Here is some additional info that may be helpful , (That someone else wrote)
http://jvmforum.com/phpBB3/viewtopic.php?f=7&t=3310
jon wilder did that a couple months ago on the marshallforum, but it takes quite some time to understand what he was talking about...
ok i'll trie my best...
you're familiar with the 70% dissipation limit, right?
this 70% limit is determined by plate voltage and plate current... with the following calculation:
plate voltage X plate current X 0.7...
this 70% of plate dissipation on idle is considered as the absolute limit...
how to read plate voltage is pretty simple... red lead on pin 3, black lead on ground, no sweat
plate current however is another story... you have to take a measurement from the center tap of the OT to pin 3... bias probes cannot do this...
measuring current is dangerous, because your multimeter becomes a dead short... touch off pin 3 with the red lead and accidently touch the black lead with your finger and you're about to get zapped...
when trying to make a plate voltage reading, while the DMM is set for current... the HT-fuse will blow (if you're lucky), otherwise you short the OT... bye bye dagnall, welcome mercury magnetics [smilie=icon_mrgreen.gif]
so they come up with a safer and more simplyfied approach... they measure the current from the cathode... only one problem... cathode current isn't the same as plate current... cathode current is plate current + screen current...
so let's calculate... marshall recommends 30mA cathode current per tube... you take readings in millivolts, because you measure across a 1 ohm resistor...
marshall states, the plate voltage on the JVM-series is 480Vdc
480volt X 0.030amp = 14.4watt
14.4 / 25 X 100 = 57.6% of plate dissipation, right?
wrong! you have to remember, we're measured cathode current, so the screen current is calculated within this figure...
max plate dissipation for an EL34 is 25watts
max screen dissipation for an EL34 is 8 watts
combine them as we did with the measurement of the cathode current... and you guessed it:
25+8=33 watts...
remember the dissipated 14.4 watt's?
14.4 / 33 X 100 = 43.6%...
way cold, but very safe... WHY?
current production EL34's are nor as sturdy as they used to be in the old days...
for example a JJ EL34 has a screen voltage limit of 450V... screen voltage is only a bit lower then plate voltage.... remember the JVM runs 480 at the plates... this is exact the reason why marshall recommends SED winged C's for their amps... old or new... the SED has a max screen voltage limit of 500v
why do i want to know the exact plate current on my JVM?
first, i'm running JJ KT77's... they have an advertised screen voltage limit of 800 volts (a bit optimistic in my opinion) but the max screen dissipation is lower... 6 watt compared to the 8 watt from the EL34... they are said to draw more current as well... in other words... it's a different kind of tube
second, i like to bias the amp, while i get the max main voltage from the wall... over here we get 220Vac, but sometimes it's quite lower (i've had AC readings of 200V) those voltage drops will affect the bias as well... setting it up a bit colder, giving you less volume and feel
third, my plate voltage isn't near the 480V marshall states... i'm just at 430Vdc with 84mV on the testpoints...
calculate this?
430volt X 0.042amp = 18.06 watt....
18.06 / 25 X 100 = 72.2%... oh dear, i crossed the limit?
no i didn't... because we measure cathode current, not plate current... and somehow i'd like to know what the limit is, before i cross it by accident...
every time i learn a bit more, i discover how cold the JVM is actually biased... every time i discover i can bias it hotter then i thought was possible... and every time i bias it a bit hotter it sounds better...