Probably the easiest way to test it is using the resistance setting, because this way you can disconnect the connector and applyt the meter probes directly to the pins on the TPS (Not on the cable). The alternative method is to measure the voltage coming from the TPS, however to do that you need to leave the connector attached and find a way to break into the cable to probe the wires, which is more difficult.
I can't recall the exact range of the TPS, I think it is something like 5 K Ohms. Basically, it is a variable resistance. There are 3 pins on the TPS. If you measure between the outermost pins, you will read the full value of the TPS resistance. There is a resistive carbon track between the two outer pins. The centre pin is connected to a wiper which as the throttle plate opens, is moved along the length of the carbon track. So at one end of the throttle pedal range you will read almost 0 K Ohms, and at the other you read almost 5K Ohms (assuming it is 5K Ohms). So you want to have the meter set to read resistance - most meters have different ranges such as 200Ohms, then say 2K Ohms, then maybe 200KOhms. Start by measuring the full resistance of the TPS between the 2 outer pins, and work down the ranges until you get to the lowest range which can still read the value without indicating overrange.
To measure the actual performance of the TPS, you want to measure between the centre pin and one of the outer pins. it doesn't really matter which outer pin - all that will differ between choosing either one is the direction of the changing resistance. One outer pin to centre pin will read high to low as the throttle is opened, the other outer pin to centre pin will read low to high as the throttle is opened.
Get someone to very gradually and steadily operate the accelerator through it's range (engine off of course). Alternatively, if you can get the meter probes securely attached to the pins somehow, or have your assistant hold them on the pins, then you can just operate the throttle by hand using the throttle linkage. This is actually the better method if possible, as it saves confused communications with said helper, and if it is your missus, it saves her getting annoyed when you ask if she is being smooth enough with her 'pedal action' 
What you are looking for is to see how the numbers change as the throttle is opened. In a good condition TPS, the value will start near the bottom or top of the range (depending on which outer pin you are measuring from) and then change smoothly, without jumping around, moving towards the other end of the range, and without any dead spots where the reading could go suddenly very high or even open circuit (overrange on the meter). If the reading appears 'noisy' then the TPS is bad. Equally if the value at either end of the range is incorrect, it is also bad.
The most likely area of wear on a bad TPS is the area which sees the most use - from idle to part throttle. Usually mid to full throttle is fine as the wiper spends less time in that area and so doesn't wear the track away.
Edited by spiguy, 25 June 2015 - 09:49 AM.
Maybe the TPS will turn out to be the problem. Do you have a mutlimeter to be able to read the resistance value through its operation range?
