Wait for the tester to warm up and become stabile. For a quick test you can skip this, but in case you want to adjust something, wait at least 30 minutes before doing so.

Important: Do NOT plug all holes with card plugs! This card uses only two holes with card plugs, which are clearly indicated. These will activate the internal voltages. The other holes are used to connect a voltmeter to. For the voltmeter, connect the ground lead (so the black lead) to where is says "ground lead" on the card, also marked with a "0". Then, the remaining open holes numbered "1" to "5" supply the reference voltages for the voltmeter.

Set the needle of the panel meter to the mains calibration mark, with the rotation switch. Attention, the meter indication depends on two things: 1) Correct Meter sensitivity and 2) Correct current applied to the meter. See the next note on this.

1) The meter sensitivity can be wrong. Possible reasons: Replaced meter, bad meter or meter calibration was changed. Meter calibration is by the magnet screw and by the panel potmeter. If there is a problem here, it is rare. You usually find out, when you connect an external DC voltmeter between the banana plug of the "Anode", and the other lead connected to the ground of the tester. (There is a connection for ground, close to the mains inlet). The panel meter should be very precise, actually outdo precision of low cost multi meters. So when your multi meter is a 5% type it can not be used for this. If the panel meter generally indicates too high or too low, or if there are linearity errors, you quickly will find out this way.

2) The current which is applied to the meter can be wrong, when the series resistor for this is wrong, due to aging. This is a a fixed resistor, factory selected. The value in the schematic is only average. There is no official instruction how to work on this. So what you read here, is only my experience with this. It is my conclusion from verifying several of the L3-3 I have here. As a reference I recommend to check several unloaded voltage taps of the transformer. If these appear all too high, or all to low, you can say the internal calibration resistor is wrong. My advised method is, to adapt this resistor such that the AC voltages on the mains transformer are right. So what voltages are that, and what values? There is no official information for this, so I just collected this information from several L3-3 in my collection. I have put this together in test cards #5 and #6. With these cards you can conveniently tap those voltages simply from the plug board, and values you need to measure are on the cards, with their tolerance.

In the internet some people write, they use the internal 6.3 Volts of the electronics for this, but this is definitely not the right way. Internal AC heater voltage is 7...10% too high and that is normal because a mains transformer is always specified under full load for all windings simultaneously. . Wheras on the bench you load it only partially, and some windings are unloaded. Moreover, L3-3 seems to have used the heater tolerance also to guard a little bit against mains variations. Thus to ensure good functioning of the electronics. So the tubes are overheated a few % and that seems normal will all L3-3 I have.

-100V. This is a negative voltage for the internal tube electronics. It is a Zener stabilized voltage, and can not be any better as a Zener diode is. However this seems a precision diode. I found it amazingly stabile in my L3-3 testers, and if you have more than a few Volts off, I would not trust it. Sure all electronics will work nice a -105V too, but there is something else wrong, if a -100V precision Zener gives -105V and you need to find out why that is.

This is the an important one. The internal oscillator is used for Gm testing. The frequency and the amplitude must be right. The amplitude must be roughly ok, and is unimportant for accuracy. That is because at the beginning of each Gm measurement, the circuit is calibrated anyway. Just the level itself should not be totally wrong. If so, there is either a problem with the electronics, or somebody played around with the settings. However when you adjust the oscillator voltage, you likely change the frequency too, and the frequency is VERY delicate because there is a sharp band pass filter at the internal voltmeter *. So do not just change the amplitude now, but use the official procedure.

Conclusion: 450mV at 1400 Hz should be there, within some reasonable range. Just not be totally off, or instable. Also check with an oscilloscope for good wave form. Though the sharp band pass filter of the internal voltmeter will reject distortion anyway, the signal should have no visible distortion just with the eye. If there distortion, that would indicate a problem with the oscillator, even when it appears to work.

* This was done to reject harmonic distortion of the tube under test. Also to reject d hum and harmonic products of the hum, as that would all be interpreted as "signal" coming from the tube, and make Gm appear better as it really is. Some tubes have 5% distortion or even more at high signal. Not to forget Variable Gm tubes, which by nature have very high distortion. So L3-3 removes this error from the signal.