Thanks, Arnaud. Here our goal was to include a computer interface for logging, and the Phidget was convenient for that; unfortunately, it only measures up to 75 millivolts, which creates problems of its own as you point out. Phidgets also have a purpose-built voltmeter, but its resolution is only 0.073v, which we didn't feel was sufficient. If I could find another fairly high-resolution voltmeter with a computer interface I could design to it. Of course there are plenty of $30 multimeters with RS-232 outputs, but I don't know if I trust them over a range and the resulting solution is somewhat clunky, e.g. bulky, among other things.
Jeff On Sun, Oct 28, 2012 at 9:17 AM, Arnaud Kodeck <[email protected]>wrote: > You should use a higher resistance value. I use 0.39 Ohm to measure 0 - > 4A range. Which lead to 1.6 V at resistance when a current of 4A is passing > through it. Then it becomes easier to protect over current. The resistance > must be able to handle a power of 7W. There is no need here for a precision > resistance, but a constant resistance value over temperature. Measure the > resistance value with an accurate/precision multimeter to convert volt into > ampere.**** > > ** ** > > Arnaud**** > ------------------------------ > > *From:* Jeff Berkowitz [mailto:[email protected]] > *Sent:* dimanche 28 octobre 2012 16:24 > *To:* [email protected] > *Subject:* Re: [Vo]:Home made data logging ammeter**** > > ** ** > > Thanks Arnaud.**** > > ** ** > > The goal of the diodes is to protect the board - it cost more than the > resistors did. But yeah, I know.**** > > ** ** > > Your point is especially true because those 1N5400s are odd - they list a > forward voltage around 1.2v, rather than 0.6v as you would expect for a > typical silicon diode. I guess they have two junctions in series?**** > > ** ** > > Otoh, they are spec'd to carry up to 200 amps forward for half a cycle > (8ms)**** > > ** ** > > You can't protect semiconductors with MOVs or fuses either ... too slow.** > ** > > ** ** > > Just a question of curiosity, but is there a good answer?**** > > ** ** > > Jeff**** > > On Sun, Oct 28, 2012 at 3:44 AM, Arnaud Kodeck <[email protected]> > wrote:**** > > Nice job !**** > > **** > > But let me explain that your protection diodes aren’t effective for > current overrun on sensing resistances. As you said, the voltage drop on > forward current of the diode is around 1V maybe a bit less … let’s say > 500mV. At 500mV, the resistance will dissipate (U^2/R) (0.5)^2/0.05 = 5 W. > At 1V, the power dissipated at resistance is 20W! So the resistance will > burn even if the diodes are closed, because maximum power rating of the > resistance is 1W.**** > > **** > > To have 500mV at the resistance, current flow will be 10A. Diodes are only > 3A …**** > > **** > > The protection will be effective only for non repetitive small spikes, but > rather more to protect your phidgets.**** > > **** > > Arnaud**** > ------------------------------ > > *From:* Jeff Berkowitz [mailto:[email protected]] > *Sent:* dimanche 28 octobre 2012 02:42 > *To:* [email protected] > *Subject:* [Vo]:Home made data logging ammeter**** > > **** > > Hi all, I built this USB-connected ammeter so we could log current flow > accurately and rapidly while doing electrolysis and also electroplating.** > ** > > **** > > http://pdxlenr.blogspot.com/2012/10/a-pretty-good-data-logging-ammeter.html > **** > > **** > > Jeff**** > > **** > > ** ** >
