Not really hard to find, but the price is rising from week to week. And when you need a batch to match some quads. DIY could get very expencive.
And people from China starts to sell fakes or sheet parts. I have bought 200 pcs 2SJ74BL (promised origin Toshiba) from a seller at DH-Gate.com. And those parts have a an Idss range from 3mA to 22mA. Temperature behavior is strange, and a friend have used them in a Hiraga 8 Watt amp, he told me that they do not bring the current like 2SJ74 from reliable sellers. So be careful where you buy your 2SJ74
So, Michael what about building your circlotron with Semisouth J-Fets? As Papa Pass did with the Nemesis in Linear Audio Volume 0? ( I have 4 matched pcs waiting )
Funny two days ago I was talking about that to a friend, and now I found these thread.
Kind regards to all
Sam
Hello Luke,
The SemiSouth version in my blog is just the original circuit with the JFETs dropped-in. This was to show what could be expected with no changes.
Was it subjectively better? Any chance I could offer you a raincheck there?
That being said, there could be a short article in our future.
I hope my answer was somewhat useful, if perhaps evasive and noncommital.
Mike
The SemiSouth version in my blog is just the original circuit with the JFETs dropped-in. This was to show what could be expected with no changes.
Was it subjectively better? Any chance I could offer you a raincheck there?
That being said, there could be a short article in our future.
I hope my answer was somewhat useful, if perhaps evasive and noncommital.
Mike
I normally build tube amps, but for a change I decided to build the FET Circlotron. Since I need to drive it from an unbalanced source, I also added a Balanced Zen line stage. I completed construction this weekend, and it certainly sounds very good. But I do have a problem with the temperature the line stage runs at. I'm using precisely the heatsinks specified in the article. The preamp heatsinks reach almost 200 Fahrenheit, which seems a bit excessive really. (Voltages and currents are all consistent with those in the article.)
I could play around with beefing up the heatsinks, but it seems to me a better option might be a different preamp driver for the circlotron. All I need is to convert unbalanced to balanced, with a little bit of gain. Can anyone recommend a design for this that doesn't run hot? Either discrete components or opamps?
Thanks!
Chris
I could play around with beefing up the heatsinks, but it seems to me a better option might be a different preamp driver for the circlotron. All I need is to convert unbalanced to balanced, with a little bit of gain. Can anyone recommend a design for this that doesn't run hot? Either discrete components or opamps?
Thanks!
Chris
Hi Chris,http://www.diyaudio.com/forums/pass-labs/176720-new-line-stage.html We did some work on this line stage a while back. I have not had a chance to get back and build a second one to listen to, but it looked promising. Very low power and gains of 5-10 are doable. Will do SE to DIFF and Diff to SE with fairly low offset. I will be back to it shortly for use with the FX5. May be a little self serving to suggest it at this stage but it is dirt simple to build and I would like to see other opinions if anyone is interested it.Cheers,DaveP.S. Something funny with the editor allowing hyperlink insertion. You will have to copy it and paste it into the address bar.
Last edited by a moderator:
OPA1632/THS4131 (full differential amp manufactured by Texas Instruments with topology licenced from Mr. Pass) is basically a UGS modul confined to single chip. I love the way it sounds and I use it whenever I need unbalanced to balanced conversion (it's not hard to devise a similar discrete circuit but THS4131/OPA1632 is really hard to beat).
I used it in my version of circlotron with success:
http://www.diyaudio.com/forums/pass-labs/157915-simple-cirklotron-lateral-mosfets.html#post2034724
nope, as with just about every decent modern opamp from Ti they only come in soic8 or dgn type packages. some of the linear tech chips come in to220 even, but they are not of use here. if you are designing a pcb i highly recommend the dgn package, better current capability due to a sink on the bottom. you can still hand solder this, but you have to put a via directly under the pad underneath so you can solder it down for heatsinking by flowing solder into the via with it flipped over, without this it performs worse than the regular soic8.
pretty crappy spec on that chip though, i cant comment on the sound as i stopped plans when i saw the numbers. numbers arent everything of course, but if im gonna use an opamp at all, it would want to measure well, the soic8 really isnt that hard to work with if you cant deal with the dgn right now. i could solder a couple soic8 to dip adapters for you if that helps you out
Last edited:
Well, now you've piqued my curiousity. Which spec do you find crappy?
http://www.thatcorp.com/datashts/THAT_1606-1646_Datasheet.pdf
http://www.thatcorp.com/datashts/THAT_1606-1646_Datasheet.pdf
Last edited:
Thanks again for the suggestions. Meanwhile, I've played around with the Balanced Zen linestage, setting the mosfets perpendicular to the board, with somewhat bigger heatsinks. I've got the temperature down to about 185F now, and hope to reduce it more with a fan I've ordered. So maybe that will be good enough for now.
By the way, the Amazing Circlotron really lives up to its name, I think; it sounds truly great! I've been trying to compare it with my vacuum tube OTL amps, both in sound, and in measurements. As best as I can judge, the sound is pretty much indistinguishable, to my ears at least. The output impedance is about 0.8 ohms, which is a bit less than my vacuum tube OTL circlotron based on the design by Alan Kimmel (about 1 ohm). Both of these are beaten my inverse Futterman type OTL based on a design by Tim Mellow, using 6C33C tubes, which has about 0.25 ohms output impedance. There is a wee bit more noise (hiss) on the output of the mosfet circlotron than on the tube amps; enough to be noticeable close up to the speakers, which are high sensitivity Lowthers. But it may be mostly noise from the line stage rather than the circlotron itself; I haven't checked into that yet. Anyway, for a first foray into solid-state construction, I have to say I'm impressed! And even though it's class A, it certainly consumes a lot less power than a vacuum tube OTL!
Chris