Maker Pro
Login Join Maker Pro
Or sign in with

Menu

Connect With Us

Network

Maker Pro

Do MOFETS have a Rdson that goes up for very low curents?

J

Jan Panteltje

Do MOFETS have a Rdson that goes up for very low curents?
I have been measuring some IRFZ44A, the data sheet says
Rds-on is constant from say 10A to 100A.

I wanted to use the voltage drop to measure the current in a circuit.
But if I measure

At 310 mA 8.3 mV drop, makes 26.7 mOhm
at 16.2 A 250 mV drop, makes 15.4 mOhm
That is about double at very low currents.

That last Rds-on is about what the data sheet says (I use 12V Vgs).


Is this inherent to MOSFETS, very low current = higher Rds-on?
 
M

MooseFET

Do MOFETS have a Rdson that goes up for very low curents?
I have been measuring some IRFZ44A, the data sheet says
Rds-on is constant from say 10A to 100A.

I wanted to use the voltage drop to measure the current in a circuit.
But if I measure

At 310 mA 8.3 mV drop, makes 26.7 mOhm
at 16.2 A 250 mV drop, makes 15.4 mOhm
That is about double at very low currents.

That last Rds-on is about what the data sheet says (I use 12V Vgs).

Is this inherent to MOSFETS, very low current = higher Rds-on?
Click to expand...


MOSFETs have their lowest resistance at zero current. IGBTs have what
looks like a diode's voltage vs current characteristic.
 
T

Tim Wescott

Measuring error.
Click to expand...

Indeed. I suggest you get a length of wire of approximately the same
resistance, and repeat your measurement. Assuming that the temperature
change is insignificant its resistance should stay constant -- if you see
a _wire's_ resistance changing with current then it's _definitely_
measurement error.

--
Tim Wescott
Control systems and communications consulting
http://www.wescottdesign.com

Need to learn how to apply control theory in your embedded system?
"Applied Control Theory for Embedded Systems" by Tim Wescott
Elsevier/Newnes, http://www.wescottdesign.com/actfes/actfes.html
 
J

Jan Panteltje

Then how do you explain this measurement?
Click to expand...


OK, found it, looks like some strong RF messed up the 16A measurement.
Just measured a whole lot of values again, now I get better values,
and indeed slightly lower at lower current (like .025 Ohm versus .028 Ohm),
but there still is this RF influence, so anyways, I just need an indication,
maybe 10 percent accuracy.
Thank you for the help, so I measured again :)
 
J

Joerg

Jan said:
OK, found it, looks like some strong RF messed up the 16A measurement.
Just measured a whole lot of values again, now I get better values,
and indeed slightly lower at lower current (like .025 Ohm versus .028 Ohm),
but there still is this RF influence, so anyways, I just need an indication,
maybe 10 percent accuracy.
Thank you for the help, so I measured again :)
Click to expand...

Invest in some ferrite beads and a few larger toroids, big enough to go
over the probe cables.
 
J

Jan Panteltje

Invest in some ferrite beads and a few larger toroids, big enough to go
over the probe cables.
Click to expand...

Probe? What probe ;-)?
I have some extra analog input channels now on the SWR meter PIC.
RF power, reflected, forward, battery voltage, and now also current
(still have to dream up a robust interface to grab the few millivolts),
and the rest like RS232 messages (it also shows inside and outside
temperature send from teh PC, but actuall yi want one more analog channel
and that is ouput transistor temp of the transmitter.
There is also a SWR alarm, and a temp alarm seems a good idea.

What I did is have the RF part (the actual SWR transformer) in a separate
metal box, that box outputs an analog DC voltage proportional to power,
forward and reflected voltages, and is connected to the display unit via
a long screened cable.
The MOSFET DC power switch is in an other place, with a big heatsink
(that stays cold actually), and I have to grab the few hundred millivolt
from there.
I noticed that 10 cm wire from PIC analog input to decoupling caps is enough to
cause weird values to be measured.
Now I have soldered some 100nF caps directly on the back of the PIC on the pins...
That helps.
 
J

Joerg

Jan said:
On a sunny day (Sun, 03 Feb 2008 22:12:11 GMT) it happened Jan Panteltje

Like this:
http://panteltje.com/panteltje/pic/swr_pic/LCD_4x20_2.jpg
Click to expand...


Nice! Yet I still prefer my old analog SWR meter, also home-built. It
doesn't show the outside temperature though.

If you see RF getting in then it must find a BJT or some other diode
path. RF tends to find this stuff like ants will find the sugar no
matter how good you hide it.
 
J

Jan Panteltje

I would expect it would be difficult to hold both ends of the sensor
to that degree of precision.
Click to expand...

Sure, but the sensors are LM135, one is outside, hanging 1 m above ground
from a about one meter long cable, that cable has the same temp as 'outside',
the other is inside, also with a cable at the same temp as the room :)

But you are right, although you can calibrate LM135 with a pot, the temp
came via an 8 bit adc (PCF8961 i2c), and, as there are no ranging amps,
the display will jump from for example 19.07 to 21.09.
But it looks good :)
That area of the LCD is filled from the PC, the PC reads the ADC.
That LCD message area, updated via serial port, is destined for other things,
if no monitor is on, it can display temp, or 'you have email',
or 'alarm the aliens have landed'.

# assuming com parameters set
while [ 1 ]
do
inside=`/bin/cat /root/.hcs/temperatures | /usr/bin/awk '{print $1}'`
outside=`/bin/cat /root/.hcs/temperatures | /usr/bin/awk '{print $2}'`

/bin/echo -e "\0001I $inside\0337C O $outside\0337C\r" > /dev/ttyS1
sleep 10

done



Does that answer your question?
 
J

Jan Panteltje

Nice! Yet I still prefer my old analog SWR meter, also home-built. It
doesn't show the outside temperature though.

If you see RF getting in then it must find a BJT or some other diode
path. RF tends to find this stuff like ants will find the sugar no
matter how good you hide it.
Click to expand...

Yes, I added the current measurement hardware now, a LM324 with
a gain of 1.955, so measuring across a 25mOhm MOSFET as shunt gives me
1023 steps with 5V reference for 102.3 A, so all I had to do was add a dot
to the ADC output to get quite accurate amperes :)
Split up some resistors and 100nF some caps to ground and the 2 wire
lead did not even notice the RF, it is 1) common mode, and 2) filtered
low pass 8k2 100nF.
The asm listing is available now, likely final version?
http://panteltje.com/panteltje/pic/swr_pic/index.html
Worked really nice tonight.
Now I need to order a nice box, found a metal one, and make a real PCB.
 
J

Joerg

Jan said:
Yes, I added the current measurement hardware now, a LM324 with
a gain of 1.955, so measuring across a 25mOhm MOSFET as shunt gives me
1023 steps with 5V reference for 102.3 A, so all I had to do was add a dot
to the ADC output to get quite accurate amperes :)
Click to expand...


Careful, the LM324 is a BJT amp, not so good if strong RF fields are around.

Split up some resistors and 100nF some caps to ground and the 2 wire
lead did not even notice the RF, it is 1) common mode, and 2) filtered
low pass 8k2 100nF.
The asm listing is available now, likely final version?
http://panteltje.com/panteltje/pic/swr_pic/index.html
Worked really nice tonight.
Now I need to order a nice box, found a metal one, and make a real PCB.
Click to expand...

For my wattmeter I made my own box from wood. Didn't have a lot of money
as a teenager so it wasn't very noble wood. But, numerous applications
of stain and several thin layers of marine lacquer later it looked
really posh. Most of my stuff back then went into such enclosures and
they kind of matched each other. Best of all it still works. It's an old
photo, not very good:

http://www.analogconsultants.com/ng/images/oak2.jpg

This one goes to two kilowatts.
 
E

ehsjr

Joerg said:
Careful, the LM324 is a BJT amp, not so good if strong RF fields are
around.



For my wattmeter I made my own box from wood. Didn't have a lot of money
as a teenager so it wasn't very noble wood. But, numerous applications
of stain and several thin layers of marine lacquer later it looked
really posh. Most of my stuff back then went into such enclosures and
they kind of matched each other. Best of all it still works. It's an old
photo, not very good:

http://www.analogconsultants.com/ng/images/oak2.jpg

This one goes to two kilowatts.
Click to expand...

Nice!

Ed
 
J

Jan Panteltje

Careful, the LM324 is a BJT amp, not so good if strong RF fields are around.
Click to expand...

Yes, true.
For my wattmeter I made my own box from wood. Didn't have a lot of money
as a teenager so it wasn't very noble wood. But, numerous applications
of stain and several thin layers of marine lacquer later it looked
really posh. Most of my stuff back then went into such enclosures and
they kind of matched each other. Best of all it still works. It's an old
photo, not very good:

http://www.analogconsultants.com/ng/images/oak2.jpg
Click to expand...

Joerg that is really beautiful!
Timeless design.

It reminded me of a picture I did see fom the Russian space centre,
everything had knobs and meters like that .
Probably more reliable then all the NASA *electronics*.
 
J

Joerg

Jan said:
Yes, true.


Joerg that is really beautiful!
Timeless design.

It reminded me of a picture I did see fom the Russian space centre,
everything had knobs and meters like that .
Probably more reliable then all the NASA *electronics*.
Click to expand...

Well, you can still experience it. AFAIK one guy at our local airstrip
here wants to see his Nanchang, similar to this one:

http://en.wikipedia.org/wiki/Image:Nanchang_CJ-6_warbird.jpg

Lots of "real" instruments in there.
 
You must log in or register to reply here.

Similar threads

Martin Rodriguez
Who knows how to test N-Ch. MOFETS with Multimeter? Please HELP!
Replies
7
Views
2K
Audioguru
Audioguru
K
Low RDSon Logic CMOS Gate
2
Replies
28
Views
2K
Fred Bartoli
F
H
Mosfet 7.5% increase in RDSON between 25C and 100C
Replies
6
Views
1K
Tim Williams
T
J
Push-Pull driver <<1ohm Rdson?
Replies
15
Views
2K
Joerg
J
J
Philips Monitor 230E1HSB/27 - Screen goes black seconds after working
Replies
2
Views
586
Harald Kapp
Harald Kapp
Top