www.marklev.com has schematics for the JC-3 and the ML-2. As I understand it, the JC-3 was the prototype and the ML-2 the (greatly fiddled-with) production version. I don't remember offhand who did the fiddling--Tom Colangelo, perhaps? It's been a while since I used those particular brain cells. In any event, the ML-2 is clearly a more complicated circuit than the JC-3, although if you squint your eyes just right you can see a resemblance.
I'm not aware of any other schematics by John being available, unless you want to count the JC-2, but that's a preamp.
Grey
I'm not aware of any other schematics by John being available, unless you want to count the JC-2, but that's a preamp.
Grey
MikeW said:
Mike,
Um, I don't know if this is the proper time and place to admit to this, but I've already done some work along the lines of the JC-3, using all FETs from nose to tail. I've got a prototype half-built. Let me know if you're interested.
I also have--at least conceptually--a bridged version in mind, though I have yet to have time to try it out. I've been wanting to try a biasing trick for a while and a bridged circuit would be an ideal guinea pig for what I have in mind.
If you want bipolar outputs and such, then I may or may not be your guy. I'd have to start from scratch on that. On the other hand, it might be a good chance to play with those big bipolars that have Charles Hansen in such a lather.
Incidentally, many thanks for the caps. I finally got my shop (woodworking--separate and distinct from the area where I do electronics, which I call the lab) dried out. I'm glad that I have all the necessary tools and stuff, but I purely resent having to apply my DIY skills to plumbing...even if it saves me hundreds of dollars. At least I didn't have to have some ugly troglodyte with a hairy butt crack stomping through my house.
John,
The caps Mike is offering are Digikey #P6939-ND. They are in good condition.
Grey
Yes, it is...if you use bipolars for the folded cascode. Using MOSFETs puts me at the mercy of varying Vgs. I either have to use painstakingly matched devices vs. specific resistor values (almost certainly the route you would choose) or make the front end load resistors variable, which creates a different set of headaches.
From a perfectionist point of view, a large quantity of matched devices are the obvious solution. However, the average DIY enthusiast, having a limited number of devices to choose from, is going to want to use the devices on hand. Or at least the minimum purchase he can get away with. It's a nice balancing act.
Couple that with difficulties getting Japanese parts and it becomes an exercise in frustration. I've got 2SK389s and 2SJ109s, both in BL, so that's pretty much a no-brainer...for me. Doesn't mean others can get hold of those parts. At present, I'm using Vishay/Siliconix J310s and J271s as cascodes for the '389/'109s. Fair enough. Except that those parts, too, seem to have fallen off the edge of the Earth. Rats. I'd like to try Toshiba MOSFETs for the folded cascodes, but have been unable to develop a source for them. Okay, put Fairchilds there. They're easy enough to find and reasonably priced. Etc.
It's a nuisance, given the parts availability problems that individuals (as opposed to corporations) face, but do-able. I'm fine-tuning the front end at present, as the open loop bandwidth doesn't suit me. It's around 80kHz, but I'd like to at least get over 100kHz. I regard 200-250kHz as the beginning of acceptable bandwidth, even if I have to use feedback to get there. Yes, I know I'm in the minority on that. Just leave me alone. I'm stubborn on that point. I come pre-equipped with a tube mindset, and that's the lowest bandwidth that I find to equal the 'ease' of a good piece of tube gear.
I agree with you (and Charles Hansen) on feedback. The less, the better. But given the choice between 80kHz and no feedback and 250kHz with a modest amount of feedback, I'll grudgingly accept the feedback as the better alternative. However, if I can get the bandwidth up without feedback, then obviously that's the best of both worlds. (Charles, I hear you laughing at my feeble 80kHz...I hope you inhale a dust mote and choke. [To be read as: I'm jealous, ya bum.])
Them what wants lotsa feedback can jolly well do it themselves. I ain't gonna waste my breath arguing with people who either can't hear or haven't tried. So there.
Grey
From a perfectionist point of view, a large quantity of matched devices are the obvious solution. However, the average DIY enthusiast, having a limited number of devices to choose from, is going to want to use the devices on hand. Or at least the minimum purchase he can get away with. It's a nice balancing act.
Couple that with difficulties getting Japanese parts and it becomes an exercise in frustration. I've got 2SK389s and 2SJ109s, both in BL, so that's pretty much a no-brainer...for me. Doesn't mean others can get hold of those parts. At present, I'm using Vishay/Siliconix J310s and J271s as cascodes for the '389/'109s. Fair enough. Except that those parts, too, seem to have fallen off the edge of the Earth. Rats. I'd like to try Toshiba MOSFETs for the folded cascodes, but have been unable to develop a source for them. Okay, put Fairchilds there. They're easy enough to find and reasonably priced. Etc.
It's a nuisance, given the parts availability problems that individuals (as opposed to corporations) face, but do-able. I'm fine-tuning the front end at present, as the open loop bandwidth doesn't suit me. It's around 80kHz, but I'd like to at least get over 100kHz. I regard 200-250kHz as the beginning of acceptable bandwidth, even if I have to use feedback to get there. Yes, I know I'm in the minority on that. Just leave me alone. I'm stubborn on that point. I come pre-equipped with a tube mindset, and that's the lowest bandwidth that I find to equal the 'ease' of a good piece of tube gear.
I agree with you (and Charles Hansen) on feedback. The less, the better. But given the choice between 80kHz and no feedback and 250kHz with a modest amount of feedback, I'll grudgingly accept the feedback as the better alternative. However, if I can get the bandwidth up without feedback, then obviously that's the best of both worlds. (Charles, I hear you laughing at my feeble 80kHz...I hope you inhale a dust mote and choke. [To be read as: I'm jealous, ya bum.])
Them what wants lotsa feedback can jolly well do it themselves. I ain't gonna waste my breath arguing with people who either can't hear or haven't tried. So there.
Grey
Grey:
with what little tinkering i've done, i'm inclined to agree with you about bipolars for the folded cascode. those video transistors with the low Cob (mentioned in several "transistor choices for VAS?" threads are pretty hard to beat.
the more i play around with feedback designs (still haven't got a no-GNFB power amp going good enough to play with yet
), the more i think borbely's servo 50 is a pretty good design, especially with some tweakings ...
mlloyd1
with what little tinkering i've done, i'm inclined to agree with you about bipolars for the folded cascode. those video transistors with the low Cob (mentioned in several "transistor choices for VAS?" threads are pretty hard to beat.
the more i play around with feedback designs (still haven't got a no-GNFB power amp going good enough to play with yet
), the more i think borbely's servo 50 is a pretty good design, especially with some tweakings ...mlloyd1
Grey I have heard a lot of Pass labs amps. I like them very much. I would like to build one of john's to compare. I do not want to get to far away from the original. Maybe the next one can be MOSFETs. I will try to get a layout going next week. Then I can search for replacement parts. Thanks for the offer, Mike.