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?
MOSFETs have their lowest resistance at zero current. IGBTs have what looks like a diode's voltage vs current characteristic.
On a sunny day (Sun, 3 Feb 2008 07:28:27 -0800 (PST)) it happened MooseFET wrote in :
Then how do you explain this measurement?
Measuring error.
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
On a sunny day (Sun, 3 Feb 2008 07:28:27 -0800 (PST)) it happened MooseFET wrote in :
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 :-)
On a sunny day (Sun, 3 Feb 2008 08:38:25 -0800 (PST)) it happened Winfield wrote in :
You are so right, RF upset the meters :-)
Invest in some ferrite beads and a few larger toroids, big enough to go over the probe cables.
-- Regards, Joerg http://www.analogconsultants.com/
On a sunny day (Sun, 03 Feb 2008 21:50:27 GMT) it happened Joerg wrote in :
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.
On a sunny day (Sun, 03 Feb 2008 22:12:11 GMT) it happened Jan Panteltje wrote in :
Like this:
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.
-- Regards, Joerg http://www.analogconsultants.com/
Temperature to 0.01=B0C?
On a sunny day (Mon, 4 Feb 2008 14:53:57 -0800 (PST)) it happened Richard Henry wrote in :
Not enough?
e
I would expect it would be difficult to hold both ends of the sensor to that degree of precision.
On a sunny day (Tue, 5 Feb 2008 08:02:05 -0800 (PST)) it happened Richard Henry wrote in :
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?
On a sunny day (Mon, 04 Feb 2008 09:41:23 -0800) it happened Joerg wrote in :
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?
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:
This one goes to two kilowatts.
-- Regards, Joerg http://www.analogconsultants.com/
Nice!
Ed
On a sunny day (Tue, 05 Feb 2008 23:49:02 GMT) it happened Joerg wrote in :
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*.
Well, you can still experience it. AFAIK one guy at our local airstrip here wants to see his Nanchang, similar to this one:
Lots of "real" instruments in there.
-- Regards, Joerg http://www.analogconsultants.com/
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.