So I have nearly all the parts and most of the know-how to finally create my first build. I'm sitting on a bunch of silicon transistors with gains in the mid 200's. I need some with gains closer to 100 to make some decent stompboxes, so I ordered some 2n2222A's from a reputable eBay seller, claiming that these transistors " can switch up to 40V at peak currents of 1A (not continuously, just peak!), with a DC gain of about 100. "
Now my understanding is that the hfe should be around 100, but I got them today, and tested them with both a built-in multimeter tester and the GeoFX transistor tester, and they all have gains around 250 with little variation. None are below 200.
So what am I missing? Is hfe of 250 considered to be "around 100"? Is DC gain not the same as the hfe referred to on these boards? How can I reliably buy transistors with gains that are actually around 100?
What does the data sheet for your jellybeans range of hfe?
If they're within that tolerance then I could be human error or a misleading description on behalf of the seller.
If it's the latter then ask for a refund if you cannot make them work within the circuit.
It's a bit hard to tell without a diagram if they can be made to work.
Good luck and welcome.
What are you planning to build with them? They might be fine.
I only have a few 2N2222s left but iirc they're in the 150 - 250 hfe range.
If you need closer to 100 hfe perhaps a different transistor would be better suited, that said if you've been sold these having been told they're were rated close to 100 I'd contact the seller., "close to 100" is fairly ambiguous language, I'd be reluctant to jump to conclusions that you've been mis-sold.
I dont think they're off spec for 2222's but if the seller said they were close to 100 that an error on their part, deliberate on not.
Thanks Kipper, I have just requested the datasheet from the seller, they should get it to me within 24hours.
In the meantime, is it worthwhile to look up the datasheets for other 2N2222A transistors, or will the characteristics vary too wildly by manufacturer?
When I type in all the printed info on my transistors (which, for the record is 2N 2222A H 331) the first thing that comes up is a datasheet with the following info:
(https://i.postimg.cc/sBnNHGkf/2n2222a.png) (https://postimg.cc/sBnNHGkf)
Either this is accurate to what I have and I need to initiate a return, or this datasheet is inaccurate to what I have, or I'm just still hopelessly misinformed. :icon_confused:
Quote from: GuitarMatt on May 30, 2020, 03:17:09 PM
Is DC gain not the same as the hfe referred to on these boards? How can I reliably buy transistors with gains that are actually around 100?
I've assumed DC gain and hfe to be the same, but I barely know what I'm doing.
The specs you posted (in my eyes anyway) show that a 2N2222 will be rated anywhere between 100 and 300 hfe.
Thanks slashandburn, I'm hoping to build silicon versions of old fuzz-face and tonebender type circuits. I could use transistors with gains even lower than 100 to be honest, but I figured ones around 100 would be reasonably versatile.
I've just seen a few posts on this forum about guys using 2n2222's with gains around 140 or so, so I didn't think I'd be off the mark too bad with this purchase, but as of now I still don't have any transistors with gains lower than 200.
Assuming I'll need to make another purchase, does anyone have any suggestions on silicon transistors that are generally useful for fuzz-face type builds with gains that are within 70 to 130?
You'll get plenty more (and better) advice from more knowledgeable people as time wears on but off the top off my head the only npn silicon transistors I've come across in that range were 2n1711, they were all around 80 to 100 hfe. I think I ended up using them as Q2 in Meathead builds.
I say build whatever your going to build with sockets and swap in lower gain transistors as and when you get them if you're not happy with how the current batch sound.
In the grand scheme things, 250 hfe is still on the low side as far as silicon transistors go.
Should have guessed a silicon fuzz face.
Not sure it apples like in the germanium pnp fuzzface.
I'm no fuzzpert.
If you use the recommended germanium gains in a silicon build it won't sound good, or at least it won't come close to a germ build. I think you want to be 150-200 and even higher.
Are these in the metal cases? I got a bit less than 1000 in a factory box STI and they are measuring 98-130. I have PN2222a and MPS2222a and they are higher. I have seen PN part numbers have a different spec sheet than a 2N sheet.
I'm even less of an expert than kipper. From what I gather the 2n2222 is a good choice for a Si Fuzz face.
Socket the transistors. Build it. Try it. Swap them out later if you're not satisfied.
For silicone FF try BC138 or BC108, I think 70s era FF used those
The text-table you cite is the 1960s RETMA(?) registration. Any "2N2222A" SHOULD fall in these specs. You will notice they are VERY wide, and mostly open-ended.
The hFE of a specific individual varies a LOT with current.
> between 100 and 300 hfe.
At 100mA current! We don't run pedals that hot.
We also do not know if the two testers test at a current remotely similar to the currents in a FF.
Until you AND the vendor agree on test conditions, you do not have any valid complaint. My guess is these are 2N2222As and that at "some" test current the hFE may be "around 100".
Why don't you just try them? (There are billions of uses for '2222 even if they do not happen to mesh with any specific fuzzface.)
When Hfe is specified for silicon devices, +/- 30% is considered relatively close. So, as others have suggested, you likely have grounds for returning if the seller promised gains of "about 100". That doesn't mean that the devices are all unusable. In my experiments:
http://diy.smallbearelec.com/HowTos/BreadboardSiFF/BreadboardSiFF.htm
the FF circuit is OK with a Q2 in the high 100s, though you should expect to have to tweak the bias resistors for best results. Also your 2N2222A are too "hot" to be good for Q1. I like a 2N4123, but that number is Not magic; many other common types get that low. Sorry, my store is not open for ordering till the lockdown in NYC is lifted. Please either check back in a week or so or do some more fishing on eBay. And do invest in a breadboard.
I breadboarded the FF circuit several days ago and have messed around with swapping transistors and component values, etc. It just seems too hot, and I would guess it's because my Q1 has a gain of 211 (the lowest gain transistor I have). At it's best, it sounds like a fuzz face trying to be a metal zone.
Thanks all for the help, I will probably not request a return, as this pack of 100 transistors was only 6 dollars, and I may find some use for them later. But I really only intended to build silicon versions of old 60's fuzz boxes, so I'll need to get some silicon transistors with gains that are truly "about 100".
Also to answer mozz, no these transistors are the typical modern black plastic veriety, not metal casings. And thanks for the info, I did not know silicon versions of FF style pedals require higher gain than germ.
Quote from: GuitarMatt on May 30, 2020, 05:13:41 PM
But I really only intended to build silicon versions of old 60's fuzz boxes, so I'll need to get some silicon transistors with gains that are truly "about 100".
Are you adverse to going germanium? Might be the answer to your prayers
Would it help to try "reverse beta" for lower gain? This is where you exchange the emitter and collector - although that can result in really low Hfe.
I'm not too against germanium, (I have some on order from Russia, but they may not get here till July :icon_rolleyes:)
My biggest hurdle with germanium is the whole charge pump thing to get it to work with a typical 9V jack supply. I haven't really researched it much, and silicon just seems easier to help me get off the ground, so to speak.
I just figured since I'm sitting on this pile of silicons, and they seem to be a dime a dozen I may as well try to build some of the germ circuits I'm planning to build later with these.
And to anotherjim, I can try to reverse beta, but I have never heard of that before. If I do that on a transistor with gain of 250, what could I reasonably expect the new gain to be? Edit: I just reversed some with gains of about 280 and they came back at 20 or lower. Interesting, but probably too low to be usable.
>> I will probably not request a return, as this pack of 100 transistors was only 6 dollars
Hold the phone. 2n2222 at 3c each? Does this seller have any more of those?
Edit: 6cents. I can't count at the best of times but Its also late and I'm drunk.
>> Hold the phone. 2n2222 at 3c each? Does this seller have any more of those?
Go crazy. Or let me know if I bought the wrong thing outright: https://www.ebay.com/itm/163662244123
I don't know if he is a "reputable" seller, looks like he just resells chinese stuff. I asked if they were the metal can type because if you go to digikey or mouser 95% of them are the metal can variety. TO-18. If i see a TO 92 package, it is most likely some other transistor that way exceeds the 2n2222a spec, and is relabeled. I really hate buying devices on eBay becuase a lot are relabeled, or fakes. Now find a seller who has old stock, old date codes, they are most likely legit and adhere to the original specs closer. I spend a lot of time unsoldering parts from old circuit boards i know were made in the 50's 60's 70's, today was pairs of OC45, OC71,OC72. Made my day.
Search the forum
You should be able to build a good Si FF with what you bought.
Beta changes with how the test is done different testers will give different readings 2nd link has a table in reply 1 that are real measurements I took.
I like 2N2222s around 200
Others have posted circuits there are calculators and LT spice etc. to help adjust FF like circuits
Two threads I started that might help
https://www.diystompboxes.com/smfforum/index.php?topic=114153.msg1059002#msg1059002 (https://www.diystompboxes.com/smfforum/index.php?topic=114153.msg1059002#msg1059002)
https://www.diystompboxes.com/smfforum/index.php?topic=114210.msg1059667#msg1059667 (https://www.diystompboxes.com/smfforum/index.php?topic=114210.msg1059667#msg1059667)
Thanks mozz, my "reputable seller" designation was based purely off his good eBay seller rating.
Is it really that common for electronic parts like this to be fake? I can understand why somebody would sell bogus parts, but couldn't they just as easily do it by trying to pass it off as NOS, or putting it in a metal casing before relabeling it? I thought that was part of why we test their gains, so regardless of their external casing, you know something about how it behaves.
How would I verify the authenticity of "old date codes"? I'm totally new to this sort of thing.
Thank you Gus, I will definitely check out those threads!
I have used (and have stock of) 2N2222 in TO-18 and TO 92. They work nicely for pedals. Don't throw them away... try them and see if you like them. That's the beauty of breadboarding!
> If i see a TO 92 package, it is most likely some other transistor that way exceeds the 2n2222a spec, and is relabeled.
"PN2222" (plastic 2222) was a thing, probably legitimately out-sold TO-18 metal 2222, and was a 2222 die inside.
Seeing the plastic part marked "2N2222" defies the JEDEC registration papers. Like you can't get an AKC registration for a Corgi with a tail. You wonder what else they are not telling you.
(https://i.postimg.cc/V579wdy3/tailcorgi.jpg) (https://postimg.cc/V579wdy3)
(Yes, I know you can get AKC papers for a rock, as long as you never enter it in a show or have an unhappy breeding.)
PRR, I don't follow. Are you saying PN2222's are likely different from the 2N2222's I have?
Also can anyone tell me what the A means at the very end of (2N2222A)? What does a B at the end mean? I thought it had something to do with the gain, but I really don't know.
QuoteNow my understanding is that the hfe should be around 100, but I got them today, and tested them with both a built-in multimeter tester and the GeoFX transistor tester, and they all have gains around 250 with little variation. None are below 200.
They are not off spec.
It's a matter of how to interpret the datasheet. In the end the tabulated data was for a different part, 2N2222, not 2N2222A.
The tabulated data on a datasheet can show a combination of minimum, typical and maximum values.
You will see that hFE can vary quite bit with collector current.
Sometimes you will see different specs and different presentation of specs in the datasheets for different brands.
Using your table,
https://postlmg.cc/sBnNHGkf
From the 150mA current, min=100 and max=300. So max is 3 times min (roughly). At a more reasonable 1mA your table has hFE min =50, so you might guess hFE max = 150 and on average hFE typ. about 100.
However, look at the figure 1 on this datasheet, showing typical value
https://web.mit.edu/6.101/www/reference/2N2222A.pdf
Here the gain is about 225. In fact in this datasheet the tabulated data for 1mA shows hFE min = 75 and hFE max = 325, so you could guess a typical of about 200. Either way you getting values of around 200.
So the short answer the 200 you measured sounds fine.
One issue is there is 2N2222A's and 2N2222's maybe your table is for a 2N2222, which IIRC has lower gain.
EDIT:
OK it's confirmed the table is for 2N2222 not 2N2222A
https://media.digikey.com/pdf/Data%20Sheets/ST%20Microelectronics%20PDFS/2N2222.pdf
Used PN2222s in the linked thread. I like this adjustment of the FF. Notice the collector voltages Q2 does not have to be 1/2 power supply voltage.
https://www.diystompboxes.com/smfforum/index.php?topic=112880.msg1044061#msg1044061 (https://www.diystompboxes.com/smfforum/index.php?topic=112880.msg1044061#msg1044061)
Quote from: GuitarMatt on May 30, 2020, 05:29:20 PM
And to anotherjim, I can try to reverse beta, but I have never heard of that before. If I do that on a transistor with gain of 250, what could I reasonably expect the new gain to be? Edit: I just reversed some with gains of about 280 and they came back at 20 or lower. Interesting, but probably too low to be usable.
I'm not a FF expert - but doesn't the circuit effectively multiply the gains of the transistors? 100x100 is same as 20x500 - in principle. Bear in mind that the average Hfe of germaniums is around 70. There's some fuzz designs out there with intentional reverse beta which used to catch me out as I assumed an error in the schematic.
Anyway, you can socket mount transistors - use or butcher an IC socket if you have to - and get building!
Quote from: GuitarMatt on May 30, 2020, 07:03:25 PM
PRR, I don't follow. Are you saying PN2222's are likely different from the 2N2222's I have?
Also can anyone tell me what the A means at the very end of (2N2222A)? What does a B at the end mean? I thought it had something to do with the gain, but I really don't know.
The "A" suffix means they are sorted for gain ranges, A is lowest, B is medium gain, and C would be the highest gain range for that particular transistor. If no suffix, they are unsorted for gain, and can be anywhere within the full range for that transistor.
That is my understanding.
Al
in the 2N/PN types, the A suffix indicates an improved process over the first, non-A spec parts. usually an obscure current or voltage spec. as far as I know, 2N/PN issue a different type number for gain-ranging.
and - the silicon die inside the xx2222A should be the same whether 2N or PN, BUT the 2N should be a metal can, and the PN should be a plastic. those transistors shown at eebay, with 2N2222 etched on, and no other markings are suspect. will almost certainly work as transnsitors, but you have no real idea what the die inside is.
the Pro Electron series transistors, ACxxx, BCxxx et al use the A, B, C suffixes for gain bukketing.
Quote from: GuitarMatt on May 30, 2020, 03:51:20 PM
Assuming I'll need to make another purchase, does anyone have any suggestions on silicon transistors that are generally useful for fuzz-face type builds with gains that are within 70 to 130?
PN2369A have hfe between 40 and 120.
Can't get the attachment to work otherwise there would be a screenshot of the circuit
Simmed the stock Si FF circuit you find on the web with LT spice using the 2N2222 models.
Change the Q1 33K to 10k and the 1K gain control to a 2K or the 1K with a fixed 1K from the bottom to ground. The cap on the control bypasses the fixed 1K all the time
The rest is the 470 and 8.2K in the collector, 100k feedback bias. Cap values and output volume control value to taste
The LT spice 2N2222 model seems to give good results in simple gain stages like a FF, often the circuit voltages match the sim with the 2N2222 model
Simmed at Q1 collector 1.6VDC and and Q2 collector 4.6VDC with the two resistor value changes.
With Fuzzes like this don't get stuck at the 1/2 supply voltage at Q2 like you might read on the web.
> The "A" suffix means they are sorted for gain ranges
That's true of many later EU and Japan parts.
It is not generally true in 2N registration. And the 2N2222A is a modestly improved (higher Vceo) 2N2222.
Selecting for hFE is doomed to be a long hunt.
Quotein the 2N/PN types, the A suffix indicates an improved process over the first, non-A spec parts. usually an obscure current or voltage spec. as far as I know, 2N/PN issue a different type number for gain-ranging.
QuoteIt is not generally true in 2N registration. And the 2N2222A is a modestly improved (higher Vceo) 2N2222.
Yes, for 2N part's it means a new revision, or in fact a new part which is "better" than it's older brothers.
Quote2N/PN issue
In the back of my mind I thought PN meant plastic and cheaper. There was also MPS/MPSA parts. IIRC pin-outs came into the picture as well.
The metal can BC108s became plastic BC548's more or less. There was an era when the plastic parts replaced metal cans.
------------------------------------
EDIT:
Here's an example of the MPS2222/MPS2222A
https://www.farnell.com/datasheets/115091.pdf
For these one's at least the pin-out looks normal for 2Nxxxx
-----------------------------------
For good measure here's the PN2222. The weird thing is this 2018 datasheet from ON Semi doesn't have the PN2222A in the title, and has a lot of stuff stripped out,
https://www.onsemi.com/pub/Collateral/PN2222-D.PDF
This older 2000 datasheet is a combined datasheet and implies the PN2222 is superceded by the PN2222A,
https://users.ece.utexas.edu/~valvano/Datasheets/PN2222-D.pdf
It seems the datasheets are suffering from cut-backs these days ;D
> I thought PN meant plastic and cheaper.
Plastic, lower cost. Not a lower quality!
Military and gear-heads clung to metal packages even after improved epoxy was shown to be better in almost all ways.
Since you've got it on breadboard anyhow, try putting a very small value cap (try 47 or 51 pf) across the collector and base of each of the transistors. It's a great way of smoothing out a silicon fuzz face and generally allows you to get away with higher hfe transistors than you might otherwise.
Found the thread with a one knob fuzz with 2N2222 transistors. I have built a few of this circuit. Parts to the left are a simple guitar sim.
Think colorsound or meathead like fuzz
I used PN2222 transistors and metal mot 2N2222 transistors no real difference it is fuzz it about distortion
https://www.diystompboxes.com/smfforum/index.php?topic=112880.msg1044061#msg1044061 (https://www.diystompboxes.com/smfforum/index.php?topic=112880.msg1044061#msg1044061)
Totally unrelated to the original problem, but I tried to pin down when the plastic packages appeared.
BC108 (TO-18 metal can) and BC548 (TO-92 plastic)
BC548's were released around 1973
History from ETI Electronics magazine in AU:
Ad's appears for BC548's around late 1973
1974 project parts lists specify only BC108
1975 project parts lists specify both BC108 and BC548 as equivalents
1976 projects parts list specifiy only BC548's, BC108's no longer appear
BC108's were being sold off in bulk in 1976 ads.
There's a few exception but that's the general trend.