Hi jacco,
Everyone reads this thread.
Delft . . .
O . . . I still have one candle given by maritime students during their study tours. Stamping reads 1996. 😀
Sorry for the off-topic, guys!
Regards
jH
Everyone reads this thread.
Delft . . .
O . . . I still have one candle given by maritime students during their study tours. Stamping reads 1996. 😀
Sorry for the off-topic, guys!
Regards
jH
Just for grins, I simulated a JFET array using 30X pieces of J110 (I kept plunking down JFETS until my hand got tired), an IRFP240 for cascode, and an IRFP9240 as a current source load. I set up the drain voltage on the J110s to be about 6V, and set up the current at about 15ma per FET (about 100mW each, the limit for a TO-92 package). It took a lot of juggling to get the output voltage balanced out to around 1/2 VCC (~20V), but when I did, the distortion characteristics looked pretty Zen-like (~0.16% at 1V out into a 16 ohm load, mostly 2nd harmonic, with higher harmonics decreasing pretty smoothly). If you were crazy enough, you could probably get a setup like this to work, especially if you forego absolute Zen purity and use a servo to center the output voltage. J110s are decent FETs with pretty high IDSS, and relatively high transconductance (> 10k mmhos). The Fairchild versions of the part are pretty cheap from outlets like Mouser. With a little bit of airflow, one could possible get more than 100mW per JFET - they will certainly conduct the extra current if you let them. J105s are even better, but you really have to strangle them with a big source resistor to keep the current within the dissipation limits. I ended uo using 24.9 ohms source resistor for the J110 model that ORCAD uses. I would post the circuit, but there's no way I could shrink it down ehough to meet the size limits and still have it readable.
wrenchone said:
Coolness!
Now try it again but as a straight source follower with a +/- 13.8 volt supply and an 11 ohm, 300mH choke load. 😀
se
I'm not tired of drawing JFETs - theyre easier than drawing
Mosfets. However, I think some rules ought to be made clear.
After some education, I learned that no wire should cross
implying a connection, in fact, wires should only connect in "T's".
This eliminates any ambiguity as to whether a dot is at the
intersection indicating a connection.
Similarly, the position of the Gate on a MOSFET or JFET should
favor the Source, so that we have no difficulty determining
which is the Source and which is the Drain. Not so problematic
on a Mosfet if you include substrate connection, which clues you
in to the Drain, but on a JFET symbol, the Gate is often in the
middle, which makes life a little more interesting.
As a service to DIYers, I propose that henceforth the Gate
connection should be placed nearest the Source on Mosfet
and Jfet symbols.
Mosfets. However, I think some rules ought to be made clear.
After some education, I learned that no wire should cross
implying a connection, in fact, wires should only connect in "T's".
This eliminates any ambiguity as to whether a dot is at the
intersection indicating a connection.
Similarly, the position of the Gate on a MOSFET or JFET should
favor the Source, so that we have no difficulty determining
which is the Source and which is the Drain. Not so problematic
on a Mosfet if you include substrate connection, which clues you
in to the Drain, but on a JFET symbol, the Gate is often in the
middle, which makes life a little more interesting.
As a service to DIYers, I propose that henceforth the Gate
connection should be placed nearest the Source on Mosfet
and Jfet symbols.
wrenchone said:
Awwwww, c'mon, man. Be a pal. Hey it's my birthday! Well, it was my birthday. Just a while ago. Alright, a couple of months ago... er, or so. Seriously. The truth is I have an incurable disease and I was rejected by the Make A Wish Foundation.
Ok, ok. If you won't do it for me, do it for the victims of Hurricane Katrina.
The starving in Africa?
Elvis?
Fine! Don't do it you selfish pig! I hope you get warts!
Sheeesh. Some people.
😀
se
When I sold my share of Threshold to Dynatech, I was a little
bit surprised by an early encounter with Richard (Dick) Torberg,
real person, who said, "And what are we doing for starving
children... and why not?" Dynatech had an interesting social
conscience, reflected in individuals as well as corporate
leadership. I think that part of it was that the company had
been founded by technical guys, the kind of people who wouldn't
kill spiders (professional respect between engineers). Helping
humans went without saying.
😎
bit surprised by an early encounter with Richard (Dick) Torberg,
real person, who said, "And what are we doing for starving
children... and why not?" Dynatech had an interesting social
conscience, reflected in individuals as well as corporate
leadership. I think that part of it was that the company had
been founded by technical guys, the kind of people who wouldn't
kill spiders (professional respect between engineers). Helping
humans went without saying.
😎
Nelson Pass said:
Excellent proposal.
I'd taken to using that convention some time back with regard to JFETs and MOSFETs. Though with some schematics it's next to impossible to avoid wires crossing. In those cases, instead of using a simple cross, I'd propose using a "jump."
An externally hosted image should be here but it was not working when we last tested it.
And just to illustrate what Nelson's talking about with regard to JFETs and MOSFETs:
An externally hosted image should be here but it was not working when we last tested it.
se
Mr. Pass,
Did Lovoltech (Low Volume Tech?) ever get back to you re distribution
of their JFET for retail sale?
it's sad that this is the first Zen variation that can't be built 🙁
Did Lovoltech (Low Volume Tech?) ever get back to you re distribution
of their JFET for retail sale?
it's sad that this is the first Zen variation that can't be built 🙁
So we're talking about drawing parts now?
My program (Circad) makes replication easy. Just cut-'n'paste. Or if I feel ambitious, I can do single parts with of a triple-click of the mouse.
As for the form thereof, Circad uses a big old arrow on JFETs, albeit on center. Still, it's not ambiguous. I have learned how to hack the image library, but it's enough work that I don't do it very often.
Grey
My program (Circad) makes replication easy. Just cut-'n'paste. Or if I feel ambitious, I can do single parts with of a triple-click of the mouse.
As for the form thereof, Circad uses a big old arrow on JFETs, albeit on center. Still, it's not ambiguous. I have learned how to hack the image library, but it's enough work that I don't do it very often.
Grey
moe29 said:
They seem to be working on high volume manufacturing accounts.
I have them on order, so when they arrive, there will be enough
for people to get some. Just be patient.
😎
Mr. Pass,
Thanks for your patience with us (me).
But you are partly to blame... for posting such a cool schematic
and write up! 🙂 🙂 🙂
oh, and another question... did you use a computer CPU heatsink
with fan to cool the JFETS?
Thanks for your patience with us (me).
But you are partly to blame... for posting such a cool schematic
and write up! 🙂 🙂 🙂
oh, and another question... did you use a computer CPU heatsink
with fan to cool the JFETS?
Low Voltage Tech has a $1K per line item minimum.
Due to shipping multiples, a min order actually is 1050 pcs on the LU1014D, the LD1014D has a shipping multiple of 2500 pcs.
They are about $1 each at these quantities, next price break is at 5K pcs.
I'm guessing a group buy at 10pcs per individual for 100 buyers at $15~$18 (including import duty, customs bond, shipping, re-pack and re-ship) might be possible. Add about $1 each for more.
Due to shipping multiples, a min order actually is 1050 pcs on the LU1014D, the LD1014D has a shipping multiple of 2500 pcs.
They are about $1 each at these quantities, next price break is at 5K pcs.
I'm guessing a group buy at 10pcs per individual for 100 buyers at $15~$18 (including import duty, customs bond, shipping, re-pack and re-ship) might be possible. Add about $1 each for more.
hitsware said:
This is true for most JFETs, but not all. Sometimes the catalogs will show the die, as for instance the InterFET catalog. If the die is visually symmetrical, then you can interchange the source and drain.
However, some JFETs are asymmetric and exhibit different behavior when the drain and source are switched. I believe that there is some asymmetry in the Toshiba parts. I could be wrong, as I have a vague memory that the long discontinued 2SK146 (two almost as long discontinued 2SK147s in one metal can) may have had the drain and source leads swapped...
I believe at least some JFETs have different capacitances when Drains and Sources are swapped. Whether this is a deal-killer will depend on the application.
I also think I remember seeing a diagram of a JFET where the Gate tapered from Source to Drain (or vice versa).
This ties in with what Charles said. I'll try to remember where I saw these things.
Grey
I also think I remember seeing a diagram of a JFET where the Gate tapered from Source to Drain (or vice versa).
This ties in with what Charles said. I'll try to remember where I saw these things.
Grey
- Status
- Not open for further replies.