Just found that Fairchild makes a SMT version mmbf4117 "for sub-picoamp instrumentation". This version would cost $1.02 (for a pair).
5G resistors run about $3 TH/SMT.
.
In small quantities, the BAV199 runs $0.25, SMT 1G $1.51, TH 1G $0.64.
5G resistors run about $3 TH/SMT.
.
Yes, I wouldn't use a 6AS7.Hi Scott,
Thanks. Yes I knew it was a cascode, but I didn't know if it were possible for the FET to have a relatively high voltage developed across it. Just curious since I hadn't seen this circuit before.
If a tube was used that had a very low gm (a different tube type), would it be possible to have too much voltage appear across the mic assembly?
-Chris
Eastern hemisphere last time I studied any geography. But that was a long time ago so they may have changed the prime meridian since then.
se
And the other hemispheres are?
Yes, I wouldn't use a 6AS7.
And of course not 6c33c-b 😀 ,
---------------------------------
BTW , I never experienced any electret capsule electrical fail incorporated in this simple design , but only few with mechanical fails , mostly caused from surrounding moisture .
tobydog,
Northern and Southern hemispheres, covers the whole damned world far as I remember.
Northern and Southern hemispheres, covers the whole damned world far as I remember.
And of course not 6c33c-b 😀 ,
---------------------------------
BTW , I never experienced any electret capsule electrical fail incorporated in this simple design , but only few with mechanical fails , mostly caused from surrounding moisture .
Of course, you guys are goofing on improbables, but I've seen pictures of a real phono equalizer built with Eimac type 450TL final stage tubes. Maybe falls into the category of an art project (in Japan, where it's allowed). Sorry for using the forbidden "a" word in serious company.
All good fortune,
Chris
It was a great smartass reply. 😛
se
Thank you for taking it as it was intended. 🙂
And the other hemispheres are?
Northern and southern. But Europe, Japan, Austalia and New Zeland don't span more than one actual hemisphere. In other words, it's just half the world.
se
tobydog,
Northern and Southern hemispheres, covers the whole damned world far as I remember.
But Europe and Asia don't span the whole damned world. Just one half of it. Need another half to get a whole.
se
I hear the 5881 is also good in signal stages, but the 45 trumps them all! Filament power baby!
Of course we have to limit the kidding to signal tubes, so because I can't find a 200 anywhere, I'll pull one of my 201's out!
Of course we have to limit the kidding to signal tubes, so because I can't find a 200 anywhere, I'll pull one of my 201's out!
If memory serves, South America did not take part in the whole WW2 mess, so that leaves out a significant part of the Western hemisphere, reducing it geographically to just North America.
That's only if you work in percentages of territory engaged in WW2. If so, then only South America sat on the bench, and workig by continents, 5 of 6 were in it. Quite enough for a decent Wordl War, I think.
That's only if you work in percentages of territory engaged in WW2. If so, then only South America sat on the bench, and workig by continents, 5 of 6 were in it. Quite enough for a decent Wordl War, I think.
Still having total mental block with this one..........You need to bias the capsule which can be 50pF or so. Now consider the noise of a parallel RC and that the Brownian noise is around the 5nV level. You soon realize that the resistor must be VERY big.
Say capsule bias is fed from a noise-free voltage source via a resistor. This resistor has unavoidable associated JN noise which increases with sqrt of its resistance. Then bigger resistor is more noisy, and 5G resistor itself generates about -60dBu of audioband noise........!?!😕
The capsule capacitance forms a low pass filter for the resistor noise. The larger the resistor the lower the corner frequency of the LPF and the lower the noise in the audio band. It is a little paradoxical, but it works.
The capsule capacitance is shunted the 5G resistor. The delta i caused by the changing capacitance develops a voltage across the R//C network.
Just think source impedance ?Still having total mental block with this one..........
kT/C
For a parallel R-C at equilibrium, the rms voltage is independent of R, namely (kT/C)^(1/2). As bkdog says, the corner frequency gets lower with larger R and thus more of the noise is infrasonic.
The T dependence is real, but cooling is not of much help unless the microphone is bathed in dry gas to prevent condensation, and then only proportional to T^(1/2) 😀
Preamp gate leakage is important. If it shows full shot noise, an equal contribution to resistor noise occurs with a leakage current of about 10.3pA for a 5G resistor at 300 K. This means choosing a JFET for low voltage noise alone may not be optimal (this is discussed by Scott W.). Since JFET leakage increases with temperature, roughly doubling for every 10 degrees K, cooling would help a lot were it not again for condensation (and power consumption).
For a parallel R-C at equilibrium, the rms voltage is independent of R, namely (kT/C)^(1/2). As bkdog says, the corner frequency gets lower with larger R and thus more of the noise is infrasonic.
The T dependence is real, but cooling is not of much help unless the microphone is bathed in dry gas to prevent condensation, and then only proportional to T^(1/2) 😀
Preamp gate leakage is important. If it shows full shot noise, an equal contribution to resistor noise occurs with a leakage current of about 10.3pA for a 5G resistor at 300 K. This means choosing a JFET for low voltage noise alone may not be optimal (this is discussed by Scott W.). Since JFET leakage increases with temperature, roughly doubling for every 10 degrees K, cooling would help a lot were it not again for condensation (and power consumption).
Last edited:
Since WWII started in 1939, you're both wrong! 😀
They were in the 1939 catalog according to an AES paper.
Still having total mental block with this one..........
Say capsule bias is fed from a noise-free voltage source via a resistor. This resistor has unavoidable associated JN noise which increases with sqrt of its resistance. Then bigger resistor is more noisy, and 5G resistor itself generates about -60dBu of audioband noise........!?!😕
It also falls out of the classical KT/C noise derivation (R goes to infinity). I know you love this stuff. 🙂
http://www.coe.montana.edu/ee/andyo/ee503/ieee_abbott1996_thermalnoise.pdf
Or simply plot the noise spectral density of a parallel R and fixed C as R gets larger as others have said. The pole goes as R and the noise as sqrt(R).
Last edited:
doesn't stop MLB
You left out the WBA and WWF.
(the US was always good for a giggle now and then, alas no more. At least the aliens still land in central park, to start the destruction of earth in America, but they may be still bound by contract)
If it shows full shot noise, an equal contribution to resistor noise occurs with a leakage current of about 10.3pA for a 5G resistor at 300 K.
Yes a good rule of thumb is when the current makes 2Vt on the R diminishing returns set in. Unfortunately the JFET datasheets don't help, the Ig spec is invariably sandbagged, for instance, SK170 typically measure only 2-3pA on an electrometer at low Vgd.
- Status
- Not open for further replies.
- Home
- Member Areas
- The Lounge
- John Curl's Blowtorch preamplifier part II