need FET, P-ch, < 10nA off current @ 13Vds from 0 to 60C

[[email protected]]

does this exist? I need the fet to act as a very open switch when off
so a very sensitive circuit breaker does not trip. The circuit breaker
remains tripped unless it sees less than 30na to the load.

 

[Winfield]

acannell wrote:

Subject: need FET, P-ch, < 10nA off current @ 13Vds from 0 to 60C

does this exist? I need the fet to act as a very open switch when
off so a very sensitive circuit breaker does not trip. The circuit
breaker remains tripped unless it sees less than 30nA to the load.

You may be OK with most MOSFETs. I've measured leakage currents at
room temperature of under 10 to 30pA for many MOSFETs, IIRC. With
a current-doubling every 10C, that would translate into under say
0.5nA at 60C. But you won't find this performance spelled out on an
official datasheet. Have you taken any measurements? Oh, depending
on the time-scales, etc., you may have to worry about capacitance.

 

[[email protected]]

acannell wrote:

Subject: need FET, P-ch, < 10nA off current @ 13Vds from 0 to 60C


You may be OK with most MOSFETs. I've measured leakage currents at
room temperature of under 10 to 30pA for many MOSFETs, IIRC. With
a current-doubling every 10C, that would translate into under say
0.5nA at 60C. But you won't find this performance spelled out on an
official datasheet. Have you taken any measurements? Oh, depending
on the time-scales, etc., you may have to worry about capacitance.

I guess I just got unlucky? The first two mosfets I picked were the
IRF4905 and SU65P04, the first has 25uA zero Vgs drain to source
current, the second has 1uA. Are you perhaps referring to gate
current?

 

[[email protected]]

acannell wrote:

Subject: need FET, P-ch, < 10nA off current @ 13Vds from 0 to 60C


You may be OK with most MOSFETs. I've measured leakage currents at
room temperature of under 10 to 30pA for many MOSFETs, IIRC. With
a current-doubling every 10C, that would translate into under say
0.5nA at 60C. But you won't find this performance spelled out on an
official datasheet. Have you taken any measurements? Oh, depending
on the time-scales, etc., you may have to worry about capacitance.

Oh wait I thought you meant the datasheet would only show leakage
current at one temperature and not the doubling per 10C, but you meant
the datasheet will show worst case leakages and not the actual
leakages, which are orders of magnitudes smaller, am I correct?

 

[Joerg]

Oh wait I thought you meant the datasheet would only show leakage
current at one temperature and not the doubling per 10C, but you meant
the datasheet will show worst case leakages and not the actual
leakages, which are orders of magnitudes smaller, am I correct?

Datasheets are really, really conservative in that respect. For example
the BSS84 states 100nA max at 25C but I have found it to be much lower:
http://www.nxp.com/acrobat_download/datasheets/BSS84_3.pdf

I don't think there is much of a selection beyond that except maybe for
some fancy boutique devices. You may have to hand-pick or contract that
out but calling a few mfgs to find the best contenders can't hurt.

 

[Winfield]

Oh wait I thought you meant the datasheet would only show leakage
current at one temperature and not the doubling per 10C, but you
meant the datasheet will show worst case leakages and not the actual
leakages, which are orders of magnitudes smaller, am I correct?

Yes, I meant, go measure actual devices. Forget the datasheet;
do your own characterization. For a production design, create
a test fixture and add the tests to the manufacturing rules.
However in that case, wise up and understand your part well.
Understand the production spreads, and a half-year of supply
in advance, so you can weather a crash in the fab processes
while you select and get in stock an alternate part.

 

[Tony Williams]

does this exist? I need the fet to act as a very open switch when
off so a very sensitive circuit breaker does not trip. The
circuit breaker remains tripped unless it sees less than 30na to
the load.

You could use a second switch to help any leakage prob.


1. P-FET upstream of the current limiter.
____________ _____________
| | | |
13Vdc -----|Series P-FET|----+---|Current Limit|--->Load
|____________| | |_____________|
______|_____
| |
|Shunt Switch|
|____________|
|
0v----------------------------+--------------------->



2. P-FET downstream of the current limiter.
_____________ ____________
| | | |
13Vdc--+--|Current Limit|----+---|Series P-FET|--->Load
| |_____________| | |____________|
| _______________ |
| | | |
+--|Shorting Switch|--'
|_______________|

0v------------------------------------------------>