Ortofon 2M Black, AT 150MLX, Stanton 881S if you can find an old body to add a Jico Shibata aftermarket needle.
I was planning to do some grid current investigation after I have measured up my batch of 6N2P's. Will take a pair of low current tubes and a pair of high current and see where gc starts and where it differs between tube samples. I guess mine are from a small number of batches but it might get us an idea of things.
In the meen time Merlin, I suggest one that looks linear in small currents but goes a bit over 1 v at that current. Maybe near where your dual diodes takes you now.
In the meen time Merlin, I suggest one that looks linear in small currents but goes a bit over 1 v at that current. Maybe near where your dual diodes takes you now.
It could, depends on how quiet the rail is. Its a practical matter to test. But look in page 3 IF/VF (fig. 4) for this example VISHAY part that was in Merlin's RS link: http://www.vishay.com/docs/81010/tsal6200.pdf
It looks like not demanding more current to go straight curve, what do you think?
I think they just draw the line down because there is a current where the diode simply wont turn on. And they dont show that. After that comes either an unlinear or unexpectable zone.
Though. The 5-6 different samples I tested last showed at least a decent narrow spread between samples. But i didnt do linearity tests
If its a line stage and has to swing mA there will be a significant delta Ohm if biased low. That is why its not optimal if not boosted. But see, Merlin does not have an obvious problem even with 1N4148s for now which are much worse for delta Ohm. So the delta problem is not challenged enough in this application. 
Morgan Jones says parasitic signal (THD) Vrdiode=(VRL*rdiode)/RL
So make signal VRL smaller or make rdiode lower (more mA bias) or increase load of the stage. In μ follower RIAA input stage we have both high RL and low signal.
Indeed. Since you separate upper tube to lower un the mu follower. Have you experimented with another top tube?