analogguru said:
Anybody knows why audio-technica dropped their patent for their AT3060 (which is very similar to the circuit used here) ?:
http://www.freepatentsonline.com/7295675.html
(At least they didn't use an opamp in the signal path of a "tube"-microphone.)
And anybody knows which grid current/(resistor) is necessary for the 6418 to ensure proper bias ?
I am not sure they DID drop the patent; you see the one YOU linked to is not the only AT3060 design! A later one has 4 current-regulating diodes and a center-tapped primary transformer...currents to plate and Grid#2 are moving in opposing directions in the primary and a resistor network was selected to BALANCE those currents within ~5% or less, so the primary DC fields should "cancel", thus AT could utilize a much lighter transformer in the later AT3060 = cheaper to make. So you linked to a patent which is not the latest one. The later one feeds 2 current regulating diode to primary CT, primary wires to plate and Grid#2, and, I think they tweak the current to grid#2 current to approximate what should be at Plate. If the same the opposing DC fields are the same?
Version 1 they used oh I want to say 10K down to the 10V Zener. The voltage divider on the grid #2 I think 43K and 100K below.
Most other part values would not matter, maybe 10pF feedback from plate to Grid#1. [1G-Ohm to GND at main input Grid, or whatever you have in high values] AT3060 Version 1 you need a transformer that does not saturate in the primary due to the one way DC flowing there. Like a $75-$150 transformer not a cheapie DI box tranny.
Zap went with QTy zero coils on the DC-DC side. If ya use coils etc, then you have more hash and magnetic fields...[I think].
You would not need a 52V rated buck regulator if you used a zener to pre-regulate the buck, or, a resistor-value you could maybe determine with a POT first. Zener makes noise but so does a buck-regulator, either way you have work to do, since the cathode is the heater too
nice
not to shove in noise THERE. They make some pretty nice like 820uF 2.5V caps [for the heater DC] you can find on carriers in the SMD world, those can be used through-hole after removal from carrier. Maybe bypass that [parallel] with 0.1uF 0805 SMD cap, and toss in a series choke or ferrite?
The UM900 is said to have an OP AMP, people buy them. Maybe that was your intent to say? Could use a MOSFET follower, or bipolar follower there too, and then yah a transformer via a cap. With a OPA you could give the mike different profiles, or just strap for unity and cap to transformer. The tube is still first in line, sounding like a tube does. I doubt you could hear the impact of a GOOD unity-gain OPA after the tube. It is certainly possible to use a OPA in Franken that happens to idle at say 1.5mA, and figure that the LM2665M6 chain is going to operate at some efficiency less that 100%. That heater you'd want at maybe oh 9mA? Or be a boy-scout and run at 10mA. 1.5mA at say 80% eff [chain of 2 reverse-wired LM2665M6] would be ~9.6mA heater? Some mike guys starve heaters a splash on purpose. No?
MY argument about the famous Naiant MSH4 is that the first stage IS actually a jFET [in a capsule] not a tube off the skin of the capsule. Vs Gefell UM900 would be LDC into tube grid#1 after a DC isolation cap since Gefell probably puts 60+ Vdc on the capsule -- unless they used the old ZapnSPark way and do not ground the non +60V side but rather shove that into grid#1 -- the capsule IS a cap.
Resistor grid #1 -to-ground? Or a series R? Grid-to-GND, why not use some Augat socket-pins and try: 2G-Ohms, 1G-Ohms, 100M, 10M, 1M ? See what you see on a scope? Listen??? None of them will hurt anything. Any of those values will allow the tube to operate.
