How does CCDA circuit compares with other clasic vintage tube mics circuits (pcb's) available

[Nen'O]

Even if this subject is well covered on the web, it would be usefull to get first hand experience from actual audio engineers.
How does CCDA compares with other clasic vintage tube mics circuits (pcb's) if used in the same body with the same capsule and transformer.
Brian Fox from FoxAudio has great article, in his words CCDA in chinese mics we find is actualy used back in 1995 by David Bock of Bock Audio in his Soundelux U95.
Production of that mic was moved in China for brief time and that is where Chinese engineers ripped off CCDA used about all chinese tube mics we find today.

https://www.foxaudioresearch.ca/CCDA.htm

So far I collected 3 pairs of donnor mics. Apex 460, t.Bone sct2000, t.Bone sct 800 and one SYT 5 from Shuaiy (with pcb made for Advanced Audio).
All mics are modded with well know mod instructions by David from Advanced Audio.
Tubes are EH 12AY7 except for sct 800 (Philips 12AT7), Transformers are from AA and UTM or stock. Capsules are all chinese made Advanced Audio AK12, Maiku K47, ebay M7 and 35mm center terminated.
Apex 460 has Advanced Audio AK 12 capsule, Transformer is AA BV 18, (8:1).
t.Bone sct2000 has Maiku K47 capsule, UTM 0547 (6.5:1)
t.Bone sct 800 - Stock transformer, capsule is 35mm center terminated.
SYT 5 - stock transformer, M7 ebay capsule.

All mics sound great to me, other recording engineers who used them are impressed and musicians loved them too. Due to different capsules and shape each pair of mics has different characteristics. No noise issue, very smooth tube sound, sct800 has a bit more agressive tube colloration but overal amazing tube mics collection for about 1000€ investment.
A question is what is benefit if I replace stock PCB's with C12, U47, U67 pcb clones available?

Thanks

 

[ruffrecords]

CCDA attempts to create equal and opposite signal currents in its two stages such that the overall HT current draw is constant.

The potential advantages are improved PSRR and separation of the gain stage from the transformer driving output stage.

In practice, the average current in a class A stage is already constant (the current just flows through the tube and through the load in varying proportions). In a properly decoupled stage, the actual variation isn total HT current draw will be minute, any local current variation being supplied by the decoupling capacitor.

A double triode CCDA design will be 3dB noisier than a single tube stage and 6dB noisier than a single stage using the two triodes wired in parallel.

Like all engineering its a compromise.

Cheers

Ian

 

[kingkorg]

any local current variation being supplied by the decoupling capacitor.

This here is gold, and so obvious, but i wasn't aware of it at all. Does this make case for larger value decoupling capacitors than typical 1uF?

 

[Nen'O]

Really like CCDA mics I have. My impression is that CCDA leaves Capsule to do the work with a touch of tube coloration
-I was told to use 2.2uF for C8, works great.
This is the explanation Dave gave me:
'This way any resonance between the inductance of the output transformer and C8 falls into the sub-sonics below the audio range.'

 

[ruffrecords]

This here is gold, and so obvious, but i wasn't aware of it at all. Does this make case for larger value decoupling capacitors than typical 1uF?

Yes it does.

Cheers

Ian

 

[Emmathom]

Well I don't see any difference between a CCDA and a classic CF (without coupling cap nor resistor...)
Could someone point the difference please ?

 

[Khron]

Well I don't see any difference between a CCDA and a classic CF (without coupling cap nor resistor...)
Could someone point the difference please ?

They're the same thing, but they're not - semantics, one might argue

CF = cathode follower refers ONLY to the second triode (that's used with only a large cathode resistor, no anode resistor, and the grid tied directly to a DC bias source and whatever signal source).

CCDA = a gain stage plus a cathode follower; since the CF is unity gain and non-inverting, but the gain stage is inverting, the total supply current seems to end up roughly constant.

Even the link from your first post says so (emphasis mine):

"As the title indicates this configuration shown below is not a "Cathode Follower" but actually a composite amplifier, consisting of two triodes"

 

[Emmathom]

Ok thanks @Khron ! CCDA is a gain stage followed by a CF... It's true that a (pure) CF could be "technically" used somewhere else (like for ex. after a Jfet gain stage...)

 

[Khron]

Ok thanks @Khron ! CCDA is a gain stage followed by a CF... It's true that a (pure) CF could be "technically" used somewhere else (like for ex. after a Jfet gain stage...)

https://www.valvewizard.co.uk/dccf.html

https://www.valvewizard.co.uk/accf.html

Sure, technically you could stick a tube after a JFET, but... why would you?

 

[rock soderstrom]

A double triode CCDA design will be 3dB noisier than a single tube stage and 6dB noisier than a single stage using the two triodes wired in parallel.

I wonder why you hardly ever see two triodes wired in parallel in mics. Is the theoretical advantage lost elsewhere, so that the 3db less noise compared to a single triode (as AF) doesn't come into play, or are there other disadvantages? Do the two triode systems have to be matched or does the Miller capacitance cause problems?

 

[rock soderstrom]

Ok thanks @Khron ! CCDA is a gain stage followed by a CF... It's true that a (pure) CF could be "technically" used somewhere else (like for ex. after a Jfet gain stage...)

You can also use just one cathode follower as a head amp in your mic. There are many examples of this. The head amp is primarily an impedance converter, the gain does not play as big a role as you might think, as this is almost eliminated in the necessary output transformer with a high step down factor.

 

[Emmathom]

You can also use just one cathode follower as a head amp in your mic. There are many examples of this. The head amp is primarily an impedance converter, the gain does not play as big a role as you might think, as this is almost eliminated in the necessary output transformer with a high step down factor.

You mean adaptative impedance ? as a CF is a x0,9 "gain stage" (if we can call it so since it lowers the gain)

 

[rock soderstrom]

adaptative impedance

Sorry, I don't know that term.

The situation is quite simple if you take a closer look. Let's take an example, a 5654/6AK5w tube based microphone with a 10:1 output transformer is estimated to have a total system gain of a few dbs. This is not extremely much more than a CF with its negative gain of almost 1. However, both circuits fulfill the primary goal of bringing the ultra high impedance of the capsule into an acceptable range.

The cathode follower can cope with a much lower step-down transformer, in fact it can even be omitted if the cable runs are not too long!

 

[Khron]

as a CF is a x0,9 "gain stage" (if we can call it so since it lowers the gain)

The cathode follower can cope with a much lower [ratio] step-down transformer, in fact it can even be omitted if the cable runs are not too long!

THAT is the key point to keep in mind. Single triodes need 6.5:1 (U47) - 12:1 (C12) ratio step-down transformers, to get the output impedance in the low-hundreds-of-ohms. Cathode followers (or CCDA circuits, or PNP emitter followers in solid-state circuits) most often have "just" 2:1 - 4:1(?) ratio output transformers.

 

[Emmathom]

Sorry, I don't know that term.

impedance matching...

 

[Cqwet Dbdfte]

Impedance matching, not really, load "matching" may be a better term.
Or not overloading the source.
It is OK to drive 150 Ohm load with a 0.2 Ohm source, thru a 600 ohm cable, as only voltage is of interest.
In the RF world matching Z means same source, cable, and load impedances to preserve power, in digital to prevent reflections and lost data integrity.
Else bad things can happen, coax catching fire, what not.
https://electronics.stackexchange.com/questions/35741/impedance-matching-confusion