For a while now, I have seen that Ixys makes a line of 200V depletion mode mosfets, but I've never seen them available at the usual distibutors (so I've drooled in vain) - until now. I just placed an order from Mouser for some of the IXTH16N20D2, a 16A, 200V, TO-247 device that looks promising as a current source load for one of the Sony or Tokin (or Russian or NP) SITs. One could also make a version of the Pass "De-Lite". I plan to use them as a modulated current source load in one of my "Fake SIT" amplifiers. The price is not all that ruinous.
On another note, both Mouser and Digi-Key are offering some Cree SiC mosfets at less than $10 (remember when they were $90?). I have plans for those guys, too - an all semiconductor transformer-coupled push-pull amp running at your usual tube voltage, probably powered by a hefty 115 - 230V isolation transformer and a small aux transformer. Don't hold your breath, but that project will happen some time in the next year or two....
On another note, both Mouser and Digi-Key are offering some Cree SiC mosfets at less than $10 (remember when they were $90?). I have plans for those guys, too - an all semiconductor transformer-coupled push-pull amp running at your usual tube voltage, probably powered by a hefty 115 - 230V isolation transformer and a small aux transformer. Don't hold your breath, but that project will happen some time in the next year or two....
Last edited:
Member
Joined 2009
Paid Member
All the big IXYS depletion mode devices have high capacitance. This may matter in a current source (the biggest intended application), or it may not. cascoding helps fix the problem. The 500V device Nelson introduced for the "Delite" project (IXTH20N50D) had more capacitance, but worked ok for the that amp. When I get around to it, I'll try them without cascoding first.
Keep in mind also that this is not an EUVL "five nines" perfection type of deal.
Keep in mind also that this is not an EUVL "five nines" perfection type of deal.
Last edited:
I was rummaging through my undifferentiated stuff this evening and came across a pair of 3.5k Tomiko P-P toroidal P-P transformers I'd forgotten I had. These might work well in an all-SS amplifier using the Cree SiC Mosfets or with the 800V ST 800V Si Mosfets I have in profusion.
These may be as crudely set up/wound as the Antek toroidal P-P transformers, or if Tomiko actually did their homework, they may approach the Amplimo or Plitron transformers in quality. I've heard no reports on these transformers from the net either way, so I may be obliged to "let the dice fly high".
These may be as crudely set up/wound as the Antek toroidal P-P transformers, or if Tomiko actually did their homework, they may approach the Amplimo or Plitron transformers in quality. I've heard no reports on these transformers from the net either way, so I may be obliged to "let the dice fly high".
I have three "AMP Cramplette" designs that utilize the 200V IXYS device for an exquisite no-parts take on Papa's modulated current source for his Zen amp series. These amps are not anywhere near EUVL's umpteen decimal points of distortion perfection standard, but should be quite listenable anyway, which is the ultimate goal of this ridiculous hobby in the first place. Once I get the simplest one of these efforts to work in the real world, I might let some schematics trickle out. If you want to beat me to the punch with some proposed schematics, have at it. None of the ones I've come up with so far are inaccessible to a person with LTSpice willing to put in some effort.
I will say this, though. I just took one of the devices I got got from Mouser and loaded it on to a heat sink with a 1 ohm source resistor and a 1k stopper resistor from gate to ground. The drain current is constant at 2A with a 1 ohm source resistor, so this means -2V Vgs for 2A Id. A person with a somewhat nominal amount of savvy can use these values for starting points and forge on.
It also means that these devices are real-deal depletion mode mosfets and not re-marked fakes, as a normal enhancement mode fet or other piece of random junk would display absolutely no drain current (or maybe a dead short for an extreme piece of bilge) under the conditions described.
It also means that these devices are real-deal depletion mode mosfets and not re-marked fakes, as a normal enhancement mode fet or other piece of random junk would display absolutely no drain current (or maybe a dead short for an extreme piece of bilge) under the conditions described.
Attachments
Last edited:
Another dodge would be to replace part of the resistive drop in the source with a semiconductor junction that has a low incremental impedance. This would allow the Ixys fet to bias up with a desired current level but with higher gain, since there is less impedance in series with the source. Assuming I want to keep the same drain current as cited in the above example (2A), one choice would be to use a power Darlington with base connected to drain (~1.1V, a TIP121 or similar would work just fine) in series with a 0.47 ohm resistor. The slightly varying voltage drop with current of the Darlington may provide distortion cancellation (or distortion augmentation, who knows?).
Of course, one could also make a Vbe multiplier using the Darlinton and a voltage divider connected from collector to base to emitter, so that it provides all the bias necessary to park the Ixys fet at a given drain current.
Of course, one could also make a Vbe multiplier using the Darlinton and a voltage divider connected from collector to base to emitter, so that it provides all the bias necessary to park the Ixys fet at a given drain current.
Sticking your choke in the ground leg, eh? That's a nifty dodge for a small negative supply. As always, I'm looking for perverse solutions, being that kinda guy. I haven't had the occasion to use a negative supply to bias any of my fake SIT designs as of yet, though it's been in the back of my mind as a possible option should I need to.
The negative supply would definitely be a possible option for the Depletion mode fet mentioned here in a "Delite" kind of application.
My first use of these fets will likely be as a topside current source load for one of my fake SIT amps. Using the combination of Darlington dropper and resistor may make it possible to have greater freedom in tailoring the gain of the topside current source should I choose to modulate it. This might be usable for tailoring the distortion characteristics of the amp. If I could figure out how to get external libraries added into my PSpice, I'd give it a go in simulation. So far PSpice doesn't recognize the downloadable Ixys libraries for some reason.
The negative supply would definitely be a possible option for the Depletion mode fet mentioned here in a "Delite" kind of application.
My first use of these fets will likely be as a topside current source load for one of my fake SIT amps. Using the combination of Darlington dropper and resistor may make it possible to have greater freedom in tailoring the gain of the topside current source should I choose to modulate it. This might be usable for tailoring the distortion characteristics of the amp. If I could figure out how to get external libraries added into my PSpice, I'd give it a go in simulation. So far PSpice doesn't recognize the downloadable Ixys libraries for some reason.
Yes, you did wait a long time. I have a possible app using these fets that is a variation of some classic Zen and First Watt schematics. It looks as if the design may be equally applicable for a single supply capacitor coupled or dual supply DC coupled approach. More details when I hone my simulation better.
Nice work as usual.
Assuming you go down the cascode path, I've been thinking lately that using a lateral mosfet with it's more desirable capacitance curves might possibly serve as the ideal cascode device to help linearise as well as reduce the capacitance of regular Vfets and Ixys depetion mode mosfets.
I really like Papa's idea of biasing and doing away with the source resistor.
I have a similar project using Ixys devices with almost triode curves waiting to be started, so good to see the ideas you have.
Assuming you go down the cascode path, I've been thinking lately that using a lateral mosfet with it's more desirable capacitance curves might possibly serve as the ideal cascode device to help linearise as well as reduce the capacitance of regular Vfets and Ixys depetion mode mosfets.
I really like Papa's idea of biasing and doing away with the source resistor.
I have a similar project using Ixys devices with almost triode curves waiting to be started, so good to see the ideas you have.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.