@Kerry: most pumps need a good impulse current draw to be activated, and once activated they can remain on with a lower voltage.
The voltage dropping to 9V by the end of run seems a bit strange — it will certainly be lower than 12V but dropping to 9V seems a bit too much. To understand the setup:
– Did you test with just one pump or multiple pumps?
– When it drops to 9V, can you test the output voltage of the transformer? You can do so by measuring the voltage between and GND pin and the pin on the back of the black power barrel (i.e. where the transformer is plugged in). Normally the voltage presented to the station would be somewhere between 0.5V to 1V lower than the input voltage (due to the diode explained below). So if the input voltage is already dropping to 10V or below, then it’s normal for the voltage to be 9V at the pump.
Now I think about it, OpenSprinkler DC circuit has a 2A diode on the input to COM path which practically limits the total current supply to all valves to about 2A — it can probably go above 2A but the voltage drop on the diode will become more significant. This shouldn’t be an issue if the total current draw is well below 2A. And if you do need more than 2A total current, there are ways to bypass the limit.