2N2222a to92 or to18

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Joe., lisa, luketeaford, Kent

Post Reply
User avatar
Alfrede
Learning to Wiggle
Posts: 43
Joined: Wed Dec 20, 2017 3:43 pm
Location: Germany

2N2222a to92 or to18

Post by Alfrede » Mon Feb 24, 2020 1:02 pm

Hi Folks,

i have a question. In same old designs i see the 2n2222a transistor, have anybody measured if the to18 or the to92 package is better for the circuits.

For example: CGS-35 ?

I found the Datasheets: 2N2222-To92 and 2N2222a TO18

i don't find any differences from the electrical values, exapt the thermal Values.

Is there anything which is better in a better way to use TO-18 instead of TO92?

Greetz

ixtern
Common Wiggler
Posts: 202
Joined: Mon Jan 22, 2018 4:34 am

Re: 2N2222a to92 or to18

Post by ixtern » Mon Feb 24, 2020 1:58 pm

Alfrede wrote:
Mon Feb 24, 2020 1:02 pm
i don't find any differences from the electrical values, exapt the thermal Values.

Is there anything which is better in a better way to use TO-18 instead of TO92?
Better look only. But maybe it's important for somebody - that's why TO-18 transistors cost much more this days.

User avatar
emmaker
Veteran Wiggler
Posts: 550
Joined: Sat Mar 10, 2012 5:07 pm
Location: PDX

Re: 2N2222a to92 or to18

Post by emmaker » Mon Feb 24, 2020 7:12 pm

Usually the metal can packages are made to handle more current than the plastic packages. So the TO92 should work fine when the 2N2222 is being used in the Synthacon VCF.

But the pins are different between the two packages. So pay attention there.

Depending who's data sheet you use the 2N2222 and 2N2222A have different gains. Same with 2N2907 and 2N2907A. Depends on the circuit whether or not that is going to be a problem.

Jay S.

User avatar
EATyourGUITAR
has no life
Posts: 4549
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: 2N2222a to92 or to18

Post by EATyourGUITAR » Wed Feb 26, 2020 1:31 pm

2N2222 has a pretty high bandwidth so it is my opinion that they can introduce or amplify rf where you don't actually want or need it. You are probably building something with a maximum frequency of 20KHz since you are on synth diy forum. Adding 5pf from base to emitter is standard practice when I want to replace 2N3904 with 2N2222. Your results may vary. Try it yourself to verify what I'm talking about.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
mskala
Super Deluxe Wiggler
Posts: 2555
Joined: Thu Mar 27, 2014 7:33 am
Location: Toronto
Contact:

Re: 2N2222a to92 or to18

Post by mskala » Thu Feb 27, 2020 7:28 am

EATyourGUITAR wrote:
Wed Feb 26, 2020 1:31 pm
2N2222 has a pretty high bandwidth so it is my opinion that they can introduce or amplify rf where you don't actually want or need it. You are probably building something with a maximum frequency of 20KHz since you are on synth diy forum. Adding 5pf from base to emitter is standard practice when I want to replace 2N3904 with 2N2222. Your results may vary. Try it yourself to verify what I'm talking about.
It's not significantly different from the 2N3904. The first data sheets I could find on Google had the 2N3904 rated at a gain bandwidth product of 300 MHz, the 2N2222 rated at 250 MHz, and the 2N2222A rated at 300 MHz. I'm not sure that high bandwidth in a transistor works the same way for stability as it would in an op amp anyway, nor that emitter-base would be a good place to put a cap to improve it. (Note the op amp equivalent, cap across the inputs or input to ground, would be a really bad idea.) The wideband gain of a single transistor is usually much lower than an op amp and they're usually operated without a strong feedback path that could create a phase shift back to the input and lead to oscillation. It's when you have multiple transistors forming a higher-gain block with feedback, or in exotic cases like a single transistor running at microwave frequencies, that you start needing to think harder about stability problems.

User avatar
EATyourGUITAR
has no life
Posts: 4549
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: 2N2222a to92 or to18

Post by EATyourGUITAR » Thu Feb 27, 2020 11:40 am

if you look at the default capacitance on a NPN BJT, the BE junction has the most capacitance out of all the combinations. this is the limiting factor for bandwidth for all silicon BJT. it is a fact. by adding 2pf or 5pf we mimic the manufacturing problems typical with most transistors. for this reason it is presumed that the function of the circuit does not change. the AC impedance of the base is only very slightly changed by some small insignificant amount. CB caps increase stability with a path for negative feedback but they greatly decrease bandwith and gain.

as for the datasheets. I have seen 2N2222 with much better specs than that. look around.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

KSS
Super Deluxe Wiggler
Posts: 1885
Joined: Mon Jan 25, 2016 7:28 am

Re: 2N2222a to92 or to18

Post by KSS » Fri Feb 28, 2020 1:44 am

Can of TO18 *might* be better in a high ambient RF or EMF situation. In your example circuit, I doubt you'll see enough difference between the two packages to matter at all.

If you can get a look at the dies and the families assigned to each one, you can go beyond the datasheet to understand relationships between different P/Ns. Which can be quite useful in determining alternative parts. And in understanding choices others have made.

Circa 1979 -according to their databook- Fairchild used a FST0145 die to make a few different P/Ns. 2N2219, 2219A were in the TO-5 pkg, 2N2222, 2222A in TO-18, and became 2N4400, 4401 in a TO-92. Die itself is 15x16.5 for all of them. There are other numbers also which use this same die.

The same databook shows a FST0144 die used for 2N3903, 3904, 2N4123, 4124, 2N3392, 3393, and even that once ubiquitous 2N2712 everyone back then pressed into service as a noise transistor. This 0144 die is 11x18, and is markedly different in appearance from the 0145. Looking like a CDS cell, with interlaced fingers, where the 0145 looks like a childs drawing of a 4 legged cat or dog in Egyptian style profile. Bounded by a round cornered perimeter.

Looking at the 0144 family, one can understand why a 3904 may sub very well in moog circuits expecting a 3392.

User avatar
Grumble
Super Deluxe Wiggler
Posts: 1615
Joined: Thu Dec 17, 2015 5:24 am
Location: U aint much if U aint Dutch
Contact:

Re: 2N2222a to92 or to18

Post by Grumble » Fri Feb 28, 2020 1:45 am

You are probably building something with a maximum frequency of 20KHz since you are on synth diy forum.
Well...yes and no.
Especialy if you want fast rise/fall times for e.g. clocks and gates, you will want a (much) higher bandwidth.

KSS
Super Deluxe Wiggler
Posts: 1885
Joined: Mon Jan 25, 2016 7:28 am

Re: 2N2222a to92 or to18

Post by KSS » Fri Feb 28, 2020 2:23 am

EATyourGUITAR wrote:
Wed Feb 26, 2020 1:31 pm
2N2222 has a pretty high bandwidth
Same databook specifically calls out the '3904' 0144 die for up to 650MHz. In the primary application section. While the range is 250 to 650. The larger -and 'squarer' 0145 die of the 2222 is spec'd lower at 250 min, 350 typical, and no max given. Suggesting the narrow interlace die of the 3904 is better for high frequencies.

Which squares with my 'first thought' response when these two P/N's are mentioned. Have always thought of the 2222 as a 'high' current switch first, which can generally amplify. And the 3904 as a medium current general amplifier, which can also switch.

Both are described as General purpose amplifiers and switches. Specifically in vtheir respective "PRODUCT CHARACTERIZATION", 0144 family is "NPN General Purpose Amplifier and Switch". And the 0145 family is "NPN Small Signal General Purpose Amplifier and Switch." Underneath this is a "PRIMARY APPLICATION" for the 0144 of "General Purpose Amplifier and Switch to IC=100ma and fT=650MHz". For the 0145, this reads "General Purpose Amplifier and Switch to 500mA Collector Current.

Belaboring these details for the OP as part of the general answer to the eternal question, which part and why? Along with the corollary -and often unasked- question Where is the real mojo, if any?

Because it all factors into the answer. The more you know, the less you guess and wonder if what someone online says has relevancy. So noticing the difference in how the mfr sums up their own product, along with the graphs and numbers can lead to better choices.

And while on the topic of old metal can mojo vs plastic packages, it can be tough to hold two truths in place at once. That the old cans may have actually been made 'better' than the current production, while still having worse specs on a part by part basis. Which makes the detrmination of which one is better, only valid on a case by case basis. Bold sweeping mojo statements generally don't address this.

User avatar
EATyourGUITAR
has no life
Posts: 4549
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: 2N2222a to92 or to18

Post by EATyourGUITAR » Fri Feb 28, 2020 9:02 am

I could be wrong about the maximum bandwidth of 2N2222. however, in my guitar pedals that had RF problems, overly noisy, oscillation, increased sensitivity, non-linear gain (not a compressor), almost all of them had one thing in common. they all had 2N2222. the cascode totem pole circuit was the worst. some people call it a mu amp. the russian germanium transistors I have that claim to be 3GHz PNP are the darkest sounding PNP amplifier that I have tested out of all the transistors I have tested. I know that it is all my fault. there is something I need to do to the circuit to get that flat frequency response up to 3GHz.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

Post Reply

Return to “Music Tech DIY”