On Fri, 28 Jul 2006 06:08:01 GMT, David
wrote:

Roy,

So if I had an adjustment on VG2 for each circuit and adjust for
required Drain current on each product ?

Or pick a set of bogy values and accept there will be a range of
operating current.

I've found DGmosFET perfomance is not greatly impacted by
variations in Idss and Gm for practical circuits.

What's the consensus regarding Common Base BJT as LNA ?

Hard matching the input at any decent current, at 4-5ma the input R
is around 5 ohms. Easily overloaded as a result of usually being used
at low current to make the match easier. It's feature is fair
stability and the device is working at at it's alpha cutoff frequency.
The latter was more important 30 years ago when UHF transistors
were harder to get. Noise performance was dependent on device
but even in 1972 I could get TIMX10s down to around 1.5DB at
450mhz.


Allison


Thanks

Regards

David


Roy Lewallen wrote:
The problem with using FETs of all kinds is the wide part-to-part
variation. Look at the specs for the BF998 - many of the critical specs
show only a maximum or minimum, but not both, or just a typical value.
You can be way off if you simply use a "typical" set of curves. If you
want to do an analytical design with a part with non-specifications like
this is to use a curve tracer to generate curves for the individual
part, then use those curves for your design. Pull another part of the
same part number from your drawer, and you'll need a different design.
This exercise is useful for educational purposes, but it isn't a
technique you can use to design something that can be easily duplicated.

That's probably why you don't see a lot of FETs being used in commercial
products, except in applications where there's a lot of feedback to
stabilize the operating point, such as source followers, or when simply
nothing else will do. Even then, the manufacturer has probably paid the
vendor to select parts with a much narrower, and well specified, range
of characteristic values. That's been my experience in designing
commercial electronic test equipment.

Roy Lewallen, W7EL

David wrote:
Tim,

Say we look at VG1s = 0.1V as per your example.

The graph for BF998 shows that if VG2s = 4V and VDs = 8V then ID
approx = 12.5mA

This would mean that unless I applied a negative voltage on the source
I would need to apply 0.1V forward bias to G1 and 4V to G2 ?
As Rs is creating a negative self bias voltage ?

If I set the bias point lower - say 5mA then VG1s is approx. -0.2V
according to the graph.

I can achieve this by using a resistor in the source of 0.2/5mA (40
Ohms) and then set VG1 = 0 (so that VG1s = -0.2V) and then 4.2V on G2
so that VG2s = 4V.

Is this correct ?

The transfer characteristic curve shows that for say 10mA. If VG2s =
4V then gm = around 24mS and if VG2s is reduced to 0V the gm reduces
to about 7mS.

Thanks

regards

David

Sorry, I have almost no experience in using dual-gate MOSFETS, for the
reasons I mentioned. Beginning somewhere around the mid-80s, when
manufacturing became highly automated, my employers strongly discouraged
designs which included any adjustments. Besides the labor required to
make the adjustment, the variable component lowered the product's
reliability, so we'd often design in a lot of parts to get around having
any tweaks. But individual adjustment is still a viable option for some
products. What I don't know is whether you'd get satisfactory
performance with widely differing devices all running at the same drain
current. That would depend on your design and application.

If you're considering making a product, I'd certainly do some modeling
with extreme component parameters to see what happens. And you might
consider some sort of device selection and/or incoming inspection or
sorting to make sure you don't get any truly extreme parts. One thing to
be careful about is that when a part is so poorly specified, other
companies might be buying large numbers of selected parts. That leaves
you with the leftovers. I've seen some really strange distributions
resulting from this -- parts with extreme characteristics on both ends,
but nothing anywhere near the "typical" values. It used to be common
with zeners, until they got better at making -- 10% tolerance zeners
would all be between 5 and 10% from nominal, in both directions, with
none closer than 5%. Those had been selected out and sold as 5%
tolerance parts.

Just a few of the things I've picked up in 30 or so years as a design
engineer.

Roy Lewallen, W7EL

David wrote:
Roy,

So if I had an adjustment on VG2 for each circuit and adjust for
required Drain current on each product ?

What's the consensus regarding Common Base BJT as LNA ?

Thanks

Regards

David

Ah, but that's a model for just one BF998. The one you pull out of the
drawer might bear very little resemblance to it.

Roy Lewallen, W7EL

jack wrote:
FWIW, I found the following SPICE model for the BF998 on the web -- note
that the two MOSFETs VMAX are dissimilar, a lot of the entries are for
parasitic elements:
. . .

FWIW, I found the following SPICE model for the BF998 on the web -- note
that the two MOSFETs VMAX are dissimilar, a lot of the entries are for
parasitic elements:

* BF998 SPICE MODEL OCTOBER 1993 PHILIPS SEMICONDUCTORS
* ENVELOPE SOT143
* 1.: SOURCE; 2.: DRAIN; 3.: GATE 2; 4.: GATE 1;
..SUBCKT BF998 1 2 3 4
L10 1 10 0.12N
L20 2 20 0.12N
L30 3 30 0.12N
L40 4 40 0.12N
L11 10 11 1.20N
L21 20 21 1.20N
L31 30 31 1.20N
L41 40 41 1.20N
C13 10 30 0.085P
C14 10 40 0.085P
C21 10 20 0.017P
C23 20 30 0.085P
C24 20 40 0.005P
D11 42 11 ZENER
D12 42 41 ZENER
D21 32 11 ZENER
D22 32 31 ZENER
RS 10 12 100
MOS1 61 41 11 12 GATE1 L=1.1E-6 W=1150E-6
MOS2 21 31 61 12 GATE2 L=2.0E-6 W=1150E-6


..MODEL ZENER D BV=10 CJO=1.2E-12 RS=10


..MODEL GATE1
+ NMOS LEVEL=3 UO=600 VTO=-0.250 NFS=300E9 TOX=42E-9
+ NSUB=3E15 VMAX=140E3 RS=2.0 RD=2.0 XJ=200E-9 THETA=0.11
+ ETA=0.06 KAPPA=2 LD=0.1E-6
+ CGSO=0.3E-9 CGDO=0.3E-9 CBD=0.5E-12 CBS=0.5E-12


..MODEL GATE2
+ NMOS LEVEL=3 UO=600 VTO=-0.250 NFS=300E9 TOX=42E-9
+ NSUB=3E15 VMAX=100E3 RS=2.0 RD=2.0 XJ=200E-9 THETA=0.11
+ ETA=0.06 KAPPA=2 LD=0.1E-6
+ CGSO=0.3E-9 CGDO=0.3E-9 CBD=0.5E-12 CBS=0.5E-12


..ENDS BF998

In ancient times I set up a VHF amplifier test circuit with a dual-gate
3NXXX and measured gain and IM distortion vs bias parameters, DC feedback
and negative RF feedback. I was able to compare many samples and found that
I could get adequate (for my purposes) uniformity for a dozen samples in a
particular application. The tradeoff is reduced gain to get "improved"
uniformity. Trying to get max results from a single stage is not as good as
two or three cascaded more modest stages. The cost of this approach is of
course not minimal. The first stage dominates total noise figure.

Bill W0IYH

"David" wrote in message
...
OK, I think I got it.

Let's say I want Vdd = 8V. Say I have an RF chock in the Drain then Vds is
almost 8V.

Say I want 5mA bias current.

I look at the Vds vs Id and find that VG1-s = -0.2V for this condition.
(So actual VDs is 7.8V)

Next I use a source resistor of VG1-s/ Id = 40 Ohms
As VG1 = VG1-s + VG1 - IDxRS the Gate voltage applied is 0V.

The graph on the datasheet mentions that this is the output when VG2-s is
4V.

So now I calculate VG2 = VG2-s + ID x RS = 4.2V

I suppose that I can make Rs smaller to the point where VG1-s is 0V on the
graph.(Occurs at 10mA for BF998).

If I want higher Q point I then add positive bias to V1. I also understand
I can get around 10dB attenuation by lowering VG2-s from 4.2V
towards 0V but need to make it negative if I want to switch the signal off
altogether (If I was to use the FET as say a ASK or OOK modulator).

Thanks



tim gorman wrote:
David wrote:


Hi,

Pretty fundamental I know but can someone please explain the steps for
setting up bias for a Dual Gate MOSFET.

I know I could place a pot on the gate and source for each circuit and
play with values but I would like a method that enables me to calculate
the values.

The main issue is how to determine values for Rs and Gate 2 Voltage.

I am using BF998 and want to have a "play" at 5V and 8V supply.

The formulae for Id is Id = Idss(1-Vgs/Vp) ^ 2
But Idss is stated as 2-18mA
Vp Gate 1 is given as a range from 1-2V
Vp gate 2 is given as range from 0.5 to 1.5V

If I apply say 4V to G1 and 0V to G2, how do I calculate the voltage at
the source to determine Vgs ?

Any help much appreciated.

Regards

David


One way of doing this is to get the datasheet for the FET you are using.
There should be a graph that shows the operating characteristic curves.
The
x-axis will be Vds and the Y-axis will be the drain current Id. The
characteristic curves will be for various levels of Vgs. Pick an
operating
point based on the type of amplifier you want. Let's suppose it will be
Class A. Assume the FET has a power supply voltage of 40v and an Idss of
10ma. Let's say that you pick a point in the middle of the operating
curves
that gives an Id of 6ma and a Vds of 20v in order to get the maximum
swing
out of the amplifer. Looking at the characteristic curves shows that this
will require a Vgs of about -1v. Now you have everything you need. If Vgs
needs to be -1v and Id is 6ma (assume Id and Is will be the same) you
need a resistor of Vd/Id (R = V/I) or about 166 ohms. The gate resistor
you see in FET amps is not really there for biasing but
more to set the input impedance of the amplifier. As long as the leakage
current from the gate to the source is small, Vgs is set by the bias
resistor in the source lead.

tim ab0wr

On Sat, 29 Jul 2006 00:00:12 -0700, Roy Lewallen
wrote:

Sorry, I have almost no experience in using dual-gate MOSFETS, for the
reasons I mentioned. Beginning somewhere around the mid-80s, when
manufacturing became highly automated, my employers strongly discouraged
designs which included any adjustments. Besides the labor required to
make the adjustment, the variable component lowered the product's
reliability, so we'd often design in a lot of parts to get around having
any tweaks. But individual adjustment is still a viable option for some
products. What I don't know is whether you'd get satisfactory
performance with widely differing devices all running at the same drain
current. That would depend on your design and application.

If you're considering making a product, I'd certainly do some modeling
with extreme component parameters to see what happens. And you might
consider some sort of device selection and/or incoming inspection or
sorting to make sure you don't get any truly extreme parts. One thing to
be careful about is that when a part is so poorly specified, other
companies might be buying large numbers of selected parts. That leaves
you with the leftovers. I've seen some really strange distributions
resulting from this -- parts with extreme characteristics on both ends,
but nothing anywhere near the "typical" values. It used to be common
with zeners, until they got better at making -- 10% tolerance zeners
would all be between 5 and 10% from nominal, in both directions, with
none closer than 5%. Those had been selected out and sold as 5%
tolerance parts.

Just a few of the things I've picked up in 30 or so years as a design
engineer.

Roy Lewallen, W7EL

Roy,

There are a large number of VHF and HF radios both past and present
that use the DGmosfets in place like the RF amp sometimes a balanced
mixer. A recent example is the TenTec 526 (6n2) 6 and 2 meter radio
as the RF amp. Another current example is the TT 1208 transverter
also as RF amp. If your not pushing the device for max gain or
ultimate IP3 possible they perform well and offer low noise for their
power needs. Neither require pots or other tweaking (other than
tuneable circuits). Can other deivces be used to do better, yes.
But, engineering is always about understanding and compromize.

I'd never use DGfets for something like a scope amp or other
instrumentation where DC operating point or balance is a
requirement. I have use them as a high impedence (1mohm)
AGC'd input for high input senstivity frequency counters.
Another place where I've used them is IF amps, they are just far
easier to AGC than CA3020 or MC1350 and quieter.

My favorite line from years back.

Good, Fast, Cheap, Pick any two.


Allison

Doesn't say much about their quality control if the devices are that
dissimilar.

FWIW, I have developed a semiconductor and tube curve tracer for use within
Multisim -- create the component in Multisim, plug it into the curve tracer,
run the test and compare with the manufacturer's charts.

I have also built a tube curve tracer -- same will apply for
semiconductors -- even published an article on it. With this i can compare
the characteristics of the device vis a vis mfr data sheet AND the
simulation in EWB.

Jack

"Roy Lewallen" wrote in message
...
Ah, but that's a model for just one BF998. The one you pull out of the
drawer might bear very little resemblance to it.

Roy Lewallen, W7EL

jack wrote:
FWIW, I found the following SPICE model for the BF998 on the web -- note
that the two MOSFETs VMAX are dissimilar, a lot of the entries are for
parasitic elements:
. . .

In article ,
wrote:
I'd never use DGfets for something like a scope amp or other
instrumentation where DC operating point or balance is a
requirement.

Interestingly enough, in Jim Williams's book "The Art and Science of
Analog Circuit Design", there's a chapter by Steve Roach of Tektronix,
which shows the use of a BF996 consumer-grade dual gate FET in a 1GHz
oscilloscope preamp.

http://books.google.com/books?vid=IS...VGMx8v3yDD4at4