Minimum practical Rds of JFET?
- Thread starter Ribbledox
- Start date
Help Support GroupDIY Audio Forum:
AMZ-FX
Well-known member
The Fairchild J111 is rated at 30 ohms on-resistance. That's about the lowest that I know about though there could possibly be lower ones. If that's too low, the J113 is 100 ohms.
regards, Jack
regards, Jack
Svart
Well-known member
hmm that is interesting, i'm having a hard time finding Rds(on) for JFETS.
for MOSFETS, N-ch it's not uncommon to have Rds(on) of fractions of an ohm. the irf640 that i use a lot of has an Rds(on) of .15ohm.
P-ch are usually a little higher, the irf9640 is .50 ohm.
I also have MOSFETS here that have Rds(on) of .082 ohms and .029 ohms.
I think 30 and 300 ohms sounds a little high but i've been looking at datasheets for 15 minutes an NONE have the Rds(on) of a JFET.
I guess you'll just have to test the specific JFET you want to use and see.
:guinness:
for MOSFETS, N-ch it's not uncommon to have Rds(on) of fractions of an ohm. the irf640 that i use a lot of has an Rds(on) of .15ohm.
P-ch are usually a little higher, the irf9640 is .50 ohm.
I also have MOSFETS here that have Rds(on) of .082 ohms and .029 ohms.
I think 30 and 300 ohms sounds a little high but i've been looking at datasheets for 15 minutes an NONE have the Rds(on) of a JFET.
I guess you'll just have to test the specific JFET you want to use and see.
:guinness:
AMZ-FX
Well-known member
Mosfets and jfets are quite different animals... a big advantage that mosfets have for switching applications is their low on-resistance as compared to jfets. Another is their power handling capacity... there aren't many power jfets! :grin:
regards, Jack
regards, Jack
Thanx!
But what I was asking was more related to what factors determine the minimal Rds (on) of JFETs.
It was mentioned somewhere else that the load impedance of the JFET plays a part in determining the value of the minimum Rds (on).
But what I was asking was more related to what factors determine the minimal Rds (on) of JFETs.
It was mentioned somewhere else that the load impedance of the JFET plays a part in determining the value of the minimum Rds (on).
Svart
Well-known member
yes you are correct AMZ, i was just trying to show a few examples to compare to since i couldn't find much datasheet info on JFETs.
as for factors determining Rds(on).. Rds(on) is the resistance to current flow through the JFET semiconductor when the gate is considered fully ON. the only factors that determine it are of the build of the JFET itself, like die size, material purity, topology considerations(like N-CH or P-CH), and probably much more like lead material and so forth.
Do you mean the current capability of the JFET? much of the same applies since it's all related via ohm's law.
as for factors determining Rds(on).. Rds(on) is the resistance to current flow through the JFET semiconductor when the gate is considered fully ON. the only factors that determine it are of the build of the JFET itself, like die size, material purity, topology considerations(like N-CH or P-CH), and probably much more like lead material and so forth.
Do you mean the current capability of the JFET? much of the same applies since it's all related via ohm's law.
Hmm...I see...
Because I came over an article were they were testing two different JFET's, one with low drain current (same as high load impedanc, I gues) and a second with a higher drain current. And the second one turned out to have a much lower Rds (on).
So I though that perpahs the value of drain current (load impedance) plays a part in how low the Rds (on) may reach.
Also, the second JFET also had a much lower pinch-off point.
Cheers!
Because I came over an article were they were testing two different JFET's, one with low drain current (same as high load impedanc, I gues) and a second with a higher drain current. And the second one turned out to have a much lower Rds (on).
So I though that perpahs the value of drain current (load impedance) plays a part in how low the Rds (on) may reach.
Also, the second JFET also had a much lower pinch-off point.
Cheers!
G
Guest
Guest
[quote author="Ribbledox"]
Because I came over an article were they were testing two different JFET's, one with low drain current (same as high load impedanc, I gues) and a second with a higher drain current. And the second one turned out to have a much lower Rds (on).
[/quote]
Yes, you can parametrize JFETs by Rds value, but this is somewhat
ineffective.
In normal MOSFETs (with no technological chanell), you put voltage
to the gate and then you make chanel. And (in triode region) you
have some small resistance. This resistance is dependent somewhat
to gate voltage, but for gate voltage large, dependence is small.
Then you can read from catalogue (for parameter = gate-source voltage)
Rds ON. It is parameter of triode - region of the MOSFET with purely
induced chanell.
In JFET what is minimal triode-region resistancy ?
It is when gate - source voltage is zero.
You can read from catalogue Id (at Vgs=0) It is Idss, or saturated
drain current.
You can read from catalogue Uds (saturation voltage of drain
at Vgs = 0) (Udss)
And if use Ohm s law, you get Rdson for Jfet:
Rdson=Udss/Idss
Every what you need is good catalogue. And eyes for read it from
graphs.
It is the point on the output characteristic graph (for Vgs=0) where line turns.
xvlk
Because I came over an article were they were testing two different JFET's, one with low drain current (same as high load impedanc, I gues) and a second with a higher drain current. And the second one turned out to have a much lower Rds (on).
[/quote]
Yes, you can parametrize JFETs by Rds value, but this is somewhat
ineffective.
In normal MOSFETs (with no technological chanell), you put voltage
to the gate and then you make chanel. And (in triode region) you
have some small resistance. This resistance is dependent somewhat
to gate voltage, but for gate voltage large, dependence is small.
Then you can read from catalogue (for parameter = gate-source voltage)
Rds ON. It is parameter of triode - region of the MOSFET with purely
induced chanell.
In JFET what is minimal triode-region resistancy ?
It is when gate - source voltage is zero.
You can read from catalogue Id (at Vgs=0) It is Idss, or saturated
drain current.
You can read from catalogue Uds (saturation voltage of drain
at Vgs = 0) (Udss)
And if use Ohm s law, you get Rdson for Jfet:
Rdson=Udss/Idss
Every what you need is good catalogue. And eyes for read it from
graphs.
It is the point on the output characteristic graph (for Vgs=0) where line turns.
xvlk
Svart
Well-known member
well they are related due to ohms law because given a sp[ecific voltage and current, you will see a certain resistance(DC) or impedence(AC).
think of it like this. one JFET which can handle more current for a given voltage will of course show lower impedence, likely because it was designed for better current handling with a larger die, better materials and/or a more efficient design.
the inverse is true for a JFET that can handle less current, a smaller die, lesser quality, and/or less effiecient design is trying to handle the same voltage at the same current as the other JFET and shows more resistance. much like a thinner wire will heat more than a thicker wire at a specific voltage/current.
you were on the right track, but the Rds(on) is one factor determining the drain current, not the other way around.
the three factors that are usually the most important to (J)FETs are:
Vdss: max voltage
Rds(on): max resistance/impedence you will see in the fully on state
Id: max continuous current rating (Idm: max pulsed current)
:guinness:
think of it like this. one JFET which can handle more current for a given voltage will of course show lower impedence, likely because it was designed for better current handling with a larger die, better materials and/or a more efficient design.
the inverse is true for a JFET that can handle less current, a smaller die, lesser quality, and/or less effiecient design is trying to handle the same voltage at the same current as the other JFET and shows more resistance. much like a thinner wire will heat more than a thicker wire at a specific voltage/current.
you were on the right track, but the Rds(on) is one factor determining the drain current, not the other way around.
the three factors that are usually the most important to (J)FETs are:
Vdss: max voltage
Rds(on): max resistance/impedence you will see in the fully on state
Id: max continuous current rating (Idm: max pulsed current)
:guinness:
G
Guest
Guest
[quote author="PRR"]Rds(min) is nearly the same as 1/Gm(max).
Short fat FETs have low Rds(min). Many UHF FETs are built long and thin to reduce Gate capacitance, and may not be low Rds(min).[/quote]
... And maximum Gm is at Vgs = 0. Be care:
Maximum Gm is parameter of saturated (pentode) region,
Not small - signal triode. But Rdson is parameter of triode region.
Your "nearly the same" can mean "diferent twice".
xvlk
Short fat FETs have low Rds(min). Many UHF FETs are built long and thin to reduce Gate capacitance, and may not be low Rds(min).[/quote]
... And maximum Gm is at Vgs = 0. Be care:
Maximum Gm is parameter of saturated (pentode) region,
Not small - signal triode. But Rdson is parameter of triode region.
Your "nearly the same" can mean "diferent twice".
xvlk
> maximum Gm is at Vgs = 0.
Actually, you can go a little positive. JFETs don't draw grid current until many tenths of a volt "forward" bias.
> Be care: Maximum Gm is parameter of saturated (pentode) region, Not small - signal triode. But Rdson is parameter of triode region. Your "nearly the same" can mean "diferent twice".
Yes, you are correct. My quick answer is not correct for FETs in the triode range; Rds will be higher in the triode range. Thanks for the reminder.
However different FET designs, from fat switches to low-leakage amps, may have a 10:1 difference in Gm or Rds. When Rds isn't specified, Gm leads you to the better (for this use) devices and gives you an approximate idea of how much you can attenuate.
Actually, you can go a little positive. JFETs don't draw grid current until many tenths of a volt "forward" bias.
> Be care: Maximum Gm is parameter of saturated (pentode) region, Not small - signal triode. But Rdson is parameter of triode region. Your "nearly the same" can mean "diferent twice".
Yes, you are correct. My quick answer is not correct for FETs in the triode range; Rds will be higher in the triode range. Thanks for the reminder.
However different FET designs, from fat switches to low-leakage amps, may have a 10:1 difference in Gm or Rds. When Rds isn't specified, Gm leads you to the better (for this use) devices and gives you an approximate idea of how much you can attenuate.
G
Guest
Guest
[quote author="PRR"]
Actually, you can go a little positive. JFETs don't draw grid current until many tenths of a volt "forward" bias.
[/quote]
I am not sure, that little positive gate bias = higher Gm.
Overall characteristic of JFET is form of S-curve.
As in analogy between technological chanell mosfet,
where enhancement mode have lower transconductancy, there
must be inflex point of S- curve.
I may be possible, that this inflex point is an Vgs=0 ?????????????...
And that positive gate voltage results in higher drain current , but lower Gm.
xvlk
Actually, you can go a little positive. JFETs don't draw grid current until many tenths of a volt "forward" bias.
[/quote]
I am not sure, that little positive gate bias = higher Gm.
Overall characteristic of JFET is form of S-curve.
As in analogy between technological chanell mosfet,
where enhancement mode have lower transconductancy, there
must be inflex point of S- curve.
I may be possible, that this inflex point is an Vgs=0 ?????????????...
And that positive gate voltage results in higher drain current , but lower Gm.
xvlk
Sounds like perhaps they were talking about Idss, higher Idss current generally reflects lower Rds.Ribbledox said:Hmm...I see...
Because I came over an article were they were testing two different JFET's, one with low drain current (same as high load impedanc, I gues) and a second with a higher drain current. And the second one turned out to have a much lower Rds (on).
So I though that perpahs the value of drain current (load impedance) plays a part in how low the Rds (on) may reach.
Also, the second JFET also had a much lower pinch-off point.
Cheers!
Rds is the relaxed state (not pinched off) of the JFET's conductive channel so simple ohmic resistance of that path.
JR
Indeed you can enjoy a slightly lower Rds that way, but be careful to keep the gate diode cut off or that DC diode conduction current will flow into the DS channel likely corrupting DC voltage for gate/mute applications.PRR said:
JR
NOTE that this thread is 15 years old and the OP has not been back since.
But Merlin's interest is always valuable.
Rds is mostly about how BIG the FET die area is. This conflicts with low capacitance which matters in some applications; and mostly with *yield* thus cost. (If a wafer tends to have 100 defects, if you make 100 big FETs per wafer you get maybe 10 good FETs; if you make 1,000 small FETs per wafer you will surely get 900--- big FETs are bad economics.)
FETs are sold in two aisles: "Amplifiers" and "Switches". They may be the same dies under the part number, just a little better for one application or the other. Low Vgs is better for amplifiers, high Vgs is better for switches. Rds spec is usually only found on parts marketed as "switches". A very high Gm "amplifier" will be a low-Rds switch, but how low?
At Mouser, these through-hole low Rds JFETs are listed in stock:
IFN5432 - 5r
J108 - 8r
IFN5434 - 10r
J110 - 18r
2N4856, 2N4859 - 25r
Availability of SMD parts is left to those who can see them.
Some of the prices are punishing.
But Merlin's interest is always valuable.
Rds is mostly about how BIG the FET die area is. This conflicts with low capacitance which matters in some applications; and mostly with *yield* thus cost. (If a wafer tends to have 100 defects, if you make 100 big FETs per wafer you get maybe 10 good FETs; if you make 1,000 small FETs per wafer you will surely get 900--- big FETs are bad economics.)
FETs are sold in two aisles: "Amplifiers" and "Switches". They may be the same dies under the part number, just a little better for one application or the other. Low Vgs is better for amplifiers, high Vgs is better for switches. Rds spec is usually only found on parts marketed as "switches". A very high Gm "amplifier" will be a low-Rds switch, but how low?
At Mouser, these through-hole low Rds JFETs are listed in stock:
IFN5432 - 5r
J108 - 8r
IFN5434 - 10r
J110 - 18r
2N4856, 2N4859 - 25r
Availability of SMD parts is left to those who can see them.
Some of the prices are punishing.
merlin
Well-known member
- Joined
- Aug 17, 2011
- Messages
- 260
PRR said:
Thanks PRR I'll look into those! For those who are interested, here's the list of Vishay obsolescence. To me it looks like all their JFETs, including J108/109/110. This poses a serious problem to those of us who use thousands per year.
https://media.digikey.com/pdf/PCNs/Vishay/PCN-SIL-0052008%20Rev%202.pdf
