Dual Gate MOSFET: 40673 specs and/or equivs

[John Mitchell]

Does anybody have the specifications of the, now obsolete, but once popular
dual gate MOSFET, the 40673?

Even better, can anybody recommend an easily obtainable modern replacement.

I want to make an oscillator with a low-Q LC resonator. Need gate2 to keep
it on the edge of oscillation by controlling the gain (voltage derived from
feedback loop).

Any ideas how I might achieve something similar using other than 2G mosfet?

Many thanks,

John Mitchell

 

[Peter A Forbes]

Does anybody have the specifications of the, now obsolete, but once popular
dual gate MOSFET, the 40673?

Even better, can anybody recommend an easily obtainable modern replacement.

I want to make an oscillator with a low-Q LC resonator. Need gate2 to keep
it on the edge of oscillation by controlling the gain (voltage derived from
feedback loop).

Any ideas how I might achieve something similar using other than 2G mosfet?

Many thanks,

John Mitchell

Plenty of stuff on Google, including this url which mentions a 40673 substitute
in the Google preamble:

http://www.ibiblio.org/pub/academic/agriculture/agronomy/ham/QRP/20030721.qrp.v02_n988

Peter

 

[Watson A.Name \Watt Sun - the Dark Remover\]

Does anybody have the specifications of the, now obsolete, but once popular
dual gate MOSFET, the 40673?

Even better, can anybody recommend an easily obtainable modern replacement.

I want to make an oscillator with a low-Q LC resonator. Need gate2 to keep
it on the edge of oscillation by controlling the gain (voltage derived from
feedback loop).

Any ideas how I might achieve something similar using other than 2G mosfet?

Many thanks,

John Mitchell

See the substitute or replacement stuff at www.nteinc.com

 

[John Woodgate]

I read in sci.electronics.design that John Mitchell

Does anybody have the specifications of the, now obsolete, but once popular
dual gate MOSFET, the 40673?

Google gives Results 1 - 10 of about 229 for 40673 MOSFET. (0.41
seconds)

Even better, can anybody recommend an easily obtainable modern replacement.

Farnell lists three types as 'RF Dual gate', but, as is rife in their
on-line catalogue, the specifications are seriously wrong - the currents
are probably milliamps, not amps! Can you imagine a 200 mW device with
an Id cont of 18 A?

BF988A
VISHAY SILICONIX
Farnell InOne MOSFET, N RF DUAL GATE SOT-103; Case style: SOT-103;
Current, Id cont: 10.5A; Marking, SMD: BF988; Power, Pd: 200mW @ 25°C;
Transistor polarity: N Channel; Transistor type: MOSFET; Voltage, Vds
max: 12V; Application code

SOT-103 10.5 A BF988 200 mW In Stock £0.73
1
663876 BF998
VISHAY SILICONIX
Farnell InOne MOSFET, N RF DUAL GATE SOT-143; Case style: SOT-143;
Current, Id cont: 18A; Marking, SMD: BF988; Power, Pd: 200mW @ 25°C;
Transistor polarity: N Channel; Transistor type: MOSFET; Voltage, Vds
max: 12V; Application code:

SOT-143 18 A BF988 200 mW In Stock £0.56
1
3164743 BF909R
PHILIPS
Farnell InOne MOSFET, N RF DUAL GATE SOT-143; Case style: SOT-143;
Power, Pd: 200W @ 25°C; Transistor polarity: N Channel; Voltage, Vds
max: 7V; Current, Id max: 40mA @ 25°C; Gfs, min: 36mA/V; Man; Available
until stocks are exhausted

SOT-143 200 W In Stock £0.173




For further help, Farnell InOne account holders can call the Technical
Support line on 0870 1200 202.

 

[Tim Wescott]

Does anybody have the specifications of the, now obsolete, but once popular
dual gate MOSFET, the 40673?

Even better, can anybody recommend an easily obtainable modern replacement.

I want to make an oscillator with a low-Q LC resonator. Need gate2 to keep
it on the edge of oscillation by controlling the gain (voltage derived from
feedback loop).

Any ideas how I might achieve something similar using other than 2G mosfet?

Many thanks,

John Mitchell

If you're not happy with any of the suggested replacements, you can use
a JFET and increase the negative gate bias, either by lowering the gate
voltage below ground or by raising the source return above it.

The same principal applies if you want to use a bipolar transistor, but
they seem to be a bit more difficult to make into good oscillators.

 

[John Larkin]

Farnell InOne

What's this InOne stuff? Newark Electronics is now Newark InOne.

What's an inone? Is that an acronym for "inventory = none"?

John

 

[Jim Thompson]

What's this InOne stuff? Newark Electronics is now Newark InOne.

What's an inone? Is that an acronym for "inventory = none"?

John

Probably ;-)

...Jim Thompson

 

[Winfield Hill]

John Larkin wrote...

What's this InOne stuff? Newark Electronics is now Newark InOne.

It's all the interconnected Farnell-owned companies, including
Newark, MCM, etc. This means you can easily order parts from
Farnell's UK inventory from your local Newark office. At no
extra cost, so far as I've been able to discover. You do need
the right part number (e.g., get it online from Farnell's UK web
site), but it's still pretty cool.

Thanks,
- Win

(email: use hill_at_rowland-dot-org for now)

 

[Watson A.Name \Watt Sun - the Dark Remover\]

John Mitchell wrote:

If you're not happy with any of the suggested replacements, you can use
a JFET and increase the negative gate bias, either by lowering the gate
voltage below ground or by raising the source return above it.

The same principal applies if you want to use a bipolar transistor, but
they seem to be a bit more difficult to make into good oscillators.

TYhat's odd. I'd say the opposite was true. And judging from industry,
they seem to think so, too.

 

[Tim Wescott]

TYhat's odd. I'd say the opposite was true. And judging from industry,
they seem to think so, too.

My experience is with one-off variable-frequency oscillators at fairly
low frequencies. From about 1MHz to 50ish the JFET seems to be easier
(at least for a one-off). At higher frequencies, and with crystal
oscillators, the bipolar seems to win out.

I've always been able to just throw some components around a MPF-102 or
a 2N4416 and get a working oscillator, whereas I need to actually _work_
to get it going well with a bipolar. That's not to say that a BJT isn't
better when you take all the production issues into account.

 

[Mark Zenier]

Does anybody have the specifications of the, now obsolete, but once popular
dual gate MOSFET, the 40673?

Yup. Apparently it's a consumer grade version of the 3N187

see

ftp://ftp.eskimo.com/u/m/mzenier/RCA40673.pdf and
ftp://ftp.eskimo.com/u/m/mzenier/RCA3N187.pdf

Mark Zenier [email protected] Washington State resident

 

[Watson A.Name \Watt Sun - the Dark Remover\]

Watson A.Name "Watt Sun - the Dark Remover" wrote:

My experience is with one-off variable-frequency oscillators at fairly
low frequencies. From about 1MHz to 50ish the JFET seems to be easier
(at least for a one-off). At higher frequencies, and with crystal
oscillators, the bipolar seems to win out.

I've always been able to just throw some components around a MPF-102 or
a 2N4416 and get a working oscillator, whereas I need to actually _work_
to get it going well with a bipolar. That's not to say that a BJT isn't
better when you take all the production issues into account.

Perhaps that's the main reason why most mfg'rs make them out of BJTs:
the very wire tolerance of FETs makes them difficult to use in
production. Probably necessitates selecting FETs for a certain Ids current.

In any case, the one fault I find with FETs is that they are often used
in front ends where lightning damage can occur. It's more difficult to
zap a BJT.

 

[Terry Pinnell]

I found a little to get you started:

Max ratings: Pd = 300mW, Vds = 20V, Vgs = 6V
Max gate reverse current = 50nA
Typical Forward Transfer Admittance = 12,000 umhos
Max input capacitance = 6pF

Spec also looks similar to 3N187, 3N200,40819, 40820, 40821, 40822,
40823, which I suspect are equally old/obsolete? (all those have same
base pin layouts; viewed from below: [tag], d, g2, g1, s.

 

[Tim Wescott]

Perhaps that's the main reason why most mfg'rs make them out of BJTs:
the very wire tolerance of FETs makes them difficult to use in
production. Probably necessitates selecting FETs for a certain Ids
current.

In any case, the one fault I find with FETs is that they are often used
in front ends where lightning damage can occur. It's more difficult to
zap a BJT.

Are you speaking of circuits out of discrete components or IC's?
Certainly if you've got a nice analog IC process that's all BJT you're
not going to want to stick a JFET in there!

I'm currently on a quest to find a lowest-component-count oscillator
that can generate a crystal controlled 144 MHz, on about a 3V rail. Got
any suggestions?

 

[John S. Dyson]

Perhaps that's the main reason why most mfg'rs make them out of BJTs:
the very wire tolerance of FETs makes them difficult to use in
production. Probably necessitates selecting FETs for a certain Ids current.

In any case, the one fault I find with FETs is that they are often used
in front ends where lightning damage can occur. It's more difficult to
zap a BJT.

A significant reason for using a PROPER BJT instead of a FET for an
oscillator would be the 1/F noise characteristics and general LF noise
characteristics. FETS tend to have
higher 1/F noise than BJTS, esp those FETS that work well at UHF or
above. On the other hand, an SiGe BJT tends to have good UHF characteristics,
but the low Rb and high beta give them some general benefit for LF noise.
Both SiGe and Si BJTs tend to be better in 1/F characteristics, while SiGe
are really good.

John

 

[L. Fiar]

I found a little to get you started:

Max ratings: Pd = 300mW, Vds = 20V, Vgs = 6V
Max gate reverse current = 50nA
Typical Forward Transfer Admittance = 12,000 umhos
Max input capacitance = 6pF

Spec also looks similar to 3N187, 3N200,40819, 40820, 40821, 40822,
40823, which I suspect are equally old/obsolete? (all those have same
base pin layouts; viewed from below: [tag], d, g2, g1, s.

Hi.

It is listed in an old Towers FET Selector book.

They also give a recommended substitute:
3N209.


LF.