John Curl's Blowtorch preamplifier part II

[scott wurcer]

Idss is controlled by another variable, does anyone know what it is?

Do you? Come on John on these short channel JFET's there is some variance from drawn geometry and you CAN'T see it in a microscope. The Idss probably has most of its variation from implant dose. The basic equations are readily available for Vp, Idss, and Cgs,gd vs geometry, junction depth, doping, etc.

 

[iko]

Hm, let me re-phrase. Since the absolute value of said capacitance isn't what I'd be after, is there enough variability in samples of the same device, say a few 2sk170 to warrant measurement and ranking? Any easy way to gauge the capacitance of interest visually? Like SY's suggestion?

 

[syn08]

What is the mechanism for the spread in Idss? It is not geometry, is it?

Surprise, surprise, short of implant dose, it's mostly geometry. Namely, the W/L process variations, and the short channel effects (aka channel length modulation).

 

[john curl]

Well, everyone, I was told this morning by Edwin Oxner, consultant to Siliconix and Interfet, author of books on FET design, and friend to Linear Systems, that Idss is primarily caused by differences in resistivity of the silicon material. He said that if they were made with GaAs, they could all have virtually the same Idss, because the material can be controlled more easily. That's what he says, I am only the messenger.
PS. He liked the pictures of Burning Amp that Kirkwood Rough sent him. Kirkwood is next on the list to ask.

 

[Key]

All those DIY'ers not in the industry posting her, raise your hands...

Raises hand - but that doesn't mean the other stuff doesn't apply to me .

 

[jacco vermeulen]

Speaking of geometry, anything that can be derived from this picture ?

(i'm a food hoarder, the kids overhere call me a fat, dumb, old, bald guy. Dirty is somewhere in there as well, but i forgot the rank)

 

[john curl]

I do know this: If I directly replaced a 2SK146-J73 pair with a 2SK240-J75 pair, with the SAME Idss, my measured distortion with a 10K input pot, would drop. This is what I did with the JC-80 discussed here on previous pages.

 

[AndrewT]

the kids overhere call me a fat, dumb, old, bald guy. Dirty is somewhere in there as well, but i forgot the rank)

are we allowed to start a lottery on who can rank them in the correct order?
Sorry, I'm more than a little off topic.

 

[syn08]

Well, everyone, I was told this morning by Edwin Oxner, consultant to Siliconix and Interfet, author of books on FET design, and friend to Linear Systems, that Idss is primarily caused by differences in resistivity of the silicon material. He said that if they were made with GaAs, they could all have virtually the same Idss, because the material can be controlled more easily. That's what he says, I am only the messenger.
PS. He liked the pictures of Burning Amp that Kirkwood Rough sent him. Kirkwood is next on the list to ask.

Not for short channel JFETs like 2SK170, that also have the channel doping controlled by ion implantation.

What Ed is telling is a good approximation for long channel JFETs, like the 2N5458.

You are welcomed, I'll eventually send you the bill

 

[scott wurcer]

He said that if they were made with GaAs, they could all have virtually the same Idss

John what do you think controls the channel resisivity?

NEC N channel GASFET -> (Don't drool folks it probably has 1000nV/rt-Hz @ 1kHz)

 

[john curl]

Scott, I am only the messenger. Do you know who Ed Oxner is? Would you like to talk to him about this? I will give you his phone #. He was driving in his car when I called, so I did not ask for further detail, but he likes to teach. How about it?

 

[syn08]

Would you like to talk to him about this? I will give you his phone #.

Naaah Scott, you'd better call Dick Sequerra

 

[scott wurcer]

Scott, I am only the messenger. Do you know who Ed Oxner is? Would you like to talk to him about this? I will give you his phone #. He was driving in his car when I called, so I did not ask for further detail, but he likes to teach. How about it?

I don't want to put him on the spot, I don't think he was fully thinking about your question. 1) I have seen no GASFET's that were any better than silicon ones for Idss spread. 2) GASFETs have little or no use at audio frequencies.

 

[QSerraTico_Tico]

Speaking of geometry, anything that can be derived from this picture ?

(i'm a food hoarder, the kids overhere call me a fat, dumb, old, bald guy. Dirty is somewhere in there as well, but i forgot the rank)

What a nice picture Jacco

 

[john curl]

Well, he doesn't want to talk to you either. He just said so. Sorry, Scott.

 

[Steve Eddy]

Speaking of geometry, anything that can be derived from this picture ?

Why does it look like S2 should be D2 and D2 should be S2? Or the other way around, S1 should be D1 and D1 should be S1.

Are they labeled correctly?

se

 

[Steve Eddy]

Well, he doesn't want to talk to you either. He just said so. Sorry, Scott.

Why? Did he say he had cooties or something?

Jeeez. Grow up.

se

 

[PMA]

If JFETs, don't care about D or S. Makes no difference.

 

[scott wurcer]

Why does it look like S2 should be D2 and D2 should be S2? Or the other way around, S1 should be D1 and D1 should be S1.

Are they labeled correctly?

se

S2 and D2 should be swapped, gate should be away from drains and the drains sould be far apart (as is possible).

 

[Steve Eddy]

S2 and D2 should be swapped, gate should be away from drains and the drains sould be far apart (as is possible).

Thanks, Scott!

You're pretty smart for a cootie butt.

se