Can it be done?
Yes.
Will it be done?
Yes.
I have cobbled together a number of things on the bench and scratched more out on paper. They will appear in due time. Nelson will, no doubt, put out ten or twenty projects over the next six months. Others will also do things, given time and devices to work with.
Grey
Yes.
Will it be done?
Yes.
I have cobbled together a number of things on the bench and scratched more out on paper. They will appear in due time. Nelson will, no doubt, put out ten or twenty projects over the next six months. Others will also do things, given time and devices to work with.
Grey
Yeah, I have been thinking something along those lines also. I wonder what it would take to use those J-fets in the BosoZ? Its easy enough to find out if they work in there without modifications of course. (Thats the approach I would take, being a rather reckless DIY'er!)
But I am not sure at all that its the best way to go! Lots of us wants to know about this I am sure. Several DIY'ers asked me about that particular question, but as you all know by now I am a really lousy designer, but pretty fast when it comes to construction
Well as it stands right now, I will just pop a few Lovoltechs into one of the BosoZ boards I have and see how it goes
A little advice would be really appreciated though Steen
The lovoltech's have the necessary power rating, but their Vds(max) is only 24V, so you would need to either lower the positive rail or cascode them to keep them alive. Otherwise it doesn't look like much alteration would be required.
Does anybody know what the noise figure is for these devices, I was just looking at the datasheet didn't see it. I did see the capacitance and gate charge specs, and the IRF610 vs LU1014 are comparabel on the gate charge specs, but the IRF610 has much lower capacitance ratings. The other issue that concerns me for use in a BZLS is the gate current levels which look a little high for a typical source component to properly drive.
My 2 cents after a very brief look.
Does anybody know what the noise figure is for these devices, I was just looking at the datasheet didn't see it. I did see the capacitance and gate charge specs, and the IRF610 vs LU1014 are comparabel on the gate charge specs, but the IRF610 has much lower capacitance ratings. The other issue that concerns me for use in a BZLS is the gate current levels which look a little high for a typical source component to properly drive.
My 2 cents after a very brief look.
The first F3 prototype used mosfets and was moved over to using Lovoltech JFETs somewhere late in it's development.
The F3 appears to use 1) a voltage regulated PS, 2) a constant current source supplying the gain device(s), 3) a cascoded gain device which makes Vds almost constant and greatly reduces the effect of the non-linear input cap, and 4) the cascode device is modulated to further reduce distortion.
The question I have is how many of these belts and pairs of suspenders were in place when the switch from mosfet to jfet took place? I mean if you truss up your gain device to this extent how much different can the two device type sound?
The F3 is clearly a magic sounding piece judging from the reviews. But how much of it is the JFETs and how much is NP's skill at creating a first class context.
I am not trying to rain on the parade. I'm just as enthusiastic about these new parts as anyone else. But I do wonder. I can't help it.
Regards,
Graeme
The F3 appears to use 1) a voltage regulated PS, 2) a constant current source supplying the gain device(s), 3) a cascoded gain device which makes Vds almost constant and greatly reduces the effect of the non-linear input cap, and 4) the cascode device is modulated to further reduce distortion.
The question I have is how many of these belts and pairs of suspenders were in place when the switch from mosfet to jfet took place? I mean if you truss up your gain device to this extent how much different can the two device type sound?
The F3 is clearly a magic sounding piece judging from the reviews. But how much of it is the JFETs and how much is NP's skill at creating a first class context.
I am not trying to rain on the parade. I'm just as enthusiastic about these new parts as anyone else. But I do wonder. I can't help it.
Regards,
Graeme
Guys, bear in mind that these parts weren't designed with us in mind. As so often happens, we're taking the scraps from the table; accidental beneficiaries of other technologies.
Noise? Whuzzat? These things are supposed to be used in computer power supplies. They could care less about noise. According to my contacts at Lovoltech, they had never considered linear use in any context, much less audio. They were designed as switches. It's quite accidental that they have any application in audio at all...unless you were thinking of building a switching power supply. (Even then the voltage would be pretty low.)
Me? I take scraps and smile. It's rare that anyone actually designs a part with audio in mind.
For those of you about to protest and list off a few Motorola/Toshiba/Hitachi devices, note one thing before proceeding: Those linear devices are/were also used in the driver circuits for mainframe hard drive voice coils. Ask yourself which is the bigger market. Would we (meaning audio folks) have gotten half the parts we now have if it weren't for crossover uses in the computer or medical industries? With the demise of the large disk drives--now replaced with monster RAID arrays like the Hitachi 9900V behind me--an important (to us) market sector has been lost. Given the size of the audio market (rather small in the overall scheme of things) and the slow trend away from linear to class D amps, it is entirely possible that our future depends on such 'accidents' as the discovery of a linear region in the Lovoltech LU1014D.
Grey
Noise? Whuzzat? These things are supposed to be used in computer power supplies. They could care less about noise. According to my contacts at Lovoltech, they had never considered linear use in any context, much less audio. They were designed as switches. It's quite accidental that they have any application in audio at all...unless you were thinking of building a switching power supply. (Even then the voltage would be pretty low.)
Me? I take scraps and smile. It's rare that anyone actually designs a part with audio in mind.
For those of you about to protest and list off a few Motorola/Toshiba/Hitachi devices, note one thing before proceeding: Those linear devices are/were also used in the driver circuits for mainframe hard drive voice coils. Ask yourself which is the bigger market. Would we (meaning audio folks) have gotten half the parts we now have if it weren't for crossover uses in the computer or medical industries? With the demise of the large disk drives--now replaced with monster RAID arrays like the Hitachi 9900V behind me--an important (to us) market sector has been lost. Given the size of the audio market (rather small in the overall scheme of things) and the slow trend away from linear to class D amps, it is entirely possible that our future depends on such 'accidents' as the discovery of a linear region in the Lovoltech LU1014D.
Grey
GRollins said:
I remember reading an interview last year with Dan D'Agostino that Krell had custom manufactured Mosfets for better audio performance.
If I remember correctly, he stated that there were no Mosfets made for audio applications and Krell contracted with TI to custom manufacture Mosfest for audio rather than leveraging what is currently available in the marketplace. The minimum order for custom TI devices was 1 million parts.
Grey, after the Jfets - any chance for a custom GB?
I wonder if Nelson has ever considered taking this path and custom building a semiconductor?
-David
$1,000,000 is slightly beyond my reach, I'm afraid. Perhaps after I win the lottery or write (and sell) the Great American Novel. Given that all I have time for are short stories, the odds of the second option occuring are about on a par with the first.
You have to take these "custom-made" claims with a grain of salt. If you have someone--let's say On-Semi--stamp a stock part with your company logo, it becomes a "custom" part. Sounds good in the literature, it may even sound good in the equipment, but it's not any different from the regular transistor that you could buy...if only you knew which one to order.
You want one that says diyAudio? That would be a trip. Strike fear and envy into the hearts of all your audio buddies.
Grey
You have to take these "custom-made" claims with a grain of salt. If you have someone--let's say On-Semi--stamp a stock part with your company logo, it becomes a "custom" part. Sounds good in the literature, it may even sound good in the equipment, but it's not any different from the regular transistor that you could buy...if only you knew which one to order.
You want one that says diyAudio? That would be a trip. Strike fear and envy into the hearts of all your audio buddies.
Grey
Looking at the ZV8 schematic in fig. 7 it strikes me that replacing the lightbulbs with a ccs would produce a first class transconductance amp.
The self biasing feature eliminates any connection (high resistance or otherwise) from the output. reduces the input cap and distortion over similar mosfet amps. It seems almost too simple.
I haven't been paying too much attention the last few weeks so apologies to all if this has been brought up already.
Graeme
The self biasing feature eliminates any connection (high resistance or otherwise) from the output. reduces the input cap and distortion over similar mosfet amps. It seems almost too simple.
I haven't been paying too much attention the last few weeks so apologies to all if this has been brought up already.
Graeme