FET transfer function - best and worst cases

P

Paul Burridge

Guest
Hi all,

Say a datasheet lists values of

Idss= 20mA - 60mA
Vp= 2V - 6V

(pulling plausible figures out of the air)

Typical values rarely seem to be listed for FETs, so I assume there's
a fairly even distribution of real-world manufacturing variations
between the two extremes.

Now the question is, one would expect the FET with the Idss of 20mA to
have a Vp of 2V in the above example, and one with an Idss of 60mA to
have a Vp of 6V, thereby preserving the characteristic shape of the
curve across the whole range of possible values. But is it possible to
get the occasional rogue real-world FET that has a Vp of 2V and an
Idss of 60mA or one with a Vp of 6V that gives an Idss of 20mA,
thereby with totally different characteristic curves?

Thanks,

p.
--

"What is now proved was once only imagin'd." - William Blake, 1793.
 
Paul Burridge wrote:
Hi all,

Say a datasheet lists values of

Idss= 20mA - 60mA
Vp= 2V - 6V

(pulling plausible figures out of the air)

Typical values rarely seem to be listed for FETs, so I assume there's
a fairly even distribution of real-world manufacturing variations
between the two extremes.

Now the question is, one would expect the FET with the Idss of 20mA to
have a Vp of 2V in the above example, and one with an Idss of 60mA to
have a Vp of 6V, thereby preserving the characteristic shape of the
curve across the whole range of possible values. But is it possible to
get the occasional rogue real-world FET that has a Vp of 2V and an
Idss of 60mA or one with a Vp of 6V that gives an Idss of 20mA,
thereby with totally different characteristic curves?
Click to expand...
More or less, yes, but the curves still all use the same basic equation.
Vt and beta (transconducance parameter) are essentially independent, so
you can have any K.W/L with any Vp.


Kevin Aylward
salesEXTRACT@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.
 
On Fri, 10 Sep 2004 17:20:53 GMT, "Kevin Aylward"
<salesEXTRACT@anasoft.co.uk> wrote:

More or less, yes, but the curves still all use the same basic equation.
Vt and beta (transconducance parameter) are essentially independent, so
you can have any K.W/L with any Vp.
Click to expand...
Did you get my e-mail, Kev?
--

"What is now proved was once only imagin'd." - William Blake, 1793.
 
Paul Burridge wrote:
On Fri, 10 Sep 2004 17:20:53 GMT, "Kevin Aylward"
salesEXTRACT@anasoft.co.uk> wrote:

More or less, yes, but the curves still all use the same basic
equation. Vt and beta (transconducance parameter) are essentially
independent, so you can have any K.W/L with any Vp.

Did you get my e-mail, Kev?
Click to expand...
Aparantly not.

Did you remove EXTRACT. The other one is kevin.aylward with the obvious
addition.

Kevin Aylward
salesEXTRACT@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.
 
On Fri, 10 Sep 2004 19:20:20 GMT, "Kevin Aylward"
<salesEXTRACT@anasoft.co.uk> wrote:

Paul Burridge wrote:
On Fri, 10 Sep 2004 17:20:53 GMT, "Kevin Aylward"
salesEXTRACT@anasoft.co.uk> wrote:

More or less, yes, but the curves still all use the same basic
equation. Vt and beta (transconducance parameter) are essentially
independent, so you can have any K.W/L with any Vp.

Did you get my e-mail, Kev?

Aparantly not.

Did you remove EXTRACT. The other one is kevin.aylward with the obvious
addition.
Click to expand...
I assume you own the domain in question? Ergo, any old string before
the @ should get through. Ah! I see. I take that back. I'll resend it
now with the appropriate bit removed.
--

"What is now proved was once only imagin'd." - William Blake, 1793.
 
In article <7pj3k0dhmuph1q2hk8dqjo07fh3kpe20pu@4ax.com>,
Paul Burridge <pb@notthisbit.osiris1.co.uk> wrote:

Say a datasheet lists values of

Idss= 20mA - 60mA
Vp= 2V - 6V

Now the question is, one would expect the FET with the Idss of
20mA to have a Vp of 2V in the above example, and one with an
Idss of 60mA to have a Vp of 6V..........
[snip]
Click to expand...

The correspondance is nearly but not quite.

They will have 100% measured Idss to select out chips
suitable for that particular part number. So you can
believe the 24-60mA Idss for your J310's. But the
-2 to -6.5V Vgs(off) min/max values are not measured,
they are taken from the nominal values for the source
chip, with the window widened-out (for tolerancing and
the problems of leakage current when packaged).

The old Siliconix data manuals would give data for both
the device and source chip. The original U310/E310/J310
devices were selected from their wafer/chip known as NZT.
(But see note at end.)

One of the graphs on the NZT data sheet is a plot of Idss
versus Gate-Source Cutoff Voltage. It shows that a 24mA
chip would have a typical Vgs(off) of about -2.7V and a
60mA chip about -4.5V.

Those are the two Vgs(off) numbers we would use when
doing the two Id= Idss(1 - etc ) curves to find the
nominal Id biassing point for an E310.

Note that the Vgs sums only work when Vds is relatively
high, normally taken as at least 2x the maximum Vgs(off).

Note: The NZT data was taken from a 1979 Siliconix book,
but I see that by 1986 they were taking J310's from their
NZA wafer/chip. The 24-60mA Vgs(off) voltages have changed
slightly for the NZA, to -1.9V and -4V respectively. This is
probably more representative of a recent J310.

--
Tony Williams.
 
Paul Burridge wrote:
On Fri, 10 Sep 2004 19:20:20 GMT, "Kevin Aylward"
salesEXTRACT@anasoft.co.uk> wrote:

Paul Burridge wrote:
On Fri, 10 Sep 2004 17:20:53 GMT, "Kevin Aylward"
salesEXTRACT@anasoft.co.uk> wrote:

More or less, yes, but the curves still all use the same basic
equation. Vt and beta (transconducance parameter) are essentially
independent, so you can have any K.W/L with any Vp.

Did you get my e-mail, Kev?

Aparantly not.

Did you remove EXTRACT. The other one is kevin.aylward with the
obvious addition.

I assume you own the domain in question? Ergo, any old string before
the @ should get through.
Click to expand...
No. I set it up to bounce back if not exactly correct. Two much spam
otherwise.

Ah! I see. I take that back. I'll resend it
now with the appropriate bit removed.
Click to expand...
And working on the new bug you discovered!

Kevin Aylward
salesEXTRACT@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.
 
On Fri, 10 Sep 2004 17:26:34 +0100, Paul Burridge
<pb@notthisbit.osiris1.co.uk> wrote:

Hi all,

Say a datasheet lists values of

Idss= 20mA - 60mA
Vp= 2V - 6V

(pulling plausible figures out of the air)

Typical values rarely seem to be listed for FETs, so I assume there's
a fairly even distribution of real-world manufacturing variations
between the two extremes.

Now the question is, one would expect the FET with the Idss of 20mA to
have a Vp of 2V in the above example, and one with an Idss of 60mA to
have a Vp of 6V, thereby preserving the characteristic shape of the
curve across the whole range of possible values. But is it possible to
get the occasional rogue real-world FET that has a Vp of 2V and an
Idss of 60mA or one with a Vp of 6V that gives an Idss of 20mA,
thereby with totally different characteristic curves?

Thanks,

p.
Click to expand...

Take a look at PHEMTs. They mostly have very repeatable transfer
curves. The Fujitsu (just renamed to Eudyna, whatever that means)
FSU-series are amazing; looking at pulsed transfer curves, you
literally can't tell two fets apart. Transconductance is like 0.1 and
capacitances are very low; much better for RF than jfets.

The PHEMTs have lots of gate leakage, 1/f noise, and a very strange
low-frequency dispersion mechanism that makes them unsuited to many
low-current or low-frequency applications.

John
 
I read in sci.electronics.design that John Larkin <jjlarkin@highlandSNIP
techTHISnologyPLEASE.com> wrote (in <ak96k0d9a1uchf14vontulbnudsoo05pp6@
4ax.com>) about 'FET transfer function - best and worst cases', on Sat,
11 Sep 2004:
just renamed to Eudyna, whatever that means
Click to expand...
Greek, 'good force'.
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk
 
On Sat, 11 Sep 2004 20:45:30 +0100, John Woodgate
<jmw@jmwa.demon.contraspam.yuk> wrote:

I read in sci.electronics.design that John Larkin <jjlarkin@highlandSNIP
techTHISnologyPLEASE.com> wrote (in <ak96k0d9a1uchf14vontulbnudsoo05pp6@
4ax.com>) about 'FET transfer function - best and worst cases', on Sat,
11 Sep 2004:
just renamed to Eudyna, whatever that means

Greek, 'good force'.
Click to expand...

My favorite Taiwanese company name is "Mean Well Electronics."

John
 
On Sat, 11 Sep 2004 14:53:24 -0700, John Larkin
<jjlarkin@highlandSNIPtechTHISnologyPLEASE.com> wrote:

On Sat, 11 Sep 2004 20:45:30 +0100, John Woodgate
jmw@jmwa.demon.contraspam.yuk> wrote:

I read in sci.electronics.design that John Larkin <jjlarkin@highlandSNIP
techTHISnologyPLEASE.com> wrote (in <ak96k0d9a1uchf14vontulbnudsoo05pp6@
4ax.com>) about 'FET transfer function - best and worst cases', on Sat,
11 Sep 2004:
just renamed to Eudyna, whatever that means

Greek, 'good force'.


My favorite Taiwanese company name is "Mean Well Electronics."
Click to expand...
Muslims reckon that if you wipe your arse with the right hand, it's
unlucky.
--

"What is now proved was once only imagin'd." - William Blake, 1793.
 
In article <nl07k0l06347ig1p6e530cplsrumf3olqe@4ax.com>,
Paul Burridge <pb@notthisbit.osiris1.co.uk> wrote:

Muslims reckon that if you wipe your arse with the right hand,
it's unlucky.
Click to expand...
Like a number of (apparently) religious strictures,
it has a basis of practicality. In any country with
little water it is not always practical to wash
properly. So you wipe your arse with your left
hand, and make damned sure you only use your right
hand to eat with.

Get things mixed up, and you are almost guaranteed
to be 'unlucky'.

--
Tony Williams.
 
On Sat, 11 Sep 2004 14:53:24 -0700, the renowned John Larkin
<jjlarkin@highlandSNIPtechTHISnologyPLEASE.com> wrote:

On Sat, 11 Sep 2004 20:45:30 +0100, John Woodgate
jmw@jmwa.demon.contraspam.yuk> wrote:

I read in sci.electronics.design that John Larkin <jjlarkin@highlandSNIP
techTHISnologyPLEASE.com> wrote (in <ak96k0d9a1uchf14vontulbnudsoo05pp6@
4ax.com>) about 'FET transfer function - best and worst cases', on Sat,
11 Sep 2004:
just renamed to Eudyna, whatever that means

Greek, 'good force'.


My favorite Taiwanese company name is "Mean Well Electronics."

John
Click to expand...
Mine is "Sure Fire Electrical".


Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
 
On Sun, 12 Sep 2004 10:05:49 +0100, the renowned Tony Williams
<tonyw@ledelec.demon.co.uk> wrote:

In article <nl07k0l06347ig1p6e530cplsrumf3olqe@4ax.com>,
Paul Burridge <pb@notthisbit.osiris1.co.uk> wrote:

Muslims reckon that if you wipe your arse with the right hand,
it's unlucky.

Like a number of (apparently) religious strictures,
it has a basis of practicality. In any country with
little water it is not always practical to wash
properly. So you wipe your arse with your left
hand, and make damned sure you only use your right
hand to eat with.

Get things mixed up, and you are almost guaranteed
to be 'unlucky'.
Click to expand...
Which is why removing the right hand surgically is a ticket to being
shunned for life. Nobody wants some dude fishing around in the
communal food bowl with his wiping hand.

Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
 
In article <kju8k0d651a41mk8l6aahsj9ecuu4mn73s@4ax.com>,
Spehro Pefhany <speffSNIP@interlogDOTyou.knowwhat> wrote:

Which is why removing the right hand surgically is a ticket to
being shunned for life. Nobody wants some dude fishing around in
the communal food bowl with his wiping hand.
Click to expand...
Blimey! Never thought of that.......

--
Tony Williams.
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top