This has me scratching my head??? NKT275's

[RedHouse]

No matter what you try these are positive + ground transistors PERIOD!!!  Dont mean to be blunt...

All of the caps need to be relative to a positive ground.. AGAIN...transistor are (pnp) transistors.....

In other circuits that would be true, and I don't mean to be blunt either ...BUT... in a FuzzFace that's not true.

You can run them either way with just cap re-orientation. The signal path is isolated by input and output caps so the transistors can't tell if "ground" is positive or negative the PNP transistors will run normally.

If you get oscillation which can happen even with the positive ground version 99% of the time all that is needed is a cap across positive-to-ground power rail, ...OR... a 50pF cap across the base of Q1 to the collector of Q2, one or both of these have worked every time for me so far, in 22 Pos-Gnd PNP FF builds.

I've built a (second) test board which I use when I match up transistors and their bias resistors, it uses sockets for the transistors and caps and trimmers so i can dial in a set of transistors before soldering them into a PCB, anyway I have a polarity switch (jumper actually) on it so I can easily change from Pos gnd to Neg gnd... there is little to no difference using AC128, AC125 or Brett's piggy-back'd 2N5087's (which is my favorite and the reason I built this new test board, besides building a lot of FF's)

I'd like to add is that the FF is a sensitive circuit, a bad PCB or strip-board layout will oscillate, and Proto-boards have some high capacitance due to their design which can mislead one into working out a circuit that oscillates later when it's a PCB.

Also testing a FF outside of a shielded enclosure invites all sorts of weirdness, hum induced oscillations, radio reception etc.

Hope this helps

[captntasty]

Just wanted to shout out a thanks for the info on good Ge trannys.  It really seems a good place to look is toward the Land of the Rising Sun for reliable trannys that don't cost 9 bucks per! Again, thanks...

[mac]

Fret,
I've tried a FF with a pair of 2n388 TI, hfe 50 & 60. It sounded excellent, and as it had a litlle less gain than the suggested pair, 70 -120, I had more control over the guitar volume.
Also I breadborded a TBMII with this Ge, hfe 40, 50 & 60, and again it sounded fantastic.
RG said in his article, get the right gain and leakage and the numbers on the case do not matter. I do agree.


mac

[Fret Wire]

Fret,
I've tried a FF with a pair of 2n388 TI, hfe 50 & 60. It sounded excellent, and as it had a litlle less gain than the suggested pair, 70 -120, I had more control over the guitar volume.
Also I breadborded a TBMII with this Ge, hfe 40, 50 & 60, and again it sounded fantastic.
RG said in his article, get the right gain and leakage and the numbers on the case do not matter. I do agree.
mac

I get the same gain and clean up regardless of hfe. If you bias right, your voltages should be the same between different hfe tranny sets regardless of the gain. And you may have to adjust the 470 ohm resistor.

As far as RG's quote, I been saying the same thing in FF threads every time I see brand specific mojo recomendations. Except I add in transistor stability and proper biasing to RG's advice. Of course it's easy to parrot what RG or whoever says, but with Ge's, I've experienced it through a lot of testing. I think the parroting of advice without understanding (or actually trying) leads to some of the incorrect advice you see in so many FF threads. Don't you agree?

For instance, in one thread I mentioned how I set up pairs from 40/60 to over 200hfe in my FF jig, and they all had the same output, fuzz, voltage measurements, tone, and clean-up. And the same people who always parrot "use 70/120hfe and adjust Q2 for 4.5v" kept repeating it without even questioning what I posted. Leads me to believe they've never actually built one. Every fuzz thread has the exact same people (me included) giving their exact same, but different advice. Every time. It's a wonder anyone's fuzz's works.

btw, are you the same Mac who posts on Fuzz Central's forum?   

[mac]

Yestarday I was playing with a breadborded FF NPN using Si & Ge. 2N2219A (110, 117), BD139 (66, 115), 2SD352 (50 & 150uA, 71 & 160uA), 2N388 (50 & 50uA, 60 & 45uA).
The 2N2219A change the fuzz very fast when the guitar volume is near 8 or 9. On the other hand the rest allow me to dial the fuzz easier, in particular the 2N388. Almost noticeable, but real. Maybe the volume pot of my guitar??? About the final gain, not sure, did not measure. Setup is epiphone genesis & yamaha rgx721g both w/DiM and Laney LC30II.
The 2N2219A sound a little brighter and fat, the BD139 have more mids. The 2N388 are very bright, almost like the 2N2219A, but the 2SD352 sound very dark.
The resistors I used for the 2N2219A, BD139 & 2N388 are near 4.7K + 470R. The 2SD352 need 15K + 470R.
I always keep in mind the cutoff freq of the original 8.2K & 0.01uF of about 1950Hz and that the the output is taken 470 omhs before the positive rail, that is, if you picture the rangemaster volume, a 5.5% of what you would have if you'll take the output at Q2's collector (0.47/8.2 + 0.47). In other words I try to keep the same proportions.
So I should have used a 4.7K + 270R & 0.02uF for the Si versions. Maybe this afternoon.

I've been testing FF over the last 2 years, si & ge, mainly because I love 60s stuff and my gear was stolen. I read a lot here, mods, advices, build reports, discussions, some good, some bad. I sometimes walked in circles. Many times I said doh!, I was wrong. I haunted AC and NKT just to find that most good and stable Ge do the fuzz despite their freq characteristics. I've wrote a small program to estimate bias with the advice of some people in this forum. Still I feel I have a lot to learn.
The mac at fuzzcentral forum has been away for a long time. That guy was a newbie 


mac

[Fret Wire]

You just reminded me of something when you mentioned your guitar volume pot. Yours is obviously fine if it's cleaning up between 7 & 10. Alot of the cheap imports come with linear taper pots which you have to turn down alot before you hear any clean up. I wonder how many fuzz builds where the poster complained of poor clean up was due to them not knowing they had a lin volume pot in the guitar?

I end up with 4.7 and 220r alot when I go silicon. Do you play with the Mayer Mods much? 2k fuzz pot and 1M output resistor.

[mac]

I tried the 2k pot or what is closer, I  used the exact 33k, 8.2k and 470R, and I set the bias with a trimmer in place of the 1k pot, and adjusted the 22uF. Some extra gain and a little thinner. I prefer leaving the 1k pot or fixed resistor and adjust the collector resistor. And I like a +500k vol pot. The 100k sound to bright to me. To be honest, I like the original FF.


mac

[dubs]

Thanks analogman...  That's what they look like.  When I got them I sort of suspected something funny.  But as they say "Denial ain't just a river in Egypt"!  I'd never seen NKT's in the tophat style case.  I figured I'd throw caution to the wind once.  Live and learn.  The supplier I got them from is Vintageparts.com out of UK (or so I was led to believe).  I'm a bit embarrased to be snookered but knew I was taking a risk - thankfully a fairly benign one.  Anyhow, thanks for the confirmation - if anybody knows a thing or two about NKT's it would be you!  Should I post a warning about this supplier?  I also got a few AC128's (which appear authentic) - most of 'em dead but a couple with low gains - 40 & 60.  I suppose the bunch was actually planted very carefully.

I know guys like fulltone use the digicorp curent production nkt275
I recently got some 70's nkt275 from vintageparts too and I think they are the real deal?? . These do have the top hat but a notch indicating the collector (?).  the writing is different to the fakes. They have a grey base not black like the fakes...
I guess the only way to really find out is to hear it..

[analogmike]

Those are not NKT transistors, someone else relabeled them. But they might sound good, have fun and good luck!

[R.G.]

The only real NKT275's I ever got did not look like that. But perhaps there were multiple makers of NKT275 back in the day.

There is no real way to tell, and after all, the type number doesn't mean much. Put them in a circuit and listen. Your ears will know.

[vanessa]

R.G.- I agree with you. I think there were multiple makers of the NKT-275 for New Market (if that was the company). A friend of mine pulled a few from an old transistor radio a while back that looked different than what Mike has (Analogmike) and what Fulltone shows on his "fakes" page. They were can types, but with the logo horizontal and I believe they had a black dot (maybe blue?). This radio had not been opened since the 60's so I would find it hard to believe why someone would re-stamp them with a different spec value just to claim they were NKT-275.

[dubs]

I think where I got them (vintageparts) are a pretty legit source as a I got some genuine mullards oc44 too. Unless they've been duped.
These nkt275 were from the 70's. they have 90's one s which they said was different.
From the various responses everyone has made, how does anyone truly know what is real and what is not?
If there are several manufacturers at the time, some of the others may be perfectly legit?
Have you got pics of them vanessa??

[R.G.]

Here's the real poop.

An NKT275, or any transistor for that matter, does not so much exist as a real device as it exists as an abstraction, a set of limits on a datasheet.

A datasheet says certain things: it's in this or that package, the pinout is thus-and-such, the gain is this min/max, it will withstand at least this much voltage, current and power, and so on.

It is perfectly legitimate for anyone to make any chunk of semiconductor, test it to those specs, and determine to call it whatever is on the datasheet. It is possible that some type numbers could be registered as a trademark, and in that case no one else could legally use the trademark without permission, but that's a trademark issue, not a semiconductor performance issue.

ANY TRANSISTOR MEETING THE SPECIFICATIONS ON THE DATASHEET COULD LEGITIMATELY BE CALLED ONE OF THAT PART NUMBER. It's certain that all of the 2N3904's on the market are not identical inside. In fact it strains logic that they should be.

If it meets the datasheet specs, it is one of them. If it meets the specs of two or more datasheets, it could be called one of the others as well, and perfectly legitimately.

What's wrong here?

What's wrong is that it is a mistake to think that the label "NKT275" means something beyond what is on the datasheet. It does not. In particular, it does not mean it will sound good even if it meets every single spec. Whatever makes a transistor sound good is NOT on the datasheet.

I keep telling people that the type number doesn't matter. The only reason to buy a specific type number is that you think it gives you a chance at a good sound. It is by no means a guarantee.

[vanessa]

I agree with you again R.G.!

I believe what you are looking at as being a great sounding transistor just happens to have a tag on it that says NKT-275. That same run of transistors could have been also made for someone else and labeled 2NXXX for purchase by that company.

[dubs]

Thanks for that R.G!
I guess it's like a Fender strat  - Fender are the original makers of strats but they are many versions  of the strat made by other people - some better, some worse!

[analogmike]

I am a collector of vintage pedals so I look at things differently. There is only one type of NKT275 I can use when restoring a vintage fuzzface, the rest won't be correct. There was only one Newmarket company and all of their transistors looked the same. Not sure if it's legal for other people to put NKT on a transistor just because it meets the specs (NKT is a brand name, not just a 2N part#). Just like there is only one correct PAF to put in a '59 sunburst (long magnet, nickle cover, preferably white bobbins, correct color magnet wire) there is only one correct transistor for a '67 fuzzface. Anything else is incorrect and the value of the pedal will be about half of an original pedal.

If you are just building a pedal for your own use, get the matched germaniums from smallbear and you will get a better sound than any of the "AC128" or "NKT275" you will find on ebay, I guarantee.

here are some of my vintage fuzz pedals



Have fun!

[col]

I take it you are UK with the GBP reference. Try this supplier from the NW:

http://hometown.aol.co.uk/oldradioparts/semiconductors.htm

he used to do matched pairs for fuzzfaces and even a fuzzface kit. I have had quite a few items off him and all have been exactly as stated and the matched pair I got off him sounded great! He doesn't list them anymore but he might still do them if you get in touch. He might also be able to suggest something from his range closer to what you want.

Col

[captntasty]

Helluva collection Mike!  Couldn't help but notice Townshend's ATBCHCE...  might very well be my favorite album of all time.   

To update on the NKT's I got from VintageParts.com - they sound perfectly fine, but no better than the sets you can get from SmallBear.
There is no "mojo" in them.  Different tranny sets I've tried ("vintage" type to more recent Ge's) have a different character.  The most noticeable difference of sound I've found is from the NOS 128's I got from VP.com - they're... more crisp?  More highs... unfortunately their gains are low (40 & 60) and the output doesn't equal the clean output of the amp without the FF engaged.  Perhaps there is a way to get more output from a FF?  I really liked the sound but the overall output of the effect is too little.

So in conclusion... well, is there ever really a conclusion to this story    

[R.G.]

Perhaps there is a way to get more output from a FF?  I really liked the sound but the overall output of the effect is too little.

The collector resistor of Q2 is made up of two resistors, a large and a small one. This constitutes a voltage divider that reduces the signal at Q2's collector by quite a bit. If you change the ratio of those two resistors to make the one at -9V larger and the one at the collector smaller, it will increase the output of the FF. It's like a fixed volume control before the output variable volume control.

[vanessa]

There was only one Newmarket company and all of their transistors looked the same. Not sure if it's legal for other people to put NKT on a transistor just because it meets the specs (NKT is a brand name, not just a 2N part#). Just like there is only one correct PAF to put in a '59 sunburst (long magnet, nickle cover, preferably white bobbins, correct color magnet wire) there is only one correct transistor for a '67 fuzzface. Anything else is incorrect and the value of the pedal will be about half of an original pedal.

That does not make much sense. My friends transistor radio (circa 60's) had NKT-275 transistors that were marked different than yours. Why would a company back in the 60's mark their transistors NKT-275 when no one at the time could care less about NKT-275's (except maybe Newmarket)? I read a post a while ago were someone came across these very same marked versions of the NKT-275. Who's to say that they are any different than those found in a '67 fuzzface?