C12A Sony C37a Hybrid

Can anyone comment on the tube input impedance criteria for microphones?

tardishead said:

Can anyone comment on the tube input impedance criteria for microphones?

Click to expand...

This is something I don't really yet understand. Grid impedance is generally assumed to be infinite so long as the anode voltage is high enough (which it is, in a tube mic).

You need to hold the DC point of the grid at some bias relative to the DC point of the cathode but at the same time we want the impedance to AC ground of the grid to be super high because, along with the tiny capacitance of the capsule itself, this forms an HPF on the audio. Hence the need for 100s of Megs or more of resistance in the grid leak path. If there are tubes out there (I don't know if there are or not) which have a grid impedance of a similar order of magnitude then this is going to mess with our LF response also as it will be in parallel with our grid leak resistor. It was my mistake with your EC70 to see a max resistance figure on the spec sheet and mis-interpret it. It is the recommended max resistance to put on the grid, not the actual resistance of the grid itself.

Oliver has this 60x rule of thumb. I'm not sure where that comes from, but I suspect it is the best compromise between two extremes - too high an HPF versus too much noise from spurious grid impulses. I think there's also an issue with tubes and with FETs that if the grid/gate resistor is too high then the device won't fire up properly in the first place, it won't actually "see" the bias.

The grid sucks a tiny current inwards, micro-amps or less. More so when the plate is starved (low voltage), less so when operated in normal plate scenarios.

The grid must be significantly negative to repel grid current. (Plate voltage helps, but is usually ample.)

For guitar-amp and such work, you can model a naked grid as >200Megs DC, falling to 200K (Miller effect) at the top of the audio band. If critical, better put your thinking-cap on.

But in context of a capacitor mike head-amp, rough approximations may not be enough. OTOH you see the oddest tubes in cap-mikes, so it can't be so very critical.

Decided to build a C37 more simple circuit. In the process discovered that the output xlr had a dry joint - that had been my problem all along. Doh
Anyway the C37 sounds great. Really natural. HK47 capsule with a 6AU6WA which is quiet enough I think. I'll burn in some tubes after I've had a bit more use of the mic. Dropped the B+ to about 165v to create 82v polaristaion on the capsule. Not as much headroom as an original but I dont think theres much chance of clipping it.
I love the sound of it. Does remind me of the C37a pair that I had borrowed but not as bright. This circuit kinda softens fast transients like drums. Klaus Heyne and others who dont like the C37a complain about its smeariness of transients. I've read that in a cathode follower the square wave fall time is not as fast as the rise time. Would this be an explanation of what it does to transients? It might be super subtle but hey the ear can pretty much detect any slight nuance even when the listener cannot describe it.
I definitely think it has its place in recording. Just think of Hal Blaine and Earl Palmer behind the drums of the Wrecking Crew sessions. Always had C37a overheads and often U47 on kick (shock horror).

I'm working on a friends C37a right now.  Its an odd one.  It works and then after about 20 minutes of use it starts to loose level then quits working.  the voltages all measure ok but it seems like some kind of high voltage leakage problem.  still sussing it out.

The sound is great to me.  The softer transients is precisely why I like it.  On Acoustic Guitar it just rules for me.  I flat pick and can have an overly bright top end due to a firm picking hand.    I find the sony makes that sound beautiful and I really notice I can play how I would like without having to soften my attack.  On my 451 AKGs I have to lighten the stroke or it gets to bright.    Its all good but I find the problem Klaus has is what I like about the mic.    Its not dull its just got a nice plunk to it without to much bite.

I also have a Klaus Heyne Neumann M249 from Germany that I purchased from him many years ago.  I appreciate his knowledge of microphones but once you get a good tube in a sony 37a,  it has a lot of places to be used, especially in todays digital DAW sound of today.  It seems like I want to slow down the transient sound a little hence the use of ribbon mic/s, tubes and such. 

Hello All ,
                  This thread caught my eye and have been following , very interesting and Great information , I've wanted to do a C12A ,the schematics I have don't specify an output transformer / ratio /  specs , was hoping someone here might make some suggestions,

Thanks 

Sony c37a uses 10k:200
C12a is 4:1
1:4 input transformer will work as well backwards. Can be small because it does not have to handle much level.
Utc a24/25
Loads of them take your pic
Neve 31267 will work.

Thanks so much  !! 

I'm really enjoying this mic for a lot of different applications
Eric's Hk47, 6au6wa about 170ht 80v capsule triad hs50 output transformer
I'm really enjoying it on loud sources doesn't have that same coloration as on typical grounded cathode designs.
Sounds really big and crystal clear and has an interesting transient response

tardishead said:

Sony c37a uses 10k:200

Click to expand...

Where did you get this information? Not long ago I measured in and out original C37a transformer. From memory the ratio was 4.25:1, unlike 7:1, as your post suggests.

Best, M

Sorry not entirely sure
I should keep a track of web resources
Normally I just collect info and move on
I use a triad hs50 which works really nice

tardishead said:

Klaus Heyne and others who dont like the C37a complain about its smeariness of transients. I've read that in a cathode follower the square wave fall time is not as fast as the rise time.

Click to expand...

That is not true in normal (linear) operation. It is true that the Slew-Rate is different because generally, the tube has the capability to pass much more current than the cath resistor. So unless you are overdriving the mic's head amp, there should be no noticeable difference between positive and negative going edges. BTW, the problem is the same for common-cath stages, with polarity reversed.
I wouldn't give much credit to Klaus Heyne's technical interpretations; he's a great microphone historian, but his technical knowledge is tainted. He also has a "funny" concept about cathode bias vs. gris bias. I believe his technical shortcomings are the reason he has not been retained as a moderator at the time of the switch from rep  to prw.

tardishead said:

Can anyone comment on the tube input impedance criteria for microphones?

Click to expand...

Just to add on Matt's and PRR's comments.
The reactive part of the impedance is essentially capacitive, with or without Miller effect (common-cath) with or without bootstrapping (cath-follower). Since it is in parallels with the capsule, it produces some attenuation. With SDC, the effect will be the most, less on LDC.
The non-reactive (resistive) part depends largely on the purity of vacuum.
It is always desirable to target for the highest possible bias resistor, not only because it defines the LF extension, but mostly for increased noise performance.
The intrinsic noise performance of a capsule depends on the spectrum of noise density. Moving the noise density towards VLF makes it less audible. That's why modern (FET) mics have an electric cut-off at a few Hz.

BTW, bootstrapping does extend the LF response but does not decrease noise.

With tubes, it is seldom possible to use reliably grid resistors of more than 300Meg, because grid current pulls the grid excessively negative, and the tube goes into cut-off. It may be possible to compensate with positive bias, but grid current in these conditions is not stable.

tardishead said:

Personally I think cathode followers are very useful and can sound huge.

Click to expand...

There is a good article on this one. It also explains some sonic limitations of some CF mics, such as C12A, or C37A:

http://www.tubecad.com/2005/June/blog0049.htm

Best, M

It also explains some sonic limitations of some CF mics, such as C12A, or C37A:

Click to expand...

Really? It just explains how to use them properly. Do you think these CF mics are designed improperly? And if so can you explain why

tardishead said:

It also explains some sonic limitations of some CF mics, such as C12A, or C37A:

Click to expand...

Really? It just explains how to use them properly. Do you think these CF mics are designed improperly? And if so can you explain why

Click to expand...

I don't want to speak for Marik, but there is a point of discussion there.
Since the CF has less gain than a common-cath stage, designers may have taken advantage of the lower output Z and  opted for transformers with a higher ratio (4:1 IS higher than 10:1  ) with two incentives: one to recoup some of the gain, the other to make the transformer easier to make, thus cheaper, and better.
However, the peak current demand on the tube may be an issue, particularly if the load impedance is on the low side.

However, the peak current demand on the tube may be an issue, particularly if the load impedance is on the low side.

Click to expand...

Which results in ?? Compression and harmonic distortion at high SPL into low impedance tube preamps? Compared to common cathode same SPL same mic preamp surely both introduce distortion?

tardishead said:

However, the peak current demand on the tube may be an issue, particularly if the load impedance is on the low side.

Click to expand...

Which results in ?? Compression and harmonic distortion at high SPL into low impedance tube preamps? Compared to common cathode same SPL same mic preamp surely both introduce distortion?

Click to expand...

This will happen sooner because of the higher ratio of transformer. More ratio-> more current demand for a given output.

Ok so CF  mic (with lower winding ratio transformer) performs better with a solid state mic pre (2k) rather than a tube mic pre (200R). Its more load dependent
But that is given that bias point, B+, I/V relationship is the same yes? C37a is biased a lot colder (3k) than C800G (1k) so comparison is not so straight forward.
These 2 mics provide the perfect comparison because they use the same tube (although different capsule).

A question I've often wanted to ask is how can you calculate the AC input voltage at the grid of a tube from a given capsule at a given SPL? I cant seem to find any equations.

Also the harmonic distortion signature will be different between a C37a(more 3rd) and a C800G (more 2nd). Would this explain why some engineers love the C37a for toms and other high SPL sources? 3rd harmonic (in small doses) provides a nice edge for toms and accentuates the impact transient whereas 2nd harmonic seems to have a thickening effect but softens the impact transient. Both have their place in creative recording techniques. I'm an engineer firstly but I try to understand what is going on technically. Its fascinating to see that the imperfections of certain equipment can be used as a creative tool