Just making an comment about NPN- DC current gain-- "new guy" observation.

[plexi12000]

i got a couple fairchild 2N5088's.  Data sheet says hfe is min: 100, max: 800

holy crap....the ones i got are well over 1200!  I guess it is true that transistor manufacturing/ tolerances aren't a super-precise thing, is it?? -lol!

[mth5044]

Note the data sheets give the values under certain parameters. Perhaps if you recreate the testing conditions, you'd get similar specs.

[R.G.]

Keep in mind - if your circuit requires a specific current gain, you're in for a world of hurt. There may be a transistor somewhere that does what you want, but then try to find *two* of them.

Then there's temperature sensitivity. In my junior year circuits class, we had a lab project to design a gain-of-1000 amplifier. A couple of the designs had DC conditions that could be changed on a meter watching them by blowing on them from several feet away.

The secret is to design circuits that don't need a specific gain, only *enough* gain. There's been some 70-odd years of work on that in the professional side of the biz. This idea is really what's behind Keen's Second Law: when in doubt, whip in a 2N5088.

[PRR]

> Data sheet says hfe is min: 100, max: 800

Hmmm... I read "300-900" at 100uA, and "no max spec" for the 1mA current many Q-testers aim for.

[plexi12000]

ok that makes sense....about actual "test conditions"...parameters, etcc.  yeah, it's all new to me.  just pickin brains...seein what you guys had to say. 

[plexi12000]

oh--- wanted to ask something else..  i have a couple 2n3078's.  these are "semi-bi polar"?   can i use these in a big muff?

just for kicks...see if they sound good/different

[smallbearelec]

i have a couple 2n3078's.  these are "semi-bi polar"?   can i use these in a big muff?

The datasheet that I found says Hfe 150 min @ 1 ma., no max. So they will surely work, but lower-gain pieces may not get as saturated as you would like. The suggested min for the BMP that I have seen is past threads here is around 200.

[Transmogrifox]

Also depends on the manufacturer.  As has been noted before, just about all semiconductor shops have a "5088" transistor, but what that actually means may vary between manufacturers.

For example, Fairchild does actually state a max of 1200 in their datasheet:
https://www.fairchildsemi.com/datasheets/MM/MMBT5088.pdf

And then ON Semiconductor even has a graph of normalized hFE (DC Current Gain) on top of page 4:
http://www.onsemi.com/pub_link/Collateral/2N5088-D.PDF

They also state a max of 1200.

This is actually a nice thread on the topic of "how to read datasheets".  I think it is an insightful observation you have stated because it points out that analog devices are squirrely things that have different properties with different conditions (currents, voltages, temperature, aging, etc).  Datasheets usually pick just a few conditions best suited to a typical set of uses for the device to give enough to get a rough idea for whether it can do "enough" of whatever it needs to do.  For example, power MOSFET datasheets often provide a transient thermal response curve because in typical applications power MOSFETs can burn insane amounts of power for very short periods of time, but you won't typically find this data for a small-signal transistor because you typically don't consider subjecting it to one-time pulses well beyond its power rating for say, 10 nanoseconds.

Different manufacturers may have different ideas of what is useful/relevant data to their customers and what they consider a typical application for a given device.

Either way my little bit of contribution is sometimes it is educational to look at datasheets from several different manufacturers and see how they agree, see where they differ.  Some simply remain silent on certain details because they don't test a certain parameter.  Second, the pages with charts of different parameters can sometimes help shed light on the range of values stated in the Electrical Characteristics table.

You will notice there aren't many stompboxes that don't use some form of feedback.  You could only do "one-off" types if you rely on specific transistor properties, constantly biasing your circuit with trimpots -- or even more humorous -- put a table of DMM displays on top of your stompbox with knobs to adjust bias at different parts in the circuit as things change with temperature.

[plexi12000]

cool--  thank you small bear!   looks like it's play time!  haha-  thank you Transmogrifox!

and thanks to everyone else for commenting..and taking the time for a "noob".  i ask the dumb questions but it kinda "opens" up...where i should look at, to try to learn more about this stuff.

i really enjoy learning new things.   unfortunately, i REALLy suck at mathematics. always did!  and that is so vital to "electronics" -  lol 

but, hell....it keeps the 'ol brain working, doin this stuff.   i guess i need to hit my library...get a book or two on basic transistor, op amp, elect. theory. that would be a huge help!

[Transmogrifox]

...unfortunately, i REALLy suck at mathematics. always did!  and that is so vital to "electronics"

You may surprise yourself how much you can learn about mathematics when it's part of something you're /actually interested/ in doing. 

The good thing is you don't need to learn very advanced math to design circuits that do cool stuff.  Most of the time you only need a sense of "if I make this resistor bigger, voltage will go down over there", and then you can find the right value on your breadboard without having to do any formal math -- let the circuit do the math for you.

Another thing that is of great value is to learn to use a circuit simulator -- which again, will do the math for you (LTspice is a great free one).  You can do a lot of fast simulations to give yourself a good starting point on the breadboard.  Sometimes it can help you find out what you did wrong when you're pulling your hair out over something you built that isn't working right -- like the "what if I had accidentally made a solder bridge here, what would my DMM say there".

[PRR]

> Fairchild does actually state a max of 1200 in their datasheet:

For the '89 part. Plexi says he has '88s.

And for 100uA--- no max given for 1mA or 10mA.

That's the ON Characteristics, large-signal. There's also a Small Signal section. At 1mA the '88 part may have an hfe (not hFE) of up-to 1400 ('89 part 1800) at 1KHz.

hFE -does- vary with current. There is a trend-curve on page 3 of that datasheet. hFE is likely to be highest in the 1mA-3mA zone. However this is very variable with processing variations, and is NOT guaranteed (it says TYPICAL).

> the topic of "how to read datasheets".

Read them like any contract: *closely*. There are promises on both sides.

YOU promise to observe the Ab Max Ratings. Exceed those voltages, currents, power and it is all your fault.

The maker promises to supply 99+% parts which fall into the Characteristics ranges. If any large number of parts in a truck-load fail to meet these values, *at the stated test conditions*, you can whine for a price adjustment. (Small buyers can't whine.)

The datasheet is written to tell you some things and skimp-over other things that the maker does not wish to point-out.

The main Guaranteed and Rating specs for a "registered" device (the 2N, BC, and 2SC series) "should" be the same for all makers. The "Typical" numbers may differ-- some makers don't bother, others are too helpful. Some sheets are skimpy about parameters that are not brag-worthy.

> aren't many stompboxes that don't use some form of feedback

There are hardly-any *transistor* circuits which do not have some feedback in them. BJT gain has always been too variable to base a design on. As R.G. says: "The secret is to design circuits that don't need a specific gain, only *enough* gain." I was taught "assume hFE>50". If my circuit barfed on a hFE=500 part, points-off on the grade.

[LightSoundGeometry]

I got a batch of those from tade and they are between 400-500 beta uf-aprox 600

so I am guessing my uF=634 is the leakage ?  whats the mathematical formula to tell what the gain is when beta is 449 and uf=634 (it was a random selected one from a lot of ten)

[Rob Strand]

Look at the on-semiconductor datasheet. There is a graph (figure 8.) called DC current gain.  It is actually normalized DC current gain.  The normalize gain is 1 at 100uA  but when you get to say 1mA the gain is 1.8.   If the hFE has a range of 300 to 900 at current 100uA then at 1mA  it will have a gain of 1.8*300 = 540 to 1.8*900 = 1620.

Also,
A common test current for multimeter HFE ranges is to set the base current to 10uA.   If the gain is 1000 then the collector current will be 10mA.