Question about capsule and capacitor

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

KrIVIUM2323

Well-known member
Joined
Jul 23, 2005
Messages
364
Location
France
Hi,
Lately i was looking at schematics of different classic mic (m49 b/c, c12,...) and i noticed something i don t get: in the m49b or c there is c1 which bridge both diaphragm together when in multi pattern mode, the same capacitor appear in other mic (g7) but is not present in the c12 schematic for example.

I see it bridge both side of the diaphragm but i don t really understand the purpose of it. At first i thought it was related with the biasing scheme of capsule which is different between most Neumann and Akg c12 but the G7 which borrow the biasing of c12 use one too.

Could someone explain me what it does in practice and why it is needed?
Thanks.
 
Feels better not to be alone!  ;D
In fact the way i see a condenser capsule is like a common plate shared by two variable condenser, abit like these electrolytics you see in guitar amp psu (except these one are fixed value not variable). In that case i ve never seen this kind of bridged cap and as they usually deals with quasi dc or dc i think this is ac related in the case of capsule... but really don t get what to be gained. But im open to stand corrected.

In fact this really is a mystery for me.As for value as this kind of cap vary from 1n to 1000n with the different one i ve seen in schematics. Even seen some comments of Neumann engineer telling the value is not really critical, but i may have not understood what was meant.
 
Oh men...

The best cardioid is pure "cardioid" (like U47 and M49b/c - with use of the switch S2) - completely without use of back diaphragm.
There's also C414E type of topology (for now the best sounding for multipattern) but it needs negative voltage - most of capsule can't work here with high voltages above 100V.
There's also ELA M251 topology which is pretty good but need separated backplates.
Cathode follower (currently working on it) can be used without any capacitor for multipattern.
G7 topology is odd, because it utilize two different types of connection, but it can use only one.
Each topology have it own limitations.  Worse response or/and lower level comparing m-p cardioid with pure cardioid.
The higher values of high ohm resistors and higher values of capacitors =  worse response.
Usual 470pF - 1nF is enough (especially in M49 topology) - some capsules and circuits allows to use lower values.
Not without reason C414E use 4.7M resistors ;)
All depends on the circuit, tube or FET and capsule.

 
Ehrm... I'm pretty sure the question was NOT about the "front diaphragm to JFET gate / tube grid" coupling cap, but the one between the front diaphragm and the rear one ;)

Or... Hmm... Looking at an M49 schematic, is that supposed to be the "rear diaphragm coupling cap", somehow?  ???
 
I'm writing about overall input topology concepts.
I'm not writing about U87 type (read totally not needed) input grid/gate cap.
C12 mentioned in the first post have input capacitor in series with backplate  and grid (worst solution btw.) to skip that type of connection you can use M49 type of input - it's much better but still not perfect.
When is perfect? When you can compare m-p cardioid to so called "pure cardiod" and response as also sensitivity is similar at least in 80%.
Cap in M49 is to get connection between both diaphragms without loading grid vith high voltage as also to make flipped phase between front an back.
C414E type of topology is almost perfect, but utilse negative voltage to get phase difference. Notice that both diaphragms are polarised and the signal is taken from backplate. Since you mentioned grid cap - here's also not needed as in U87 ;)
 
Kriv

It's just a DC blocker to keep the double voltage off the front side of the capsule, to get a figure of 8 the beck capule needs to be polarised at twice the voltage as the front. If you didn't have that capacitor you'd end up with a super charged capsule with 116 volts on either side. The M49s is a weird one because you can switch it internally  to be a series connected omni. I guess this was so the studio could decide - if it was planned only orchestral work then the pattern selector was unecessary and both capsule are linked via the switch like a u47.

The reason its not there in the C12 is because the c12 polarises the diaphragm not the back plate and the signal is taken from the backplate not the diaphragm so the two never need to be connected - it could be argued that omni on the C12 isn't true omni because of the two separate voltages on the two diaphragms even if they're identical. theoretically they can never be 100% as one, but the plus side is it isn't lossy like the omni U47 so the output isn't decreased

The M49 really was a redesigned 47 Neumann got a lot of complaints at the time regarding the u47 - the body was too big, you could either have figure 8 or omni, not both, the tube was pretty much obsolete by 1955, at the same time AKG had come out with the C12 which is just about as perfect as you could get and they'd modified it to be the ELA-M for Telefunken, and Neumann didn't want to loose the market.

So here was the M49 aU47 but better, or so they thought.
 
Out of interest here is a BBC document from 1954 detailing the U47 it's interesting that this tech notion believed the U47 was a new cheaper version of the M49 and he was clearly working with a shorter body model so it must have been marketed this way in Britain at the time.

http://downloads.bbc.co.uk/rd/pubs/reports/1954-23.pdf
 
Timjag said:
Kriv

It's just a DC blocker to keep the double voltage off the front side of the capsule, to get a figure of 8 the beck capule needs to be polarised at twice the voltage as the front. If you didn't have that capacitor you'd end up with a super charged capsule with 116 volts on either side. The M49s is a weird one because you can switch it internally  to be a series connected omni. I guess this was so the studio could decide - if it was planned only orchestral work then the pattern selector was unecessary and both capsule are linked via the switch like a u47.

The reason its not there in the C12 is because the c12 polarises the diaphragm not the back plate and the signal is taken from the backplate not the diaphragm so the two never need to be connected - it could be argued that omni on the C12 isn't true omni because of the two separate voltages on the two diaphragms even if they're identical. theoretically they can never be 100% as one, but the plus side is it isn't lossy like the omni U47 so the output isn't decreased

The M49 really was a redesigned 47 Neumann got a lot of complaints at the time regarding the u47 - the body was too big, you could either have figure 8 or omni, not both, the tube was pretty much obsolete by 1955, at the same time AKG had come out with the C12 which is just about as perfect as you could get and they'd modified it to be the ELA-M for Telefunken, and Neumann didn't want to loose the market.

So here was the M49 aU47 but better, or so they thought.

Almost each line you are wrong :D ;)

It's not just DC blocker - you need to get flipped phase - if you will use to low capacitor value then you will get practically one pattern on each selection - sound wouldn't change.

M49 isn't weird but brilliant! Where you get that "omni"?
S2 is the switch for cardioid ("pure cardioid") operation only - disconnecting completely back diaphragm from the circuit.
Of course PSU switch need to be set for 0V. Sounds much better comparing both options of cardioids as aslo sensitivity is higher.

C12 have exactly permanently polarised  backplate and 1nF - input capacitor - here's as a "dc blocker". It doesn't need capacitor  between diaphragms because front is permanently grounded (read 0V) so the phase is flipped. 

U47 type of omni implementation have lower sensitivity, but it's much more "omni" than in other impelementations - measured in anechoic chamber ;) 

M49 wasn't redisigned U47 - it have different origin and it was just differnt model.
 
Timjag said:
Out of interest here is a BBC document from 1954 detailing the U47 it's interesting that this tech notion believed the U47 was a new cheaper version of the M49 and he was clearly working with a shorter body model so it must have been marketed this way in Britain at the time.

http://downloads.bbc.co.uk/rd/pubs/reports/1954-23.pdf

Not exactly but in fact it was a cheaper "head amplifier" :D :D :D
 
Guys,
thank you all, i got it now and... feel a bit like i had a brain fart!  ::)

Ln76d:
Cap in M49 is to get connection between both diaphragms without loading grid vith high voltage as also to make flipped phase between front an back.

About the m49 if the back membranne isn't connected to the front one through c1 (1nf) there is no multipattern possible as there is no mix of front and back signal so only 'pure' cardioid (front side of capsule) pattern availlable.

I think i've been misguided in studying the G7, as i was never bothered to see that same topology (the m49  use) in the u48 schematic.  ???  :p

Ln76d:
The best cardioid is pure "cardioid" (like U47 and M49b/c - with use of the switch S2) - completely without use of back diaphragm.
There's also C414E type of topology (for now the best sounding for multipattern) but it needs negative voltage - most of capsule can't work here with high voltages above 100V.
There's also ELA M251 topology which is pretty good but need separated backplates.
Cathode follower (currently working on it) can be used without any capacitor for multipattern.
G7 topology is odd, because it utilize two different types of connection, but it can use only one.
Each topology have it own limitations.  Worse response or/and lower level comparing m-p cardioid with pure cardioid.
The higher values of high ohm resistors and higher values of capacitors =  worse response.
Usual 470pF - 1nF is enough (especially in M49 topology) - some capsules and circuits allows to use lower values.
Not without reason C414E use 4.7M resistors ;)
All depends on the circuit, tube or FET and capsule.

Ok. I've had to read it multiple times but this is enlightning for me about other things i've had wandering.
I now do understand why the k67 backplate is used in so  many clone of ck12: the need for separate isolated backplate, using an 'easy' to machine/reproduce design.

I'm working for some time on a CF headamp too and came to same conclusion as you.  ;D

Now about the value of this diaphragm bridging capacitor: it seems counter intuitive to me that the lower the value the better the response. Could you expend on this a bit please?

Ln76d
I'm writing about overall input topology concepts.

And i thank you for doing this!

C414E type of topology is almost perfect, but utilse negative voltage to get phase difference. Notice that both diaphragms are polarised and the signal is taken from backplate.

Ok, i ve read many times this kind of comments from you about the c414 topology. I don't know the circuit but i'm interested in understanding it. When you talk about negative voltage are you talking about relative negative voltage to some biasing or real negative rail?

Are you talking about this schematic:
http://gyraf.dk/schematics/AKG_C414_old.GIF

If yes could you explain further as i don't see where there is negative voltage in there?

Timjag:
to get a figure of 8 the beck capule needs to be polarised at twice the voltage as the front.

Please excuse me but i don't follow you there... In m49 in fig of 8 you have backplate at 58 or 60v, back membranne at 116v or 120v and front at +2.4v. The polarisation voltage is not twice the front membrane, it is twice the backplate. ;)


Timjag:
The M49s is a weird one because you can switch it internally  to be a series connected omni. I guess this was so the studio could decide - if it was planned only orchestral work then the pattern selector was unecessary and both capsule are linked via the switch like a u47.

If i remember correctly the m49 is the first mic with remote pattern control and i think the switch made a comeback once the Neumann engineer whitnessed that in 'pure cardioid mode'(or u47 mode) there was a difference of 4db of sn. The thing which i find interesting is that in the  latter 'c' version the switch is omitted in the schematics i have.

Timjag:
The M49 really was a redesigned 47
As is the u67... the 87, the... ;) I'm kidding about the 87 but most of the tube circuits are evolutions upon the 47, or are presented like this in the Neumann history i've read.

Timjag:
The reason its not there in the C12 is because the c12 polarises the diaphragm not the back plate and the signal is taken from the backplate not the diaphragm so the two never need to be connected - it could be argued that omni on the C12 isn't true omni because of the two separate voltages on the two diaphragms even if they're identical. theoretically they can never be 100% as one, but the plus side is it isn't lossy like the omni U47 so the output isn't decreased

I agree.

Timjag:
The M49 really was a redesigned 47 Neumann got a lot of complaints at the time regarding the u47 - the body was too big, you could either have figure 8 or omni, not both, the tube was pretty much obsolete by 1955, at the same time AKG had come out with the C12 which is just about as perfect as you could get and they'd modified it to be the ELA-M for Telefunken, and Neumann didn't want to loose the market.

So here was the M49 aU47 but better, or so they thought.

The complaint came from the film and broadcast (tv) about the size... which is ironic if you ever seen an Elam! The c12 was indeed smaller in form factor, but to my knowledge in EU most used the Elam rather than the c12 in film or broadcast (mainly because of the 'standard' ac701 tube and psu compatibility).

Anyway i really thank you! I learned a lot and you made things clearer for me.

Timjag:
You're barking mad. :eek: ;)

I agree too! ;) But Ln76d is mainly right even if definitely biased upon some subject.
 
Note the directions of the rectifier diodes in the DC-DC converter ;)

KrIVIUM2323 said:
Ok, i ve read many times this kind of comments from you about the c414 topology. I don't know the circuit but i'm interested in understanding it. When you talk about negative voltage are you talking about relative negative voltage to some biasing or real negative rail?

Are you talking about this schematic:
http://gyraf.dk/schematics/AKG_C414_old.GIF

If yes could you explain further as i don't see where there is negative voltage in there?

But i would guess it's a bit of both, as in, both absolute negative voltage, as well as relative negative voltage (depending on the context / "how you look at it").
 
Ok, i spend a little time looking at the 414 schematic,
so, transistor and other components at left of trasnformer form an oscillator, the output is then send to the transformer which step up the waveform which is then send to a serie of voltage doublers.
Two of them are 'positive' one of them is 'negative'.
The 'negative' one bias the rear diaphragm, the front one is either grounded, 'negative ' biased or biased by the 'positive' ones.
Thats it?
You are right Khron, it is both relative and a real negative rail.

I must confess i m a bit obsessive about glass and vaccum for microphones, much less about sand... :)
So it is not really a dc to dc converter, much more an ac to dc converter. What is the frequency the oscillator is running? And what are the expected output voltage?
Same range as for tube mic? About 60v/90V/120V of relative bias?
I was going to ask why using a transforer for step up but this must be a p12 powered mic so the need for a step up and probably galvanic isolation for rest of circuit... Clever circuit design.

Ok so Ln76d, why a negative bias is so superior to a positive one in the case of a mp mic? After all this just a reference swap no? What iam missing?
 
No, it is very much a DC-to-DC converter ;) The phantom-power isn't AC, is it? And could R3 and C3 perhaps form a low-pass RC filter, to smoothe the DC? ;D And there's that Zener diode there too...

Not quite sure about the AKG ones, but the U87 and Schoeps-type oscillators are switching around the low MHz range (i've heard of 2MHz or so?), and are usually adjusted to put out +/-60v (relative to ground, ie. 0v).

Not sure about the requirement for galvanic isolation, but i have my doubts. The U87A DC-DC stage definitely doesn't use an "isolation transformer" (so-to-speak), but rather a coupled inductor. The Schoeps variant, at least the one found in Rode as well as Chinese condensers and whatnot, makes do with two separate inductors.
 
Ok dc to ac to dc. Let s call it a Dc to Dc.  ;D

Mhz range ok. It seems enough out of audio range not to worry to much. You are probably riht abut the inductive properties, make more sense at this frequency.
Thank you Khron and all, i ve learned a lot today.
Time to sleep, less dumb than yesterday.
 
That cap let's you tie the two halves of an isolated backplate capsule together (in parallel) when the two sides are sitting at different polarization voltages.

There are many ways to polarize a capsule to get different directional modes, and as a result you get different topologies / connection methods.
 
That is ok Khron.
It is just that as powerlines are ac then rectified and filtered... ;)
Butt yeah this is dc to dc, i agree (works with a battery).
 
Back
Top