Adrians free time projects

Best sample out of 16, selection done after 100h burnin.

Best sample out of 10 burnt-in triode systems.

Note the re-adjusted gridcurrent model. It is more accurate below approximately 100uA compared to the old 12AX7.TSi4 model. To distinguish the new from the old model without enlarge the suffix, I replaced the "i" by an "r" which means "re-adjusted".

I apologize for the not so well fitted 12AX7.TSi4 model.

Very linear triode, which has (for this kind of tube) a well suppressed island effect. High gm.

This triode offers both high mu and high gm, with still quite moderate heating requirements. Note that it is designed for grounded grid configuration, so the usage in standard configuration might show a significant high input capasitance.

For its relatively high mu, the 6BN4A performes amazingly good under low plate voltage conditions.

This Model is based on measurements of the most typical triode out of 10 triode systems, each burnt-in for 100h. It is optimized for any grid voltage up to +4V.

This Model is based on measurements of the most typical triode out of 10 triode systems, each burnt-in for 100h.

best sample out of 10 triodes, each burnt-in for 100h.

This model is based on the most representative triode out of 4 tubes, all burnt-in for 100 hours.

This model is based on the most representative triode out of 5 tubes, all burnt-in for 100 hours. Note that this tube turned out as quite linear (with an exponent of only 1.37) when used below Va=100V.

This model is based on the most representative triode out of 5 tubes, all burnt-in for 100 hours.

Note that this tube is very special regarding its manufacturing tolerances: While it is simply the best in terms of contact voltage variations, its plate resistance has a quite high tolerance.

This model is quite accurate up to Vg=4V and offers usable accuracy up to Vg=16V.

This model is based on the most representative triode out of 5 samples, burnt-in for 100h.

This triode has the same long plate construction as the NOS Telefunken ECC82 (plate length is approx. 15mm), and seems to be of quite similar construction like the JJ ECC802S.

Best sample out of 16, selection done after 100h burnin.

This model is based on the most representive sample out of 5 tubes, each burntin for 100h. So far, all electrode constructions I have seen were identical regardless of the brand. So, the suffix is shortened as the both letters for the manufacturer is not needed for this tube.

Note that this tube shows gm drift (about +10%) when the anode runs hot. For bias points with anode losses above 1W, it is better to use my PC900.i6 model (based on "hot plate" measurements)

It is a Valvo tube. The Siemens verson is totally identical, while the Telefunken showes higher gm and contact voltage. Best sample out of 6 burnt-in triodes.