When adjusting the tank drive level, I didn't manage to get more than about 500 microamperes RMS through the transducer, before serious distortion appeared. This is approximately 16 dB below the recommended drive level. This has to be compensated for by adding gain in the recovery amplifier, and the tank sensitivity to microphonics and external fields will therefore be degraded by the same amount. The distortion was caused by the driver op-amp running out of voltage headroom. Further investigation showed that the combination of this reverb tank and the RC network across it has a strong resonance around 10 kHz, where the impedance peaks at 8 kohm or so. I replaced the 3.3 nF capacitor (C108) with a 1.5 kohm resistor, to form a simple pole around 6 kHz. This limits how high the impedance of the tank can rise at high frequencies, and allows a higher drive level within the voltage rails of the op-amp.
Secondly, the recovery amplifier has a pole (lowpass) at 3.3 kHz formed by R109 and C110. This cuts too much of the high-end in the reverberation signal for my liking, so I replaced C110 with 470 pF. There's little benefit to going above 10 kHz, so ideally a 680 or 820 pF cap should be used here. I only had 470 pF in the right package, so I ended up at 16 kHz.
There's no low-pass filtering of the signal going into the reverb driver. To avoid spending headroom on processing frequencies that won't make it through the springs anyways, I added a pole to the reverb send signal, 1.5 nF across R6.
Source: https://www.facebook.com/groups/436675570022014/permalink/1144002955955935/
When adjusting the tank drive level, I didn't manage to get more than about 500 microamperes RMS through the transducer, before serious distortion appeared. This is approximately 16 dB below the recommended drive level. This has to be compensated for by adding gain in the recovery amplifier, and the tank sensitivity to microphonics and external fields will therefore be degraded by the same amount. The distortion was caused by the driver op-amp running out of voltage headroom. Further investigation showed that the combination of this reverb tank and the RC network across it has a strong resonance around 10 kHz, where the impedance peaks at 8 kohm or so. I replaced the 3.3 nF capacitor (C108) with a 1.5 kohm resistor, to form a simple pole around 6 kHz. This limits how high the impedance of the tank can rise at high frequencies, and allows a higher drive level within the voltage rails of the op-amp.
Secondly, the recovery amplifier has a pole (lowpass) at 3.3 kHz formed by R109 and C110. This cuts too much of the high-end in the reverberation signal for my liking, so I replaced C110 with 470 pF. There's little benefit to going above 10 kHz, so ideally a 680 or 820 pF cap should be used here. I only had 470 pF in the right package, so I ended up at 16 kHz.
There's no low-pass filtering of the signal going into the reverb driver. To avoid spending headroom on processing frequencies that won't make it through the springs anyways, I added a pole to the reverb send signal, 1.5 nF across R6.