Hi Jack, I have created some diagrams to help communicate the setups that I am going to describe. It on my Google drive at:
https://drive.google.com/file/d/0B5Pc25a4cOM2MTlIX1pwMC1PdHc/edit?usp=sharing These setups presume that when you measure between the high current bars and the pins of the 120VAC input plug, that in all cases you measure an OPEN circuit (infinite resistance). If this is not the case, then we need to re-think the setups ... but it should be the case. Referring to the set of diagrams in the file above, the setups are described as follows: Setup 1: This is to measure the peak voltage out of the welder during a spot weld. Do this with a voltmeter and the circuit shown. The voltage you measure will be the AC peak voltage. The actual voltage that is present will be about +/- (the measured voltage + about 1V). This measurement will be done without the oscilloscope so as to see what voltage is coming out of the welder to insure that your oscilloscope can handle the voltage range. Setup 2: Measure the resistance of the welding bar as a current shunt. Even though the actual current will be AC, the resistance for a current shunt can be measured with DC. The actual current that is required in the measurement is not critical as long as the value supplied (I_BC) is known. The current should be the max. the source can provide to get best accuracy. So, put your supply into current limit mode and crank up the current limit until the max for the supply is reached. Measure the voltage (V_BA) across the points B-A. The resistance will be R=V_BA /I_BC . This is your current shunt resistance. Setup 3: Measure the spot welding waveforms. Use a heavy wire to connect from point B to the oscilloscope ground terminal. No current should be flowing in this conductor, but you want its resistance to be lower than the resistance from the probes shields to that ground point. That way if you get an anomalous ground current, it won't flow through the probes. In fact, I would begin this test with no ground connections for the probes. In fact, once you get a handle on the voltages, you will probably not want to use 10x probes, you will want a straight through connection. The voltage sources you are measuring are extremely low impedance and can easily drive the low impedance of the oscilloscope without the 10x probe. You will get cleaner signals without the 10x probes, but I would measure with the probe first. The voltage you measure in channel 1 as shown will be the voltage as a function of time. I would trigger on this voltage's rising edge. The voltage on channel 2 is the NEGATIVE of the voltage across the shunt resistor. When calculating power, you will need to multiply this trace by -R to get the current vs. time. This was done to avoid the need for signal subtraction in only a 2-channel scope. It may be necessary, as I mentioned in a previous post to create a compensating loop to subtract out current induced error voltages in these measurements. Think of that as a possible future improvement in accuracy of what you are measuring. Bob On Sun, Aug 31, 2014 at 6:57 AM, Jack Cole <[email protected]> wrote: > Bob, > > I'm getting ready to work on implementing what you suggested. > > Could you take a look at this sketch to see if this is what you are > suggesting for hooking up the oscilloscope? > > > http://www.lenr-coldfusion.com/wp-content/uploads/2014/08/power-measurement.png > > I won't be able to do 10 amps for calibration, but I can do anything up to > 5 amps with my lab power supply. > > Best regards, > Jack >
