Test of 12AX7 tubes
I tested 36 new-production 12AX7 tubes, 11 brands. Below are some charts and tables about the measurement.
36 tubes is a very small batch, please let me know your experience if you have other results. Click here to send me a message. I would also be interested in any insights on the effects of the off-standard tube parameters in actual (guitar) amplifiers.
Those not interested in the technical details can skip forward to the charts and conclusions.
My measurement method somewhat differs from the standard method, but it simulates real-life conditions better.
As you probably know, the standard 12AX7 tube gives 1.2mA of plate current for 250V plate voltage and -2V grid. Traditionally, tubes are tested with a fixed plate and grid voltages (e.g. 250V and -2V). But in real-life circuits preamp tubes are auto-biased with a cathode resistor. The actual quiescent point greatly affects tube parameters, transconductance, plate resistance and even amplification factor. To simulate real-life conditions I created a test circuit very similar to the circuits these tubes are used in actual amplifiers.
In this circuit a standard 12AX7 tube yields exactly the standard voltage and current values. In the case of real, non-standard tubes, the quiescent point is shifted very similarly to what happens in a real amplifier circuit.
I measured the usual values, mu, gm and rp. Moreover I also measured the grid-cathode voltage, which I considered as an indication of the headroom. The usual common-cathode amplifier begins to distort heavily when the grid approaches the cathode (blocking distortion). The higher is the quiescent voltage between grid and voltage the later the distortion begins, which means more headroom. Since the standard tube's grid-cathode voltage is 2V, I defined relative headroom as 20*log10(Ugk/2).
I also calculated the actual gain of the common-cathode amplifier (bypassed cathode), which is (mu*RL)/(RL+rp).
Charts and conclusions
- Sovtek WA, WB, WC have very low transconductance, around 1250 instead of 1600. WC's mu is standard (95-105) WB's mu is slightly lower (95-100), while WB's mu is higher 100-115. The transconductance is so low that it cannot be tolerated if you need a real 12AX7.
- On the other hand, Sovtek LPS tubes have higher then standard transconductance, around 1800. The mu is around 100-110.
- The third isolated group contains the JJ and TT Blackplate (7025) tubes. They have standard gm but low mu (85-95)
- All the other tubes are distributed around the standard values, with slightly higher mu (100-110).
This chart contains the same information as the mu-gm chart, but the allows us to see what happens to the plate resistance. The standard plate resistance is 62.5 kohm
- The problematic Sovtek WA, WB and WC tubes have very high plate resistances because of the low gm. Among them, the price the WB tubes pay for their above-standard mu is the even higher plate resistance.
- JJ, TT 7025 Blackplate and Sovtek LPS tubes have similar, lower-than-standard plate resistances
- JJ tubes have higher-than-standard headroom and lower-then-standard gain. This explains why these tubes are harder to get into distortion.
- Sovtek WA, WB, WC tubes tend to have lower gain and lower headroom.
- Most other tubes have slightly higher gain (~0.5dB) and slightly lower headroom (~ -0.3dB) but I need much more measurements to establish tendencies between brands.
Tamas Kenez, 2007-APR-02