No, I haven't been eaten by the dreaded Gonzo Fish, just overwhelmed by things that should have been done yesterday (or, in many cases, three or four weeks ago).
I will get another variation on the New-Tron posted as soon as I can, but perhaps I shouldn't make promises as to exactly what day. I still haven't managed to get my notes transcribed into a formal schematic. That alone will take an hour or two in order to make it pretty. I haven't managed that much time in a couple of weeks.
In any event, I'm still around, just trading stocks, playing farmer, writing a new story, not to mention getting the Lovoltech JFETs out and a few other odds and ends. It would be nice to sleep, but so far I seem to be getting by on five or six hours or so per night, though there will be hell to pay when I finally crash.
Grey
I will get another variation on the New-Tron posted as soon as I can, but perhaps I shouldn't make promises as to exactly what day. I still haven't managed to get my notes transcribed into a formal schematic. That alone will take an hour or two in order to make it pretty. I haven't managed that much time in a couple of weeks.
In any event, I'm still around, just trading stocks, playing farmer, writing a new story, not to mention getting the Lovoltech JFETs out and a few other odds and ends. It would be nice to sleep, but so far I seem to be getting by on five or six hours or so per night, though there will be hell to pay when I finally crash.
Grey
Muppets From Space
Based on a song from The Muppet Movie where Gonzo sings about flying through space to a place he's only dreamed about, Muppets From Space follows Gonzo's quest to find out where he's from. Early on in the movie, Gonzo spills his Kap 'N' Alphabet cereal and it mysteriously spells "Watch the Sky" and "RU There." That night he climbs on the roof of the house that he shares with the rest of the Muppets and watches the sky. "Lightning" strikes him and he's transported into space where the cosmic fish validate his belief that he's an alien. Everybody, including his best friend, Rizzo the Rat, thinks he's gone a little crazy. But Gonzo continues to follow his beliefs and appears on TV with Miss Piggy to let the world know he's an alien. This sets off an alien craze, and Dr. K. Edgar Singer, a scientist working for the government investigating alien sightings, abducts Gonzo. Finally at the end of the movie, when Gonzo's family comes in a spaceship back to Earth, he must decide whether to go into space with them or remain on Earth with his friends.
Based on a song from The Muppet Movie where Gonzo sings about flying through space to a place he's only dreamed about, Muppets From Space follows Gonzo's quest to find out where he's from. Early on in the movie, Gonzo spills his Kap 'N' Alphabet cereal and it mysteriously spells "Watch the Sky" and "RU There." That night he climbs on the roof of the house that he shares with the rest of the Muppets and watches the sky. "Lightning" strikes him and he's transported into space where the cosmic fish validate his belief that he's an alien. Everybody, including his best friend, Rizzo the Rat, thinks he's gone a little crazy. But Gonzo continues to follow his beliefs and appears on TV with Miss Piggy to let the world know he's an alien. This sets off an alien craze, and Dr. K. Edgar Singer, a scientist working for the government investigating alien sightings, abducts Gonzo. Finally at the end of the movie, when Gonzo's family comes in a spaceship back to Earth, he must decide whether to go into space with them or remain on Earth with his friends.
Ohh,,, O.K.,,, I guess I did not really want to know... Thanks anyway... It's father's day here but it,s not... the kids are probaly to old for that and the grandkids probably missed that...
Trouble is, I'm being de#$%^&* and all the kids seem to be going with it... I'm just having trouble to see it as more DIY time for me or somethin😕 😕 😕
But,😀 😀 😀 Happy Father's day to all, and of coarse our favorite "PaPa" of modern Class A DIY. He must be one H#$% of a Dad to someone...
Trouble is, I'm being de#$%^&* and all the kids seem to be going with it... I'm just having trouble to see it as more DIY time for me or somethin😕 😕 😕
But,😀 😀 😀 Happy Father's day to all, and of coarse our favorite "PaPa" of modern Class A DIY. He must be one H#$% of a Dad to someone...
Okay, let me dust off my keyboard and see if I can catch up.
--Availability of the 2SK389 in Europe must be vastly better than here in the US. I am assuming the price is better, as well. They are far from plentiful here and the buyer has no control over the grade (GR, BL, VI). The situation is intolerable. I'm glad that folks overseas have access to these parts, but here in the US, it's nearly a non-starter.
--While on the subject of Toshiba parts, I have been unable to find the 2SK2013/2SJ313 at any price.
--And a little birdy tells me that Linear Systems is coming out with P-ch devices. Remember, you heard it here first. I am still measuring interest in a buy for the Linear Systems JFETs. If you live in Europe and feel that you've got plenty of Toshiba parts, more power to you. We here in the US don't have that luxury.
--On the subject of the resistors to ground at the output: The DC offset of the New-Tron is virtually zip. Much less of a problem than the Aleph-X. Most any value will work in that regard. As for lower values and the return of Gate current, remember that these are JFETs and have a very high Zin. I measured the current at one point, just as a matter of academic interest, and found it to be negligible. I didn't write down the figure, so I'll have to go back and recreate the measurement from scratch if anyone thinks it's important. That might take a day or two or three, given the way my life is going. Given that it's been forever since I've gotten into this thread, consider yourselves warned.
--If you want to use the New-Tron, Isotope One for low impedance speakers, parallel outputs or increase the bias. Don't try to run it class AB.
--I had no idea that there was such a thing as a Gonzo Fish. I made that up on the spur of the moment. I haven't watched television but twice in the last thirty years. Neither time involved Muppets.
And...
There has been talk about using a current source and a resistor to set the bias for the output stage. I'm including a schematic of the New-Tron, Isotope Two, which shows exactly such an animal. It doesn't really matter whether you use a fixed CCS and and a variable resistor or vice versa. This one uses a variable CCS and fixed resistor. This is not a "build and forget" kind of circuit. You'll have to set the bias yourself. Like the Zen #9, set the bias to minimum and bring it up slowly. You'll want to bring both sides up together.
I don't remember how many biasing schemes I've got--this is just one of many. And I have non-Circlotron circuits as well. What I don't have is time, but that's an old story.
Incidentally, this is a Class A circuit, too.
Grey
P.S.: Again, no promises as to how long before I get back in here. I'll do the best I can.
--Availability of the 2SK389 in Europe must be vastly better than here in the US. I am assuming the price is better, as well. They are far from plentiful here and the buyer has no control over the grade (GR, BL, VI). The situation is intolerable. I'm glad that folks overseas have access to these parts, but here in the US, it's nearly a non-starter.
--While on the subject of Toshiba parts, I have been unable to find the 2SK2013/2SJ313 at any price.
--And a little birdy tells me that Linear Systems is coming out with P-ch devices. Remember, you heard it here first. I am still measuring interest in a buy for the Linear Systems JFETs. If you live in Europe and feel that you've got plenty of Toshiba parts, more power to you. We here in the US don't have that luxury.
--On the subject of the resistors to ground at the output: The DC offset of the New-Tron is virtually zip. Much less of a problem than the Aleph-X. Most any value will work in that regard. As for lower values and the return of Gate current, remember that these are JFETs and have a very high Zin. I measured the current at one point, just as a matter of academic interest, and found it to be negligible. I didn't write down the figure, so I'll have to go back and recreate the measurement from scratch if anyone thinks it's important. That might take a day or two or three, given the way my life is going. Given that it's been forever since I've gotten into this thread, consider yourselves warned.
--If you want to use the New-Tron, Isotope One for low impedance speakers, parallel outputs or increase the bias. Don't try to run it class AB.
--I had no idea that there was such a thing as a Gonzo Fish. I made that up on the spur of the moment. I haven't watched television but twice in the last thirty years. Neither time involved Muppets.
And...
There has been talk about using a current source and a resistor to set the bias for the output stage. I'm including a schematic of the New-Tron, Isotope Two, which shows exactly such an animal. It doesn't really matter whether you use a fixed CCS and and a variable resistor or vice versa. This one uses a variable CCS and fixed resistor. This is not a "build and forget" kind of circuit. You'll have to set the bias yourself. Like the Zen #9, set the bias to minimum and bring it up slowly. You'll want to bring both sides up together.
I don't remember how many biasing schemes I've got--this is just one of many. And I have non-Circlotron circuits as well. What I don't have is time, but that's an old story.
Incidentally, this is a Class A circuit, too.
Grey
P.S.: Again, no promises as to how long before I get back in here. I'll do the best I can.
Attachments
GRollins said:--And a little birdy tells me that Linear Systems is coming out with P-ch devices. Remember, you heard it here first. I am still measuring interest in a buy for the Linear Systems JFETs. If you live in Europe and feel that you've got plenty of Toshiba parts, more power to you. We here in the US don't have that luxury.
As of 28/12/05, at least, they did not have any plans to produce such a device.
Actually, I asked Paul Norton about what the cost would be for a complement to the LS843 (very nice dual monolithic N-JFET with 1mVgs matching), or to the LSK389. Needless to say, their prices were rather prohibitive (minimum orders on the scale of hundreds of thousands of dollars).
However, it was (IIRC) indicated that they would eventually produce a complement to the LSK170. Presumably, that will also be available in similar bins.
On a side note, I have found other (smaller) foundries that are willing to produce custom transistors at reasonable prices (a few thousand euros per wafer), and am planning to source some custom complementary output transistors this way.
I guess one could theoretically use one of these to source custom complementary small-signal dual-JFETs designed for audio use as well, but I'm thinking it would be harder to organize a group buy, since people might be skeptical about the sonic properties of a new transistor up front.
If it should happen to be of any interest to you, you could e-mail me.
If you are prepared to go that far to get custom transistors made, that I should guess that power JFETS (like the obsoleted Sony 2SK60) would be much more interesting. You would not imagine how many people are prepared to pay up to US$20 for one of those !!!!
I also did not believe how much more linear they are than MOSFETS, until I measured one.
But maybe I am not the only one with such an idea. Who knows; maybe we'll see a Pass Lab Power JFET amplifier in a year with Semelab custom made power JFETS ???
Patrick
I also did not believe how much more linear they are than MOSFETS, until I measured one.
But maybe I am not the only one with such an idea. Who knows; maybe we'll see a Pass Lab Power JFET amplifier in a year with Semelab custom made power JFETS ???
Patrick
It would seem easier to just GB some european parts (small signal)??? But, as long as we are talking custom dual devices. We might be able to convince someone ??? to bond adjacent die from the same wafer, of their existing product, into a dual package they already offer??? The NRE would be much less.
Would that produce sufficient matching.😕 😕 😕
Would that produce sufficient matching.😕 😕 😕
Linear Systems will be releasing their version of the 2SJ74 "this year." The information I have is somewhat sketchy, but I believe the '109 will be released next year.
Incidentally, the '74 I used in the circuit above was a GR--which is all I can get. If you use a higher current device, you might need to consider increasing the value of either the current adjusting pot or the fixed resistor.
Other P-ch JEFTs--from Motorola/OnSemi, for instance--will do just as well. Just allow for some fiddle factor with the resistors. You'll want something on the order of 7V at the Gate of the cascode MOSFET, plus or minus a volt or two to allow for parts variations.
Other current source designs will also work. I used the 2SJ74 because it was handy and was fairly compact voltage-wise. If you start using some of these really, really complicated current sources from other threads with stacks of LEDs and multiple transistors, you'll find that the current source will need several volts across it in order to operate. The best solution in that case will be to use a separate rail for the CCS. That way the current source won't suffer when the output approaches the rail and the CCS runs out of voltage.
Grey
Incidentally, the '74 I used in the circuit above was a GR--which is all I can get. If you use a higher current device, you might need to consider increasing the value of either the current adjusting pot or the fixed resistor.
Other P-ch JEFTs--from Motorola/OnSemi, for instance--will do just as well. Just allow for some fiddle factor with the resistors. You'll want something on the order of 7V at the Gate of the cascode MOSFET, plus or minus a volt or two to allow for parts variations.
Other current source designs will also work. I used the 2SJ74 because it was handy and was fairly compact voltage-wise. If you start using some of these really, really complicated current sources from other threads with stacks of LEDs and multiple transistors, you'll find that the current source will need several volts across it in order to operate. The best solution in that case will be to use a separate rail for the CCS. That way the current source won't suffer when the output approaches the rail and the CCS runs out of voltage.
Grey
Oops. I believe I said thousands of dollars for a single wafer. Obviously, the wafer itself is less than a tenth of that price. The real cost is the development cost and the mask production, the latter being what costs a few thousand dollars.
Actually, those are readily available at about GBP25, along with their complementaries. But, yes, I am trying to find someome who has the production facilities and expertise to design a linear, complementary power JFET. Many companies fit the bill, but only a handful are willing to deal with small-quantity orders. If the IP rights for the 2SK60 has expired, I'm guessing you could probably produce it rather cheaply.
For Nelson and co, that might not be such a bad idea, as they could probably afford to have them experiment with trying to realize a Silicon-Carbide power JFET, which could operate at significantly higher temperatures.
I'm actually from Europe. 🙂
Bonding adjacent die is not a problem, nor is bin-sorting and bonding, and you can reasonably expect similar matching as for the Toshiba devies.
However, this does not give the same level of performance. What you actually need is a single-die dual- (2xN/2xP) or quad (2xN+2xP) device. This is because the bond pads take up a lot of space, meaning that you'll end up with poor thermal coupling and hence poor thermal matching.
The ideal solution for a low-NRE design would probably to select the best device they have for the task (2N6550, IF9030, U430, 2SK112, 2SK147, etc. make sense) and make chips of 2x2 devices, internally parallelled, and stick them in a DIL6 package.
Yet the real problem is that you'll probably want complementary devices, and that's going to require a bit of R&D. If you find a basic process that has a large margin to the performance you need, I think you could get a reasonable approximation by making the N-polarity channel three times as long (to compensate for the difference in apparent channel length due to electron/hole mobility differences) and fiddling with the dopant concentrations.
For audio applications, it's not entirely unlikely that a simple approximation would yield adequate results, though.
EUVL said:If you are prepared to go that far to get custom transistors made, that I should guess that power JFETS (like the obsoleted Sony 2SK60) would be much more interesting. You would not imagine how many people are prepared to pay up to US$20 for one of those !!!!
I also did not believe how much more linear they are than MOSFETS, until I measured one.
Actually, those are readily available at about GBP25, along with their complementaries. But, yes, I am trying to find someome who has the production facilities and expertise to design a linear, complementary power JFET. Many companies fit the bill, but only a handful are willing to deal with small-quantity orders. If the IP rights for the 2SK60 has expired, I'm guessing you could probably produce it rather cheaply.
But maybe I am not the only one with such an idea. Who knows; maybe we'll see a Pass Lab Power JFET amplifier in a year with Semelab custom made power JFETS ???
For Nelson and co, that might not be such a bad idea, as they could probably afford to have them experiment with trying to realize a Silicon-Carbide power JFET, which could operate at significantly higher temperatures.
flg said:It would seem easier to just GB some european parts (small signal)??? But, as long as we are talking custom dual devices. We might be able to convince someone ??? to bond adjacent die from the same wafer, of their existing product, into a dual package they already offer??? The NRE would be much less.
Would that produce sufficient matching.😕 😕 😕
I'm actually from Europe. 🙂
Bonding adjacent die is not a problem, nor is bin-sorting and bonding, and you can reasonably expect similar matching as for the Toshiba devies.
However, this does not give the same level of performance. What you actually need is a single-die dual- (2xN/2xP) or quad (2xN+2xP) device. This is because the bond pads take up a lot of space, meaning that you'll end up with poor thermal coupling and hence poor thermal matching.
The ideal solution for a low-NRE design would probably to select the best device they have for the task (2N6550, IF9030, U430, 2SK112, 2SK147, etc. make sense) and make chips of 2x2 devices, internally parallelled, and stick them in a DIL6 package.
Yet the real problem is that you'll probably want complementary devices, and that's going to require a bit of R&D. If you find a basic process that has a large margin to the performance you need, I think you could get a reasonable approximation by making the N-polarity channel three times as long (to compensate for the difference in apparent channel length due to electron/hole mobility differences) and fiddling with the dopant concentrations.
For audio applications, it's not entirely unlikely that a simple approximation would yield adequate results, though.
GRollins said:Linear Systems will be releasing their version of the 2SJ74 "this year." The information I have is somewhat sketchy, but I believe the '109 will be released next year.
You could give them a call at 510-490-9160 if you're looking to buy. I don't think it could hurt to let them know that there is a market for it at least 😉
Actually, given the number of people on these forums that would like a set of these, you might be able to get a large enough group buy going that you could get them to make DIL12 chips equivalent to a 2SK389 and a 2SJ109 in a single package (perhaps even on a single wafer), as this would presumably have a lower associated NRE cost than what I contacted them about.
Other current source designs will also work.
I typically use balanced complementary circuits for which the 2N5911, 2SK389 and LS843 dual N-JFETs tend to work out nicely. I agree that complicated current sources tend to eat up a lot of voltage for little benefit, compared to a cascoded JFET source or an InterFET JFET tetrode.
GRollins said:--Availability of the 2SK389 in Europe must be vastly better than here in the US. I am assuming the price is better, as well. They are far from plentiful here and the buyer has no control over the grade (GR, BL, VI). The situation is intolerable. I'm glad that folks overseas have access to these parts, but here in the US, it's nearly a non-starter.
I am planning to order some JFETS at Borbely shortely. If people are interested I can order some extra and send them through.
2SK389
Dr Strangelove: "I am planning to order some JFETS at Borbely shortely. If people are interested I can order some extra and send them through."
I would be interested. Have an idea of prices?
Dr Strangelove: "I am planning to order some JFETS at Borbely shortely. If people are interested I can order some extra and send them through."
I would be interested. Have an idea of prices?
Hi pimarzu,
You can take a look at the prices here.
Borbely is not the cheapest, Borbely wil add 16% VAT and customs to my place will probably add another percentage. So if you combine all this plus shipment etc. You get a pretty good idea of the total price. Indeed they don't come cheap...
But he has got the hard to get VI grade JFET's.
If more people are interested, please send me an email.
You can take a look at the prices here.
Borbely is not the cheapest, Borbely wil add 16% VAT and customs to my place will probably add another percentage. So if you combine all this plus shipment etc. You get a pretty good idea of the total price. Indeed they don't come cheap...
But he has got the hard to get VI grade JFET's.
If more people are interested, please send me an email.
Borbelys JFETS
Hi Strangelove, thanks.
I followed the link, there is a lot of interesting stuff.
For which project will you use these grade VI JFETs?
Hi Strangelove, thanks.
I followed the link, there is a lot of interesting stuff.
For which project will you use these grade VI JFETs?
Don't know yet. I just want some to experiment and I want some to have in stock for future projects. Perhaps an Aleph-J or someting like this threads Tron.
Borbely aint got the best prices. There are better sources of the Toshiba devices indeed. We are blessed I think here in the EU regarding the J-fets. Most of them are easy to get, and a few of them are harder to get, but obtainable🙂 I know, 'cause I went through all that, and now I have 40+ of all the rare gem's😀
In my oppinion if you were to make a GB, the best thing would be to let the source/shop handle it?. Thanks Patrick😎
Steen🙂
In my oppinion if you were to make a GB, the best thing would be to let the source/shop handle it?. Thanks Patrick😎
Steen🙂
Hi Steen,
Indeed Borbely hasn't got the best prices. But he has the hard to get V grades.
I don't want to make this into a real GB. I'm way too disorganized and buisy with all my things here to handle a real GB. I must admit the thought appeals to me, but for now it is not very wise.
I just want to place an order and hope to a few with me. I just need a few parts, and Borbely has high minimum order.
If you know another source with the V grades from which it is easy to order, please let me know.
Indeed Borbely hasn't got the best prices. But he has the hard to get V grades.
I don't want to make this into a real GB. I'm way too disorganized and buisy with all my things here to handle a real GB. I must admit the thought appeals to me, but for now it is not very wise.
I just want to place an order and hope to a few with me. I just need a few parts, and Borbely has high minimum order.
If you know another source with the V grades from which it is easy to order, please let me know.
- Status
- Not open for further replies.
- Home
- Amplifiers
- Pass Labs
- The New-Tron