The WM-2 QRP Wattmeter by Oak Hills Research is a very popular piece of test equipment. It can display forward or reverse RF power on 10W, 1W and 100mW scales. It is designed to be run from external power or an internal 9V battery. If it is run from the internal battery and the user forgets to turn the rotary Power/Range switch back to the "OFF" position, the battery can be depleted.
A post last night to qrp-l mentioned a mod to prevent this from happening. A momentary push button is mounted on the front panel, along with a small LED. When the button is pushed, power is applied to the meter circuit, and the LED lights for a minute or two. Then power is automatically removed from the meter circuitry, saving the battery. This is generally enough time to take the readings needed.
Actually, the mod wasn't mentioned at all in the post. A list member happened to post a link to a picture of a rig he just finished building, and there was his WM-2, sporting an LED and a pushbutton. Another list member inquired about the button and LED. The owner said that some time back, someone had come up with the mod, and had provided the 'hard to find' parts. The original mod instructions or details had been swallowed up by the Internet.
So..... I took this as a personal challenge. I felt that the gauntlet had been thrown. Last night, I threw together a very simple circuit that does this job.
Basically, a garden variety 2N7000 FET is in the ground lead of the power for the meter. The FET has a relatively large capacitor (4.7 - 100 uF) from gate to source. Across the cap is a large value resistor which slowly discharges this cap. The cap is initially charged through a small resistor and pushbutton. The "charging" resistor is in the range of 100-470 ohms (not critical). How long the meter stays on is determined by the time constant of the capacitor and discharge resistor across it. With a 1 megohm resistor and 47 uF cap, the 2N7000 will conduct for about 1 minute. The values are not critical. 10 megohm and 4.7 uF will yield the same timing. For twice the on-time, double the value of the cap or resistor.
When the button is pressed, the cap is charged to the input voltage of the meter through the small resistor. This voltage triggers the gate, turning on the FET. The 2N7000 can safely handle about 200 mA drain current, so powering the meter circuit and an LED is not a problem. When the button is released, the capacitor begins to discharge through the resistor across it. As its charge voltage drops below the gate threshold voltage, the FET shuts off. Since this is being used in an RF environment, bypassing the FET G-S and D-S with .01-.1uf caps is generally good practice.
I found a nice efficient LED in the drawer here. It's plenty bright enough with about 4mA flowing through it. I used a 2.2K Ohm limiting resistor in series. As the cap discharges below the FET threshold voltage, the LED dims and goes out. This takes about 10-15 seconds after the timing period described above.
When off, the leakage current through the 2N7000 is less than 1 uA. The life of a good Alkaline 9V battery should be essentially it's shelf life.
I have used this same type of circuit for many applications. It is simple and reliable. Total cost of parts, maybe a buck or two. Less (or zero), if you have a reasonable junque box.