on it so high that it starts to damage the device.  The 
line is never perfectly flat as you change Vce, but somewhere in the 
middle you pick a point and in this case you would say that the device 
has an Hfe of 100.  Higher gain just means that the base is interfacing 
it's current contribution tot he junction in such a way as to produce a 
more efficient flow of collector current, which could mean several 
things.   Effective area of the junction, how the other pieces fused to 
it (as in the example above where the process wasn't good at 
all...varying degrees of this happen obviously)  and I suppose the 
geometry of the lattices involved at the atomic level.
     I don't know that there is a substantial difference in noise spec 
on a given transistor until it starts to enter the 'floor scraping' 
category :). --Bob

Fiercefish wrote:

> Hi all,
>         I have just bought a batch of 2n3904's for my soundlab 
> minisynth project, I have a meter that can measure the hFE of them so 
> will this allow me to select matched pairs by choosing those with the 
> closest value? Also I apparently need to find some noisy 3904's for 
> the noise generator, what sort of hFE reading (if any) would yield the 
> most noise eg high or low? Sorry if these seem like dumb questions, I 
> do not know if the hFE value is related to these factors but it is the 
> only method I have of testing them, unless there is another way of 
> course. I'd be greatful for any input or other suggestions, thanks.
>  
> PS Does anyone know where I can get some 2k tempco resistors?
>  
> FF
>  
>
>------------------------------------------------------------------------
>
>No virus found in this outgoing message.
>Checked by AVG Anti-Virus.
>Version: 7.0.308 / Virus Database: 266.10.2 - Release Date: 21/04/2005
>  
>