Neumann K87 and K67 capsule capacitance

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Yes, but usually "circa" ("ca,", latin for approximately) is non capital letters and no points... "approx," is probably a later translation you can see in Gotham distribuited schematic.
 
No sir, KK87 and K67/KK67 are 50pF per side... but, as said, do not really matter.
Fwiw...

"Measurement note: the “ideal” capacitance value is provided by the manufacturer. The “actual” value was recorded by me, measuring the capacitance of each diaphragm to backplate, individually, then averaging the two."

http://recordinghacks.com/capsules/
 
It says that in the schematic but that's not true in real life. It seems like either they started out with that capacitance during r&d but they decided to decrease the distance between the backplate and the diaphragm later on and didn't update the schematic, or maybe the schematics just never had accurate capacitance. Either way, I have a very early capsule here with the wire mount and even it has a capacitance of 67. My later k87s have capacitances between 65 and 75.
I have NOS K67's here that are 53pf or less
 
I have NOS K67's here that are 53pf or less
really! how old? do you think they changed it at a certain point in time? I've seen a few really old broken ones that had 50um spacers under the diaphragms (and in the middle) but they were broken so I couldn't get capacitance on them. are there capsules with even thicker gaps?
 
Last edited:
Mine are fairly old. The K67 was first made with brass rings, then fiberboard, then plastic. At different stages I have had all three types but the last were plastic rings and still measured low 50's.
 
Mine are fairly old. The K67 was first made with brass rings, then fiberboard, then plastic. At different stages I have had all three types but the last were plastic rings and still measured low 50's.
Interesting, I've had a few brass k67s in that I am sure weren't reskinned and they were all in the high 60s. Maybe it's a difference of how we're measuring. The oldest I measured recently I measured at 68 on the working side and it had plastic pegs between the backplates to align them. The gray on the screws was undisturbed so I don't think it was altered in any way. Pretty consistent at high 60s-70 through all the types for me...Maybe my method is just off by about +15pF? wires? parasitic capacitance? measuring small capacitance is hard and i am stupid.
 
Last edited:
I don't really understand the difference we are seeing. Any good capacitance meter can read 50pf so I believe your readings. I know that higher capacitance on these would make them collapse using 60-70 volt polarization. The only way to prevent this would be higher tuning freq.
 
Ok, I think the issue about Neumann "litte" capsule capacitance values in old schematics resides in measurement conditions: we all know that capacitive reactance is inversely proportional to frequency. Probably Neumann's engineers in the '1940s till '1970s measured their capsules with very high frequency equipment, such as MHz, and this was common for people who worked in the military (radar was developed well in the '1930s...), Bear in mind that Telefunken VF14 (U47) was an RF pentode designed for military radio communications, and Ac701k triode (KM56, M49 etc...) was itself developed for use in measurements instrumentation... so engineers were used working with high-end testing standards. As stated before, the clue is to know precisely the measurement conditions and test gear they used as standard, because I really do not believe they were so imprecise in their schematics!

P.s: when you read the U89 schematic (1990, so modern measurements standard) magically capsule capacitance is indicated as a real 80pF... so, no extimations or 30%error approximations in Neumann labs!
 
Last edited:
I was positively impressed by this statement:
I have NOS K67's here that are 53pf or less
This is exactly my experience... so, I'm wondering what test rig does Tim Campbell uses? I myself have access to an old Bruel&Kjaer lab rig... maybe THIS is the real clue: no chinese multimeters working at max 100KHz here...
 
Last edited:
My test setup is lower frequency so I think we have our answer! From what I've seen my measurements are consistent but perhaps not accurate, And the same is probably true of the many others who have checked this. That'll teach me to cheap out! Luckily I did all of my r&d on the same equipment that I did my initial measurements so the capacitance on my capsules should technically still be right. The readout numbers are just wrong...🤣
 
Last edited:
Yeah I repeated a few measurements, replaced the wires, moved things about etc and it only varied 1pf or less, so the precision is good. That's good. It's just not accurate. Scared me for a second there. This is my mistake, I'm just glad it probably doesn't affect anything because all the mistakes canceled out 😅
 
George if you notice your schematic is from 1960 the first year of production. The Gotham one posted appears to be from 61 so it looks like they revised the capacitance. Also I notice on the Gotham schematic the head assembly is called KK67 as we know it but on yours the head assembly is called KK60. Perhaps they originally used a slightly different capsule.
.Screenshot 2023-03-31 at 18-38-15 Neumann K87 and K67 capsule capacitance.png
 
Last edited:
George if you notice your schematic is from 1960 the first year of production. The Gotham one posted appears to be from 61 so it looks like they revised the capacitance
.
Yes.

I mainly wanted to show that it was 75 in this earlier document without any intention of drawing conclusions or suggesting any.

It’s an open question who “they” is considering the later document is Gotham, whereas the one I uploaded is strictly a Neumann document.

One thing that did also come to mind is that there is supposedly an early prototype U60 that has a prototype edge terminated capsule. I view that as a little dubious, or unverified, but it could possibly also come into play with the value discrepancy. Probably unlikely.
 
i'm going to come back to this in a week with different equipment and re-test my references and prototypes. i'm fairly certain that it's just the difference in frequency changing the readout, but it never hurts to double check. i had wondered about this sort of thing, which is why i made sure not to change setup in between measurements of samples and prototypes. precision is more important, but i guess i didn't expect the readout value to be off by this much.
 
Last edited:
One thing that did also come to mind is that there is supposedly an early prototype U60 that has a prototype edge terminated capsule. I view that as a little dubious, or unverified, but it could possibly also come into play with the value discrepancy. Probably unlikely.
That would make sense. The U89 capsule is edge terminated and higher capacitance.

i'm going to come back to this in a week with different equipment and re-test my references and prototypes. i'm fairly certain that it's just the difference in frequency changing the readout, but it never hurts to double check. i had wondered about this sort of thing, which is why i made sure not to change setup in between measurements of samples and prototypes. precision is more important, but i guess i didn't expect the readout value to be off by this much.
I wouldn't worry about this so much. Your capsules sound and measure correctly and that is the important thing.
I am always amazed by people that get distracted by the wrong minutiae. One guru swears that capsules sound different by what rings hold the membrane, others about the sounds of threads.
 

Latest posts

Back
Top