Why this transistor?

Lots of NPN transistors, including the ones you mention, would work equally well there. He probably had a spare 2N2369 in his junk box.

The op-amp needs to work from a single 5V supply, with (ideally) rail-to-rail input and output range - or as close as you can get to it.

I think he just has a big bag of 2N2369s

martin

Because he had one already probably !

Why did you expect to see him use a 2222 or 3904 ? Other devices do exist you know ! Most are quite considerably superior.

Graham

Ok then, I was just wondering if there was something special about it. It does specify a much faster Toff, but that is with a much lighter collector current (if that matters?).

OK, well that's not a big problem either. I notice he only wants a range of

Yes, after taking a look at that, it does seem to be his favorite. ;-)

I refuse to answer a post that is cross-posted to sci.electronics.basics and .design

So I'm doing what you should have done in the first place and answering this only where it belongs, in .basics

YOu assume there is some great design decision when a circuit uses a 2N2222 or a 2N3904.

But the reality is that for whatever reasons, those have come into common use, going back decades. Either someone picked them out of a databook at some point, or they were especially easily to get. It might have even been a simple as at some point in the distant past someone found them on a scrap board, used them, and then wrote a construction article that used them simply because he had them. Or, in the early days of transistors, when it was all such a new thing, there'd be articles where someone would go through the databook, extracting useful for hobbyist devices, and maybe even cross-referencing them to price.

It doesn't particularly matter. After a certain time, those common transistors were being picked because they were being picked. FOr a lot of general applications, they were a general type of transistor that worked. So someone needs a transistor, and they look at the construction articles and "pick" a transistor based on what others are using. Or, the places that sold to the hobbyist had started carrying that device because they saw it was getting some use in the hobby magazines, so it was likely to sell. Since it was easy to get, it was easy to use. One didn't have to dig through datasheets trying to decide.

Whether availability proceeded useage or the reverse, it all combined. Lots and lots of transistor types, but generally only a handful are seen in the hobby world because they are already being used and they are available. What better reason to continue to use them?

The hobby magazines, when they were common, would basically publish the same circuit every few years, with some variation. The variation was usually because someone wanted to build the thing, and they didn't have exactly what was in the last article, so they built it their way with what they had on hand, and wrote about it. The magazines wanted some basic type articles of things people might want, so that wireless microphone or Grid Dip Oscillator or even regenerative receiver would be published every few years and there'd be little variation but some. SOmeone didn't "design" it, they'd usually pick a circuit out of a book or magazine, and use what they had. They did require a certain amount of background to know what substitutes would work in the circuit, but it doesn't take very long to get a feel for that sort of thing.

So if the 2N2222 and 2N3904 (and things like the 2N706 before that, and even the CK722 in the very early days of hobbyist use of transistors) were chosen because they were available and cheap and commonly used, then it's obvious that they aren't unique devices. They work in a lot of cases because they are a pretty bland type of transistor, being used in pretty bland and non-critical circuits.

That means an awful lot of other transistors work too. Which means that someone could "choose" some "oddball" transistor simply because they have it on hand and don't have a 2N2222 (odd as that might sound) and saw that it was similar to the commonly used transistors.

IN some cases, somone might specify some general type transistor not commonly seen because the circuit comes from a semiconductor manufacturer and they want to promote that device.

Or a variant, a kit manufacturer that has construction articles in the hobby magazines ends up finding a good generic device and can get a really good price on them if they buy in enough quantity. PAIA Electronics used to do this with their electronic music kits, putting in really oddball transistors but then mentioning they could get them real cheap. But if the transistors don't go into the places selling to hobbyists, then not only is the single quantity price likely higher than the commonly available devices, but they may not even be easy to get for the hobbyist. What those common transistor types often mean is a code. It means "use a general type NPN silicon transistor with no real outstanding specs because they aren't needed". "A 2N2222? I can use that small signal transistor I pulled off that VCR last week". Any time someone specifies a "1N34" diode, it really means "use a small signal germanium diode here because it really does need a germanium diode for this application". You hoard germanium diodes as they become less and less available, but you don't care what number they have (if they even have a number) and realistically the hobby dealers could list a "1N34" and it really wouldn't matter what the part actually is, because it is the code to say "this is a germanium diode and if you need one, we can sell you one".

IN some cases, someone putting together something may even use the code in reverse, using that house-branded transistor off that VCR and then specifying a 2N2222 or something like that so others could build the circuit without having to find that VCR and get that transistor.

Elektor magazine in Europe did (or maybe still does) institutionalize all of this, where the schematics would often have transistors denoted by "TUP" or "TUN" (I'm not sure I've got that right) which would mean "Transistor, Universal, Pnp [or Npn]" and then on the page where they'd explain their schematic symbols they'd have a bit about how the specific type of transistor wasn't important, just the general range of transistor.

This doesn't mean one can do wholesale substitution without a care. It means that when someone gains a level of understanding, they understand what the parts are doing and they get an idea of what sort of variation is acceptable (this isn't design, this is rule of thumb), they can make substitutes. It also puts them in good stead for realizing that a circuit over there is audio and can get away with a more leeway than that circuit over there that is running at a very high frequency and likely does need more care in layout, that it likely is more critical of smaller variations, and the active devices may need to be a of a better spec than generic transistors (or at the very least, you pull the transistor off something that is using higher frequency transistors like the rf preamplifier of an FM broadcast band radio). Or, you discern from looking at the schematic that it's an audio power amplifier and a low power generic transistor won't have the current capability required. And you realize that you may need to make some slight adjustments when you fiddle with things, to compensate for the changes you did make.

Just because a schematic shows 1.1K resistors doesn't mean you have to use them, or that there's some "design" reason that they were used. I once got a really good deal on a lot of 1.1K resistors, about thirty years ago, and I use them because I have them and they have full leads. For pretty much all useage, they work fine as "1K" resistors that others might specify.

Michael

...Jim Thompson

Try running a 2N2222 at 30MHz square wave/saturated and you see why he uses a 2N2369. Storage time! The alternative is a Baker Clamp.

Don't know about the op amp. I'm sure something cheap and nasty would do the job.

Graham H

That seems to be the consensus.

Generally, when it doesn't really matter, these parts tend to be the ones actually specified or listed as alternates. Why do you think I'm asking? I'd like to know about some decent/better alternatives myself. I was hoping that someone that had used the 2N2369 would speak up and say "ya, ist gut" or something like that. ;-)

Ok, good to know. I'll be sure to stock up on the quality part numbers you provided. Maybe this is why 3904's and 2222's rule?

Like this Eeyore.....this is the kind of thing I was hoping to see......an actual reason to use that part over a more common jelly-bean. I'm guessing this is where the Toff thingy comes in? Of course it's now a moot point since it's an obsolete part, but maybe the designer liked the metal cans.

I still have a bunch of them myself ;-)

...Jim Thompson

Thanks Graham. :-) The fT's aren't that far seperated, so I'd ass-u-med similar characteristics in that area, but I'm learning differently now. ;-)

Ok, he sure makes a deal over it, but doesn't explain why.

Is Toff the thing I should have been looking at to determine this? Gone are the days of loads of charts relating everything, now I mostly see a few paramaters specified that (at least AFAIC) leave no way to extrapolate anything useful. :-?

Yes. Though I've only "looked" at a 2N2369 in the past 20 years ;-)

...Jim Thompson

Hello Anthony,

If I had to choose between the 2N2369 and the 2N2222 (for your application), I would choose the first one.

The 2N2222 is designed for higher current operation (larger die size, higher capacitance). Because of higher Ft and lower reverse capacitance (2n2369 device), the gain of your circuit will be higher, so you need to pick up less signal from your local oscillator. Check the datasheet from Fairchild MMBT2369A (same chip in different package).

At low current (for example 3 mA), the Ft of the 2N2222 device drops to below 200 MHz. It remains high for the 2N2369 device (they specify at 500MHz at 10 mA). When you overdrive (as Jim mentioned), the 2n2369 device remains fast, so you get good pulses out of it. The 2N2222 device will perform less (at your relative low currents).

The 2n3904 is in-between the two.

Regarding the opamp, I do not have the datasheet, maybe somebody knows whether it is equal to the OP-07.

Best Regards,

Wim PA3DJS

When using the 2369 or similar gold-doped transistors with fast switching times just keep in mind that their hfe is much lower than what you may be used to from the 3904. That's the price to pay for speed. IOW there is no free lunch:

BTW the TO-92 version of this one seems to be on the way to memory lane, could become hardcore unobtanium soon. Unless Jim can spare some from his bag.

The author might have used it in the circuit you posted because he needed the speed but not much in terms of current gain. Just one comment here: It isn't a good practice to drive a logic gate straight off the collector of a transistor. There should be a fast Schmitt first. If the input signal is smallish this counter might otherwise display next week's lottery numbers instead of the frequency.

That'd be my guess too. Most of us design with whatever we're comfortable with, & have in the parts drawers. In my case, I use a lot of BC5xx transistors because they're what I grew up up with, & I can buy them in bulk locally at dirt cheap prices.

When you're working from a circuit like that one, as long as you've checked the spec's of the original part, there's no harm in substituting something equivalent that you already have in stock.

Yep. They're ubiquitous in American designs. Yurrupeen designs tend to have lots of BC548s & BC549s for the same reasons.

Very nice write-up, Michael, & extremely helpful for beginners. If there's a FAQ for this group, your post should go in it.