Possible alt. source for 2SK170

[stoc005]

Has anyone looked at the PF5102 as an alternative for the 2SK170?? It looks good to me and simulates well in place of the 2SK170 in the Symasym Plus 1.3 DTV. It would appear that it is a little easier to get and cheaper than the 2SK170 here is the USA. Mouser and Newark carry it for around 0.12 USD. TO-92 package. Wish it was SMT. Not the same pin order as the 2SK170...be careful.
Below is the LT Spice model for the part if you want to try it.

.model PF5102 NJF(Beta=14.28m Betatce=-.5 Rd=1 Rs=1 Lambda=8m Vto=-.9723
+ Vtotc=-2.5m Is=205.2f Isr=1.988p N=1 Nr=2 Xti=3 Alpha=20.98u
+ Vk=123.7 Cgd=10.76p M=.4069 Pb=1 Fc=.5 Cgs=9.56p Kf=493.1E-18
+ Af=1)

[mlloyd1]

Depends on the application needs and the specs. The 2SK170 is a low noise, high Gm part useful for (duh!) low noise, high gain applications. Most American part number jfets don't have as good a noise spec and do not have as high gm, but are better in other ways that are useful in other applications (for example lower input capacitance for high frequency applications or lower on-resistance for analog switch or variable resistor applications).

That being said, wasn't this jfet used in John Roberts' phono preamp designs many (like 30+) years ago? I am thinking I saw schematics in Popular Electronics and/or Audio magazine in the 70's.

(Man, I am really showing my age with that comment ... )

BTW, many have learned not to put their faith in simulated circuit performance based on commonly internet available SPICE models for jfets.

$0.12 is cheap enough to "buy and try" !

mlloyd1

Edit:
Well whadayaknow, I found the schematics I remember, including comments from the designer himself.
John Roberts' Parametric EQ
Scroll to the bottom of the page ...
And after all that, i see it was the PF5103 instead of the PF5102. Oh Well ...

[zhoufang]

Any good replacement for 2sj74?
I will be the among the thousand of people looking one.

[stoc005]

Possible alternative for 2SJ74 is J270. I've seen more than one person recommend it as a sub for the 2SJ74. Widely available in the US and cheap. The data sheet omits a lot of needed data.... TO-92 package only.

Clearly you didn't dig out the data sheet to see if the PF5102 was close to the 2SK170. Instead of binning the Id(vg=0) using letter suffixes like the 2SK170, the PF510X is a family of parts with different Id(vg=0) for each number. It is a low noise family, which is why it was used 30 years ago in a RIAA preamp. The PF5103 has a higher Id(vg=0) than the PF5102. Other than that, it's the same part.

Simulation is a first step in seeing if the part can be used as a sub, IMHO. I'm sure lots can go wrong once you plug in the sub but at least you know it might be DC stable.

[mlloyd1]

Originally Posted by zhoufang

Any good replacement for 2sj74?
I will be the among the thousand of people looking one.

in a word, no.

high gm, low noise, p-channel jfets might as well be on the extinct species list

at one time, toshiba had a part that used the same die as 2sj74 but in a physically smaller package, so power dissipation was smaller. i'm not sure they even make those anymore.

earlier, one company that makes replacements for the 2sk170 sampled a limited quantity of a replacement for the 2sj74. i believe that company is still debugging their process for this device.

mlloyd1

[zhoufang]

Originally Posted by mlloyd1

at one time, toshiba had a part that used the same die as 2sj74 but in a physically smaller package, so power dissipation was smaller. i'm not sure they even make those anymore.
mlloyd1

That's 2SJ108.
The package is still around, but this part is already EOL when I checked it in the beginning of 2010.

Originally Posted by mlloyd1

earlier, one company that makes replacements for the 2sk170 sampled a limited quantity of a replacement for the 2sj74. i believe that company is still debugging their process for this device.
mlloyd1

That's Linear System.
They have been talking about the P part for years already. Let's hope it will come true one day.

[regiregi22]

Just wondering, why does such a big company stops production and offers no good alternative?
I mean, semiconductor industry tradition is to make better and better parts as time passes and technology evolves. Slower and worse parts as discarded as the better ones arrive.
Why do they discard GOOD components and offer no replacement? That approach goes in the opposite direction of natural finantial evolving.

Regards,
Regi

[jacco vermeulen]

smd/smt, not enough cash in discretes.
Audio is merely a single squashed bug on the electronics window wiper.

[StevenOH]

<<.....PF5102 as an alternative for the 2SK170?? It looks good to me and simulates well in place of the 2SK170.....Possible alternative for 2SJ74 is J270. I've seen more than one person recommend it as a sub for the 2SJ74.....>>

If people will be following the recommendations like these, they will be in a serious trouble (if not already).
YOU need to have the "parts knowledge" to see, what M(akes)S(ense), and what **.
Those "alternatives" do have their place in the "parts universe", but they are no match for the "originals".

As for the SMD equivalents, besides Linear Systems SMD parts, look at these numbers:
2SK431
2SK322
2SK932
And for one step lower (2SK117, 2SK163 types):
2SK209
2SK426
2SK880
2SK2145 - duals
I would not go any lower for real Low Noise apps.