Think I found the problem. The zener I was using, a 1N4733, was clamping to 4.3 volts. I tested it by putting the zener in series with a 3300 ohm resistor ( negative end of the zener was attached to the resistor) and attaching the resistor's other lead to the positive end of a 9 volt battery and then attaching the positive end of the zener to the battery's negative terminal. The voltage across the zener was roughly 4.3 volts. Guess its time to find a new zener diode.
Edit: When I put my 6.8 volt zener diode (1n4736) in place of the 5.1 volt zener (1N4733), it worked fine and clamped to just under 6.8 volts.
According to the Fairchild datasheet, the 1N4836 is rated at 1 Watt and Izt is 49 ma.
To get that much current your resistor would need to be 80 ohms, and that is only if the battery is still 9 volts when you are drawing that much current .
(9V-5.1V) / 49ma = 79.6 ohms.
Zeners with lower power ratings have lower test currents.
Lowering the resistance from 3300 ohms to 510 ohms increased the voltage to about 4.8 volts. Lowering the resistance of the 1k and 10k ohm resistors in the schematic should work, right?