Looking for a FET...can't find it anywhere!

I'm modifying an Oktava MK-319 and read to use a Toshiba 2SK170BL, and not use anything else. I tried the two places that were recommended (bdent.com and mcmelectronics.com) and they don't have it. Neither does Digi-Key, Small Bear, or Mouser.

Does anyone know where I can find one of these? I even search yahoo with zero luck!

Thanks in advance.

This has been an active topic but not in the Subject line (?); maybe we need a FET Meta.

This part is not supported well by Toshiba America but is widely available in the Far East. One vendor I have used is Chips Technology Circuits in New Jersey. Last I looked they didn't have a website but I will send you the phone number if you like.

There was another source that did have a website but I have forgotten what it was---someone will chime in I am sure.

There is a substitute from a US maker, Linear Technology Systems, the LSK170, and they do have a website. People in here are just beginning to use their parts and the initial press is favorable.

EDIT

PS: if you ask for the 2SK170 first, and then narrow it down to the BL Idss code, you may get better results.

Thanks for the info....what is that companies number?

ACME has them $1.19 a pop.

http://www.acme-sales.net/acmecart/html/main/portal.html

I used them for my MC012 mod and they worked beautifully. The mics are far superior to what hey were.

Shane

The price looked good until it cost 11.99 to ship it!!!! I'm in the U.S.A too!!!!

Why is everyone looking for the 170s? Someone writes something and it gets to the web so you all believe it?

Some of the other (mouser etc carrys)fets do not sound bad at all in microphones, maybe even better than a 170 in a microphone circuit.

The Oktavas I have and worked on, stock fets were OK.

[quote author="Gus"]...The Oktavas I have and worked on, stock fets were OK.[/quote]I know this first hand as well. Tried several 170s in my 319 because of what I had "read". Went back to the stock FET because it sounded better. I concluded: use ears - not eyes! :wink:

170's are superbly low voltage noise but high capacitance---you have to know what you are doing. It will not always be the most suitable part in every circuit.

I'm doing the Scott Dorsey mods to this mic and he specified that out of all FETs used the 170BL is the best for the application.

I was just posting some of what I learned about fets I tried in microphone circuits.

The BL is a mid value IDSS part. Make sure you measure your drain and source voltages with the stock fet first and compare them to the circuit with the 170.

Whew! Ciss and Crss are pretty high on the 170! 30pf for Ciss and 6pf for Crss. I can see why you issued a caution, bcarso. That's a Miller Effect waiting to happen...

What about the 2N4416? I've been fooling around with those in some breadboard apps, and like the sound quite a bit. Ciss and Crss are pretty low, 2pf and .7pf respectively.

En is higher, though--6nv/square root of Hz on the 4416, as opposed to .95 on the 170, so I can see the attraction.

The good 4416's used to be quite good---TI made the best. Since they quit, and others went to ion implantation instead of diffusion, there's been higher voltage noise especially at low frequencies, as well as higher gate leakage.

The 4416A is selected for higher breakdown voltage, although still only 35 Vdg iirc.

Some circuits (such as the bootstrapped cascode of the SE pre I did that you built) will get rid of Miller, and with series feedback get rid of most of Cgs. But, there is some question as to how audible this trickery is ;-).

Some circuits (such as the bootstrapped cascode of the SE pre I did that you built) will get rid of Miller

Click to expand...

I've read from several sources--H&H, Borbely, Pass, and a few others--that cascode and differential circuits help get rid of Miller. I often wonder why the cascode doesn't make it into more designs.

What are some of your favorite FETs, bcarso? I assume from previous posts that you like the 170--those LSK170's look like they have less input C than the 2SK170 does.

> Why is everyone looking for the 170s? ... Some of the other ...fets do not sound bad at all in microphones, maybe even better than a 170 in a microphone circuit.

The 170 is king in Transformerless Dynamic mike-amps. Low-low voltage noise, and who cares about pFd on a 150Ω line?

It would not be right for a capacitor mike capsule. As you say, the smaller (and more common) FETs are your starting point for that kind of work.

PRR I quess I was not the clearest about the fets. I meant a fet used at the capsule end of a condenser microphone.

I have not realy noted a difference in noise(with my ears) with different fets NS process 50 or 2sk170s I have used in microphones.

For people more interested look at a KM84 schematic then think of what is going on with a NS process 50 and 170 fet. If you have a km84 check it with the pad in and out.
Also the miller cap is different with a tube and a solid state the tube has a Vacuum miller cap(s).

[quote author="Gus"]
Also the miller cap is different with a tube and a solid state the tube has a Vacuum miller cap(s).[/quote]

See however PRR's ref from the Radiotron Designer's Handbook 4th Ed. regarding the modulation of the plate-grid capacitance with signal/operating point. It is a small but non-zero effect. As a result, the VHF tubes might be favored, besides their low voltage noise, although the grid current is not so small as some and this may trouble you at low frequencies.

As far as FET's, and just from a noise perspective, in the midband to high frequencies the best practical noise match to a predominately capacitive transducer is achieved for a given FET gate length by matching the capacitance of the FET to the source C. If no matching FET exists, paralleling may get you there, unless the smallest FET is already too big.

> not the clearest about the fets. I meant a fet used at the capsule end of a condenser microphone.

And I didn't really read the original post, which does seem to be about inside-mike work.

Offhand, 40pFd-90pFd seems like a lot for the typical music capsule. In classic voltage-amp condenser mike design, you minimize the amplifier input capacitance to maximize output voltage: 30pFd loading on a 30pFd capsule gives -6dB broadband loss compared to a 0.0pFd load. Of course 0.0pFd implies the input is not connected to anything at all; we need SOME input loading. 10pFd-20pFd is probably typical, and easy with Proc50-size parts, not easy with 170 size parts.

Hi therecordingart,

MCM does in fact show about 1300 in stock.

Search for 2SK170, not 2SK170BL.

These are in fact, BL grade.

I know, because I just bought some to use where J111's would have sufficed. Clients... :roll:

going back to the original question, i'm guessing that one reason mr. dorsey, whose oktava mic mods were published in recording magazine, specified the 2sk170 because it was available pre-sorted for Idss. that way, no selection of FET or resistor values is required by the user. the article was targeted at home recordists (not chronic DIYers as found here).

maybe this came up now because there is a second installment of his mods in this month's issue, this time to the mk219 and 319 mics, for which he also recommends the 2sk170bl?

ed

The Idss bracketing is still pretty wide at a factor of two, 6-12mA, but that's tighter than any other manufacturer I know except the alternate source ones from Linear Integrated Systems.