Both sides of the tube – parallel ! schem. added
- Thread starter e.oelberg
- Start date
Help Support GroupDIY Audio Forum:
Biasrocks
Well-known member
Gus said:
Lucas CS1 I believe.
http://www.compasspointstudios.com/lucas/cs-1_mic.html
Mark
It's been done a number of times.e.oelberg said:
Three advantages:
The output impedance is halved
The noise factor is 3dB lower
It offers some redundancy if one half of the tube goes down
ruairioflaherty
Well-known member
abbey road d enfer said:
Just curious...are there in your opinion any disadvantages to this way of doing things? Engineering always seems to be a game of compromises so I'm wondering what gave in this instance? (if anything).
Cheers,
Ruairi
Yes, generally not a big problem. In this case, Miller effect is part of the feedback loop; as long as it's taken care of, it's ok. I can't really think of any other disadvantage, apart from the increased power consumption.
Butterylicious
Well-known member
Fender reverb driver too.
> any disadvantages to this way of doing things? Engineering always seems to be a game of compromises so I'm wondering what gave in this instance? (if anything).
In any rational world, two triodes cost more than one.
However, there are very few small single triodes in current production; boat-loads of twin triodes.
In a bigger box, you would find another use for the "excess" triode. Tremolo. Insert buffer. Or another input, with knob and jacks, which looks more impressive (higher price) on the showroom floor.
But not much you can do in a microphone. Gain stage - cathode follower actually adds more trouble than it solves.
And if this thing sells for much more than $99 (it surely does), the cost of a twin-triode is nearly negligible.
And in this irrational world, surely smaller than any assured-supply small single triode or low-noise pentode (which could be triode-strapped).
In this application you want to balance several conflicts:
1) low-low grid current
2) high plate current for low voltage noise and high drive
3) low heat near capsule
4) cost versus value
#2 is also related to cathode area. At the same current (heat), a larger cathode has lower current density and lower voltage noise. We would like a single triode with an over-large cathode. But this world is not rational. Anyway cathode cost is a large part of tube cost. And most uses do not need over-generous cathode. Using two cathodes as one "forces you" to pay for an extra grid and an extra plate, but translates to a broad-market twin-triode instead of a narrow-market fat-cathode single triode. So it is cheaper. Even if money is low-priority, supply and spares are always a concern.
The ~~5V bias means grid current will be low. And splitting the same plate current into two grid streams may not mean more total grid current. And at 5V, in practice grid current on modern tubes will be dominated by gas, not the stray grid current of a "perfect" tube. Surely the tubes are "selected". If this is a $99 mike, the Banjo Mart customer "selects" by taking back the worst hissers for exchange. If this is a $9,999 mike, there is a strict spec for the tube factory and additional burn-in and testing in the mike assembly process.
That's all trivial. What strikes my eye is the clever remote pattern control, eliminating a grid capacitor.
In any rational world, two triodes cost more than one.
However, there are very few small single triodes in current production; boat-loads of twin triodes.
In a bigger box, you would find another use for the "excess" triode. Tremolo. Insert buffer. Or another input, with knob and jacks, which looks more impressive (higher price) on the showroom floor.
But not much you can do in a microphone. Gain stage - cathode follower actually adds more trouble than it solves.
And if this thing sells for much more than $99 (it surely does), the cost of a twin-triode is nearly negligible.
And in this irrational world, surely smaller than any assured-supply small single triode or low-noise pentode (which could be triode-strapped).
In this application you want to balance several conflicts:
1) low-low grid current
2) high plate current for low voltage noise and high drive
3) low heat near capsule
4) cost versus value
#2 is also related to cathode area. At the same current (heat), a larger cathode has lower current density and lower voltage noise. We would like a single triode with an over-large cathode. But this world is not rational. Anyway cathode cost is a large part of tube cost. And most uses do not need over-generous cathode. Using two cathodes as one "forces you" to pay for an extra grid and an extra plate, but translates to a broad-market twin-triode instead of a narrow-market fat-cathode single triode. So it is cheaper. Even if money is low-priority, supply and spares are always a concern.
The ~~5V bias means grid current will be low. And splitting the same plate current into two grid streams may not mean more total grid current. And at 5V, in practice grid current on modern tubes will be dominated by gas, not the stray grid current of a "perfect" tube. Surely the tubes are "selected". If this is a $99 mike, the Banjo Mart customer "selects" by taking back the worst hissers for exchange. If this is a $9,999 mike, there is a strict spec for the tube factory and additional burn-in and testing in the mike assembly process.
That's all trivial. What strikes my eye is the clever remote pattern control, eliminating a grid capacitor.
PRR said:
yes very nice, I have something similar in my relay based sort of ELAM 251. First I did the whole pattern switching with releais, but the way from capsule to grid became too long so the mic started to become noisy. Now I have a relay inserting the 1000pF for Figure of 8.
nicholas
ioaudio
Well-known member
PRR said:
the neumann m49c was the first which featured this kind of switch (+4dB), but without relais at that time.
i think david bock used a relais in his mics, but i don´t know who was the "first" one.
nicholas, the rise of noise in your mic is due the insluation of standard relais. i see that o.a. used standard relais too, but maybe he modded them for high Z use? special (costly) highZ reedrelais can be found from a german manufacturer.
-max
ioaudio said:
no I chose hamlin relays with pretty high isolation resistance (Gigaohms), I was aware of that problem. No I try to figure out a nice solution mixing
remote controlled pattern with some relays for the 1000pF and for a cardoid with back and half of the plate disconnected.
nicholas
lets see if it works, nicholas
ruairioflaherty
Well-known member
PRR said:
Thanks for the detailed answer as ever and D Enfer too. As it happens this is a limited run $3500 mic. I've read bits and pieces about it on other forums.
All the best,
Ruairi
Hi Max
Can you tell which relais that is?
I did the cardioid only switch in my microphone with a reed relais from meder and the mic was noisy afterwards. Without the relais everything is fine, but I do not want to loose the option of the disconnected rear diaphragm.
Thanks Tobias
ioaudio
Well-known member
http://www.meder.com/fileadmin/meder/pdf/en/Products/Reed_Relays/Relay_HI_Series_E.pdf
