It's difficult to tell what the issues are without seeing a schematic of your circuits. First, if you have both generators hooked up to each other, I would take that apart and test each separately. Also, if you are charging any batteries or capacitors or anything, I would remove those and just try to make the circuit as simple as possible with an equivalent resistor (to your complete circuit), shunt resistor, and single generator. For the purposes of finding out how much energy you'll generate, this is actually a pretty good approximation. If it's still giving you bogus results with such a simple setup then I can't think of anything else to do about it, save it being a Vernier hardware issue. I also didn't quite understand what you meant by "there is no one negative ground." Granted electronics is more my hobby than my expertise, so hopefully someone can provide a better answer for you.

If the first setup tests out fine, then I would slowly reconnect components to your circuit until you can identify the step that kills your measurement. I realize that this is kind of a blind test, but without more information, I can't think of anything else to do.

If no-one else gives you a better answer, then maybe you can email ScienceLine with more detail about your setup and perhaps I can help. But some people that are more familiar with electronics issues may be able to give you a straight answer with the level of description you've provided.

I looked at the site to see your test setup, but I could not find a schematic of your test setup. (It does look like a fun project; you might try putting some galvanized sheet metal on the sides of the solar panels to increase the output and the acceptance angle of light, if you aren't going to track the sun. Since a lot of the light conversion in Si is near infrared, the reflectors don't need to be too good, as long as the cells don't get too hot.)

I am also guessing that you have one of two problems:
1. Common ground offsets from the sensors.
2. Induced ground loop in either the sensor or the connection to the PC.

Both issues cause problems in similar ways, and there are several solutions. First, where does the power to run the sensor come from? If it is a battery in the sensor, problem 1 is not your issue unless there is a common ground from the data to one side of the sensor. If it comes from the PC -- the problem gets more interesting.This is especially true if you have connections of any sort between the generators -- for example a common battery. Most data loggers use inexpensive A/D chips developed originally for hand multi-meters or panel displays -- these require decoupled power on their inputs versus their measurement terminals.

The second kind of problem comes from current coupled into your wires from nearby EM sources like power wiring, transformers, and overhead wiring. It is amazing how much current is coupled in this way -- my grandfather used to run drainage pumps by hooking to his barbed wire fences (which ran parallel to overhead power lines). The other side was the local irrigation pipe. My lab bench has a grounded sheet of steel under the Formica top, specifically to reduce the coupling-- yet I still get 100-400mV AC from 2' test leads. Since the cables you ran to the sensors are probably twisted and run on a common path, they couple to the same fields -- so you won't see a problem with 1 cable. However, running two can cause havoc, from power coupled into the sensors (if they have a virtual ground, the voltage will show up on the measured side), or coupled into the PC. (The PC can be destroyed this way --). In very high field areas like power substations, the best solution is wireless connections between the sensors and the logger.

To see what is going on, first use an inexpensive multi-meter to measure the AC voltage at the PC from the "ground" connections. Also look for DC offsets here -- that is a tell-tale of problem
1. Since these are 'data' lines, they should not be effected by connecting 1 of them, unless the grounds are coupling power.

There is a lot of information on the 'ground loop' problem on the net.Again, I don't know the format of your data cables, but there are relatively cheap isolating repeaters for most data formats -- i.e. USB, serial, parallel, 488... As a last tip -- don't make the assumption that you can ground out the currents -- it is usually easier to introduce a resistor or other isolator as the currents are proportional to the flux and inverse to the resistance. A bigger cable is usually a bigger problem. Hope this helps

I'm an electrical engineer who wires and fixes lab equipment, so I can help. I suspect you're on the right track, that there's a problem with a shared ground connection. Can you send me a diagram for how things are connected and how you are measuring your voltages? In particular, I would need to know which connections are tied to each other, and also which pieces have some connection to ground. If you could label the voltages you measure at each point, that would very helpful as well..

I can't access your link, and unfortunately, without knowing more of the setup (and not being an electrician), I can't tell you what the problem is, although it sounds to me like the two voltages are interfering with each-other in some way.

I CAN tell you, however, that the result of your experiment is going to depend on a number of factors other than location, among these things like time of year and the angle of your solar panel. For instance, does your windmill track the prevailing wind direction in order to get the most efficiency possible? Do your solar panels track the sun? Are you doing this in summer (more light) or winter (more wind)? By"location", do you include things like altitude, as there is going to be more wind at higher elevations?

If you are using the differential voltage probes from Vernier, then I don't know what is causing the problem.

However, if you are using the non-differential voltage probe, then I would guess that it is an issue of the common connection of the probes shorting out part of the circuit. The issue is described in the following technical note from Vernier's web site:

When you measure with two or more voltage probes, and one of the terminals of a voltage probe is grounded (as is the case with this probe, according to the Vernier web site), you need to be careful with those ground contacts of the voltage probes. Make sure that the ground contacts of all of the probes are connected to the same point in the circuit.

Please feel free to follow up if this doesn't solve the problem.