I guess if it comes right down to it, we could investigate the sm parts. For those of us with large fingers, use smt adaptors. Some have areas for other components as a bonus.
Hi syn08,
bdent would be a more efficient way to spend your dollars. At least you know what you are getting for sure.
-Chris
Hi syn08,
bdent would be a more efficient way to spend your dollars. At least you know what you are getting for sure.
-Chris
"I purchase most of my Japanese parts from www.bdent.com"
I got a nice price on some 2SJ103's there.
I got a nice price on some 2SJ103's there.
> Also, it is hard to assign a match, without a number that you can directly read.
A small gum label on the transistor allows me to identify each transistor.
> Can you draw out what you are doing?
Apply the working voltage (e.g. 12V) using a floating power supply across JFET drain & source, with a source resistor of say 10R (+ve to the drain for an N-JFET, -ve to the source resistor free end). Connect the ground of the functions generator & the scope to the JFET source, channel 1 & functions generator output to the gate, channel 2 to the "Power Supply" end of the source resistor. Apply (in case of N-JFET) a triangular wave of 10Hz between 0V & -0.5V to -2V (depending on JFET type, check datasheet) to the gate. Measure channel 1 & 2 in XY mode, and you should get the Id vs Vgs curve as in your datasheet.
It is not any different to the testing circuit published by Nelson in his MOSFET testing article. Only difference between JFET & MOSFET is Vgs polarity. But that should be obvious.
Patrick
A small gum label on the transistor allows me to identify each transistor.
> Can you draw out what you are doing?
Apply the working voltage (e.g. 12V) using a floating power supply across JFET drain & source, with a source resistor of say 10R (+ve to the drain for an N-JFET, -ve to the source resistor free end). Connect the ground of the functions generator & the scope to the JFET source, channel 1 & functions generator output to the gate, channel 2 to the "Power Supply" end of the source resistor. Apply (in case of N-JFET) a triangular wave of 10Hz between 0V & -0.5V to -2V (depending on JFET type, check datasheet) to the gate. Measure channel 1 & 2 in XY mode, and you should get the Id vs Vgs curve as in your datasheet.
It is not any different to the testing circuit published by Nelson in his MOSFET testing article. Only difference between JFET & MOSFET is Vgs polarity. But that should be obvious.
Patrick
It is good that some people here have their own collection of jfets, etc. I certainly do, collected over the years. When I say that I source parts as part of my living, I mean sourcing devices for Parasound and Vendetta Research. We go all over the world for these parts. However, we need more than 25 or so parts each. We need 100's to 1000's, and most sources are tapped out.
Still, there are the 2SK170, 2SJ74 single parts. These parts are: quiet, hi Gm, and useful for MANY applications. The slightly cheaper 2SJ103 and 2SK246 can be useful also, but they are NOT as quiet, and have low Gm. Getting some of each is a very good idea, because the 2SK246,2SJ103 parts are higher voltage and are PERFECT natural cascodes for the 170, 73 parts respectively.
Anyone who wants to experiment with jfets of this type, should act soon to get a small selection so that they can put something together in the future.
American parts that were designed almost 40 years ago, such as the 2N5564, J174,J271, J113, etc (remember, J in the USA parts, ONLY means plastic package style, not N or P) are poor complementary matches and significantly higher noise. They are useful, but you get more for your investment with the Toshiba parts, even now.
Still, there are the 2SK170, 2SJ74 single parts. These parts are: quiet, hi Gm, and useful for MANY applications. The slightly cheaper 2SJ103 and 2SK246 can be useful also, but they are NOT as quiet, and have low Gm. Getting some of each is a very good idea, because the 2SK246,2SJ103 parts are higher voltage and are PERFECT natural cascodes for the 170, 73 parts respectively.
Anyone who wants to experiment with jfets of this type, should act soon to get a small selection so that they can put something together in the future.
American parts that were designed almost 40 years ago, such as the 2N5564, J174,J271, J113, etc (remember, J in the USA parts, ONLY means plastic package style, not N or P) are poor complementary matches and significantly higher noise. They are useful, but you get more for your investment with the Toshiba parts, even now.
john curl said:
I ordered from bdent.com today and was able to order the 2SK170 V, 2SJ74 V as well as the 2SK246,2SJ103, although they don't specify what IDSS grade on those. I'm hoping to match the mismatched one I purchased from MCM a few years back.
I really appreciate the link. I'd like to keep experimenting for many years and as these things are disappearing, it's time to buy.
As John has pointed out, in between the lines, what was once cheap is no longer and it's only going to get more expensive. So stop sniveling and purchase (or not).
Enjoy, Mike.
MikeBettinger said:
May I suggest a few other gems from these guys?
2SA970/2SC2240 Excellent low noise small signal bipolars in TO92, made in Toshiba. Uce=120V, ICmax=100mA, Cob=3pF, Ft=100MHz, NF=2dB
2SA1407/2SC3601 Bipolars medium power in TO226. Very expensive but well worth for an ultimate performance VAS, made in Sanyo. Uce=200V, ICmax=150mA, Cob=2.5pF Ft=400MHz
2SA1930/2SC5171 Bipolars medium power in isolated TO220. Quite expensive, but excellent for an ultimate performance driver, made in Toshiba. Uce=180V, Ic=2A, Cob=16pF, Ft=200MHz.
I'm in no way associated with this company, it's just that I did a lot of shopping lately.
For those who need 2SK1058/2SK162 laterals, I got my last batch a few weeks ago from www.richtechparts.com They are a chinese shop in Vancouver, BC but they sell original Renesas parts at a very good price. I'm very happy with them. They come from the same tube and are closely matched out of the box.
Mike, I'm glad that you made a fet purchase. I was more than subtle about this for good reason.
The problem last came up with the 2SK389, 2SJ109.
These were VERY common parts for more than a decade. Charles used them, I used them in the CTC Blowtorch, Parasound used them in every preamp and power amp for the last 10+ years. Other individuals used them in quantity as well.
A couple of years ago, Charles Hansen warned me that he was informed by Toshiba that these parts were being discontinued. It seemed impossible, but I duly informed Parsound America to inform Tiawan that they should get the stocks of these devices for any future run in advance. They ignored me.
Last year, the fur hit the fan. Instead of 79 cents for a 2SK389, the price increased to about $10 ea and we still bought them. We then had to increase the price of my new JC-2 preamp because of this, and it created production delays. All because Toshiba, after giving notice, pulled the plug.
Now here, on the Blowtorch thread, an all FET preamp that costs today over $15,000 to purchase new, people are concerned that they may have to pay $1 or more for a part that is similar to the 2SK389, except that you have to match the parts together.
I know from experience that sooner or later the 2SK170 and the 2SJ74 parts will also become difficult to source, but at least there are still some available to amateurs who only need a small quantity to build their own projects.
Like tubes, fets are very special devices. There is little wrong with bipolars, I designed with them full time, until I could find fets that would replace them. For input stages, fets are quiet, require no input coupling cap, and very linear, without even adding resistive degeneration. They bias more simply as well. To ignore this important component, just because of somewhat higher cost, need to match pairs, or storage issues, is short sighted. Just wait until we manufacturers scour the world for the last of these devices. Then the price will be $10 ea, not $1 or less, and the vast majority of you will be priced out of the market.
For example, just last night, a friend called me to ask if he could buy the remaining stocks of my 2SK389. He NEEDS specially selected 2SK389's for a special instrument that he makes to measure ultra low frequency radio waves. The low 1/f characteristic of the 389's is everything to him. He has searched the world already, and knows that there are few left. I now have to chose between, helping Parasound when they get desperate, use them in a Vendetta design, or sell them to my friend. The situation is that serious.
The problem last came up with the 2SK389, 2SJ109.
These were VERY common parts for more than a decade. Charles used them, I used them in the CTC Blowtorch, Parasound used them in every preamp and power amp for the last 10+ years. Other individuals used them in quantity as well.
A couple of years ago, Charles Hansen warned me that he was informed by Toshiba that these parts were being discontinued. It seemed impossible, but I duly informed Parsound America to inform Tiawan that they should get the stocks of these devices for any future run in advance. They ignored me.
Last year, the fur hit the fan. Instead of 79 cents for a 2SK389, the price increased to about $10 ea and we still bought them. We then had to increase the price of my new JC-2 preamp because of this, and it created production delays. All because Toshiba, after giving notice, pulled the plug.
Now here, on the Blowtorch thread, an all FET preamp that costs today over $15,000 to purchase new, people are concerned that they may have to pay $1 or more for a part that is similar to the 2SK389, except that you have to match the parts together.
I know from experience that sooner or later the 2SK170 and the 2SJ74 parts will also become difficult to source, but at least there are still some available to amateurs who only need a small quantity to build their own projects.
Like tubes, fets are very special devices. There is little wrong with bipolars, I designed with them full time, until I could find fets that would replace them. For input stages, fets are quiet, require no input coupling cap, and very linear, without even adding resistive degeneration. They bias more simply as well. To ignore this important component, just because of somewhat higher cost, need to match pairs, or storage issues, is short sighted. Just wait until we manufacturers scour the world for the last of these devices. Then the price will be $10 ea, not $1 or less, and the vast majority of you will be priced out of the market.
For example, just last night, a friend called me to ask if he could buy the remaining stocks of my 2SK389. He NEEDS specially selected 2SK389's for a special instrument that he makes to measure ultra low frequency radio waves. The low 1/f characteristic of the 389's is everything to him. He has searched the world already, and knows that there are few left. I now have to chose between, helping Parasound when they get desperate, use them in a Vendetta design, or sell them to my friend. The situation is that serious.
Hi John,
Actually I'm very pleased to actually be able to find them, buying them is a no brainer.
Years ago I bought a 100 ea of the 2SK389, 2SJ109's. I've built many projects, given them to friends, left them in proto boards, etc. never thinking they would be discontinued any time soon... Now I'm thinking of going through my ex-project collection and reclaiming what I can. It's not like I have a real need relative to your position (between Parasound, Vendetta and a friend). It's just that I heard my first preamp (the original GAS Theadra) that used them and was blown away at the low level detail that was coming through. It wasn't for a few years and many bipolar input stages that I realized it was the FETs (I can be a slow learner) and the cascode that was responsible.
The problem is that there is no point in building an input stage with anything else. Although I have stuck with bipolars for the cascodes, I will finally be trying Borbely's FET cascode on my current project. That is, if the 2sK246/2SJ103's I just ordered are any kind of match and the Vgs is high enough.
I agree with you 100%. Get over the price and stock up now. Otherwise history is bound to repeat itself.
Regards, Mike.
Actually I'm very pleased to actually be able to find them, buying them is a no brainer.
Years ago I bought a 100 ea of the 2SK389, 2SJ109's. I've built many projects, given them to friends, left them in proto boards, etc. never thinking they would be discontinued any time soon... Now I'm thinking of going through my ex-project collection and reclaiming what I can. It's not like I have a real need relative to your position (between Parasound, Vendetta and a friend). It's just that I heard my first preamp (the original GAS Theadra) that used them and was blown away at the low level detail that was coming through. It wasn't for a few years and many bipolar input stages that I realized it was the FETs (I can be a slow learner) and the cascode that was responsible.
The problem is that there is no point in building an input stage with anything else. Although I have stuck with bipolars for the cascodes, I will finally be trying Borbely's FET cascode on my current project. That is, if the 2sK246/2SJ103's I just ordered are any kind of match and the Vgs is high enough.
I agree with you 100%. Get over the price and stock up now. Otherwise history is bound to repeat itself.
Regards, Mike.
syn08 said:
Excellent!
2SA970/2SC2240, I have a good stock of these and agree 100%
2SA1407/2SC3601 I haven't tried these. What current do you suggest running these at?
2SA1930/2SC5171, These I'm using as drivers in my current amps. I'm getting excellent results which I'm sure they are contributing their part.
Thanks for the suggestions, I'll post a few of my current favorites when I get home. That is, if my daughter doesn't have my second grandchild tonight.
Regards, Mike.
- Status
- Not open for further replies.