The thing that keeps slipping through the cracks is that the 2SK389 and 2SJ109 are out of production. There won't be any more. The price for what's out there will only climb. That's why I'm interested in the Linear Systems parts. True, the footprint is different, but I don't care as long as they behave nicely in a circuit. I make new boards for new projects and can make a board for the LSK as easily--or perhaps more so--as for the 2SK/J parts. You're not dealing with that silly offset for the front leads.
As far as I know, there are still 2SK170 and 2SJ74 parts available. However, I would not be at all surprised to find that they, too, are out of production and that Toshiba is simply working through their stock.
Grey
As far as I know, there are still 2SK170 and 2SJ74 parts available. However, I would not be at all surprised to find that they, too, are out of production and that Toshiba is simply working through their stock.
Grey
Hi Grey,
For what it's worth, I just found this advice about Linear Systems on another forum...
http://www.diyhifi.org/forums/viewtopic.php?p=15297#15297
Should we trust it
?
Cheers
For what it's worth, I just found this advice about Linear Systems on another forum...
http://www.diyhifi.org/forums/viewtopic.php?p=15297#15297
Should we trust it
?Cheers
CheffDeGaar said:Hi Grey,
For what it's worth, I just found this advice about Linear Systems on another forum...
http://www.diyhifi.org/forums/viewtopic.php?p=15297#15297
Should we trust it
Cheers
I haven't compared the LSK389C's with the 2SK389V's, so I couldn't say anything about their performance, but unless Mr. Curl would like to step in to correct me, I'm going to assume (based on the various JC schematics floating around) that his problem is with the absence of a complementary part, rather than the performance of the part itself.
Linear Systems has some pretty impressive production facilities, and I suspect any perceived "inferiority" of the LSK389 sonics can be attributed to listener bias. The 2SK389's and 2SJ109's have a nearly mythical status among the DIY community, and this can not possibly contribute constructively to an objective assessment of their sonics, in my humble opinion.
If people are interested in hearing about the results, I'll try requesting a sample of the LSK389C's (or make a small-quantity buy) and comparing them to the 2SK389V's in a simple headphone amp circuit to evaluate them in a double-blind ABX test.
This, of course, depends on finding a source for the 2SK389V's, though, plus I'm right in the middle of moving house, so don't hold your breath unless you'd like to help out
Input as to the ideal circuit for testing would be appreciated; otherwise, I'll go with testing the simplest circuit possible (a volume control and a buffer) first.
GRollins said:
Since the Linear Systems parts keep coming up, I guess I'll post some replies from Paul Norton from later in the thread on that forum, in case someone missed out.
First off, with regards to availability, minimum orders (or, rather the lack of such), and the expected arrival of the complementary parts:
So Greg's source seems to check out.
With regards to samples and availability, specifically the issue raised by the post linked to by CheffDeGaar, slightly abbreviated:
And, finally, by another member of that forum, with regards to the sonics, again abbreviated:
The main complaint seems to be the same thing that I experienced when trying to get a quote from their distributor in Denmark: they are so intent on dealing with big companies that they ignore requests from individuals or small companies.
However, since Mr. Norton has posted himself at that forum, one may hope that he will read and implement the suggestion about setting up an online order system for dealing with small quantities, or do something like that via MCM or something.
Baring that, one apparently needs to contact them directly for these small quantities. Somewhat amusing, as I thought a major point of having a distributor network was to offload the production company.
My experience with the Linear Systems parts has been quite satisfactory. I agree that there might be just a little bit of listener bias involved. I regard the pinout change as a nuisance at most; in ordinary usage in my designs, I whomp up new artwork for each new design, so it's no skin off my nose to do the LSK part.
Good luck finding high current '389s...they were the first to go.
John Curl does indeed make unpleasant noises about the Linear Systems parts, but does not supply any specifics. Unless and until I have more particular information as to how the Linear Systems parts are inferior, I'm going to go with my experience, which has been positive so far.
I think it's likely that the advent of the P-ch parts will alleviate many peoples' objections.
Grey
P.S.: So far it has been a jinx to say that I'm going to get a circuit posted in this thread. Every attempt to do so has resulted in an avalanche of fresh problems here that have to be attended to Right Now, with an emphasis on Immediately.
With that in mind, I am NOT going to point out that I have this coming week off. Nor am I going to say that I've got a few parts coming in that will be of use for these amps. And I am most certainly not about to say that I will post a new circuit at any point in the near future.
(ahem...)
Good luck finding high current '389s...they were the first to go.
John Curl does indeed make unpleasant noises about the Linear Systems parts, but does not supply any specifics. Unless and until I have more particular information as to how the Linear Systems parts are inferior, I'm going to go with my experience, which has been positive so far.
I think it's likely that the advent of the P-ch parts will alleviate many peoples' objections.
Grey
P.S.: So far it has been a jinx to say that I'm going to get a circuit posted in this thread. Every attempt to do so has resulted in an avalanche of fresh problems here that have to be attended to Right Now, with an emphasis on Immediately.
With that in mind, I am NOT going to point out that I have this coming week off. Nor am I going to say that I've got a few parts coming in that will be of use for these amps. And I am most certainly not about to say that I will post a new circuit at any point in the near future.
(ahem...)
GRollins said:
Thanks for that feedback. It does seem that most of the criticisms out there are somewhat lacking in detail/substance.
Actually, I always found the Toshiba devices' pin layout to be rather annoying, at least when prototyping. With my fine motor damage, I can pretty much forget about doing point-to-point work with them.
The DIL and TO-72 packages are a different matter, though.
Yeah. They're still available some places, but at rather excessive prices, and the ones I've used lately came from mixed lots.
Makes sense. The JC input stages seem to pretty universally use complementary parts, and so the LSKs will not be all that useful to him until their complements are available.
Indeed. Now all we need is or them to announce the availability of dual complementaries in a single package. That'd make for a lovely buffer.
Hi suiraMB,
I just sent an email to Paul Norton to find out what happened to my sample requests. I have heard nothing since August, 01 of this year.
I must thank you for the link you provided. It was helpful.
I have ordered everything in TO-72 packages as they lend themselves to prototyping easily.
We shall see.
-Chris
I just sent an email to Paul Norton to find out what happened to my sample requests. I have heard nothing since August, 01 of this year.
I must thank you for the link you provided. It was helpful.
I have ordered everything in TO-72 packages as they lend themselves to prototyping easily.
We shall see.
-Chris
anatech said:
Based on my own experiences, a reply will come, albeit they certainly seem to have a small backlog on these kinds of mail. Which makes sense, unless it was a corporate inquiry.
For actual order placement, I'd suggest calling their toll-free number.
What link was that? I believe the link provided here was provided by CheffDeGaar.
Actually, according to the press release, the LSK389 samples are provided in TO-71, packages, which for prototyping purposes is the same thing. You may want to send an e-mail clarifying this.
For production, I rather like the dual in-line packages. Among other things because it is easier for me to get dual in-line heatsinks than TO-71 heatsinks, and I prefer to run the FETs hard but cool.
Also, the flat surface of a dual in-line package allows a large number of devices (e.g. a multiway S-K crossover filter) to be coupled to a single, semi-large heatsink (e.g. a long Austerlitz KS32) using silicon grease/pads or indium-solder (as suggested one of the Lovoltech threads).
In fact, a meter-long cut of the KS32, plus a series of 30mm wide PCBs (or a single, 90cm-long one) and two squares of bitumen, lets me make a 1000mm*32mm*53mm rod with a multichannel, multiway crossover that I can hide virtually anywhere.
Unfortunately, there seems to be some confusion as to the availability of LSK389 in PDIP packaging, as the "Products" page states that "All Dual Devices are available in PDIP, TO-71, TO-5, SOIC-8, SOT23-6, Tested Die", but the datasheet states that it is available as TO-71 and SOIC-8 only.
At least for some applications, I'd be interested in seeing the LSK389 compared to an LS843. The latter has 1mV dVgs matching and 5µV dVgs/dK thermal matching, so you should be able to get insanely low offsets with a voltage divider (input-gate-output) and a source-output resistor.
Edit: Don't tell Mr. Curl
... He'll presumably not be happy about the idea of a significant resistor in series with the gate.Something I thought I'd posted, but apparently hadn't:
Depending on how the LU1014Ds are biased and how hard you're driving them, you may want to reduce or delete the Gate resistors. The Lovoltechs can draw a surprising amount of gate current under certain conditions.
For the record, I have not had any oscillation problems whatsoever. In my favor, however, is the fact that I live in a nearly pre-industrial part of the country. Furthermore, my house is in a valley. RF is just not a problem for me. I had the Gate resistors in as insurance for those who live in the modern world.
Treat those resistors as user adjustable, perhaps even being so bold as to try a straight connection.
Grey
Depending on how the LU1014Ds are biased and how hard you're driving them, you may want to reduce or delete the Gate resistors. The Lovoltechs can draw a surprising amount of gate current under certain conditions.
For the record, I have not had any oscillation problems whatsoever. In my favor, however, is the fact that I live in a nearly pre-industrial part of the country. Furthermore, my house is in a valley. RF is just not a problem for me. I had the Gate resistors in as insurance for those who live in the modern world.
Treat those resistors as user adjustable, perhaps even being so bold as to try a straight connection.
Grey
GRollins said:
Hey Grey,
You weren't possibly refering to 2SK2145:
http://www.semicon.toshiba.co.jp/docs/datasheet/en/Transistor/2SK2145_en_datasheet_071101.pdf
and/or 2SK3320 ?
http://www.semicon.toshiba.co.jp/docs/datasheet/en/Transistor/2SK3320_en_datasheet_071101.pdf
Sorry Grey,
I didn't know you were refering to MOSFETs.
I thought you were talking about JFETs (small sig)
I guess you are correct (on the part No's)
and they are complementary.
http://www.semicon.toshiba.co.jp/docs/datasheet/en/Transistor/2SK2013_en_datasheet_061121.pdf
http://www.semicon.toshiba.co.jp/docs/datasheet/en/Transistor/2SJ313_en_datasheet_061116.pdf
I didn't know you were refering to MOSFETs.
I thought you were talking about JFETs (small sig)
I guess you are correct (on the part No's)
and they are complementary.
http://www.semicon.toshiba.co.jp/docs/datasheet/en/Transistor/2SK2013_en_datasheet_061121.pdf
http://www.semicon.toshiba.co.jp/docs/datasheet/en/Transistor/2SJ313_en_datasheet_061116.pdf
Re: GB
Individual data sheets for the Hitachi MOSFETs can be found here:
http://www.classiccmp.org/rtellason/transistors-2sh-k.html
I do have a hard copy book though (somewhere) and I do
recall they had some example applications with a few amps.
Sorry, don't have anything to scan it with if I can find it though
pimarzu said:
BTW, does anybody have a copy of the original Hitachi Application Note for MOSFET Audio Amplifier?
Thanks
Individual data sheets for the Hitachi MOSFETs can be found here:
http://www.classiccmp.org/rtellason/transistors-2sh-k.html
I do have a hard copy book though (somewhere) and I do
recall they had some example applications with a few amps.
Sorry, don't have anything to scan it with if I can find it though
Nope, I meant MOSFETs. Unless and until they start making high current JFETs (...that stay in production...), you have to switch over to MOSFETs if you want to carry current.
Since I wrote the above-referenced post, I've located and bought some of the 2SK2013/J313. I used them in the GR-25 and like them. The front end for the GR-25 is a good match for the New-Tron output stage.
Hitachi makes a series of similar MOSFETs that should work quite nicely if you can't find the Toshiba parts. No, they're not any easier to find than the Toshiba parts...at least here in the US.
Grey
Since I wrote the above-referenced post, I've located and bought some of the 2SK2013/J313. I used them in the GR-25 and like them. The front end for the GR-25 is a good match for the New-Tron output stage.
Hitachi makes a series of similar MOSFETs that should work quite nicely if you can't find the Toshiba parts. No, they're not any easier to find than the Toshiba parts...at least here in the US.
Grey
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.