Neumann Vintage U87 Clone : Build Thread.

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
poctop said:
+1 For Geekmacdaddy,  i specially like the place when we hear " Its Hotter Now"  8)

Good Job ,  thanks For you nice Picture,

Dany,
Best,

Thanks to you, my solder sniffing brother, for all of your hard, brilliant, wonderful work. The feeling of accomplishment in orchestra with beating the system by about 2,800.00 USD is quite envigorating. I'm all over the '67 when it's ready. 2 probably. Sniff on!
 
Went back in, Swapped out the 220pf (C6) for a Wima 100pf, and hopefully biased at 13.7 volts and here's a clip

https://dl.dropbox.com/u/87086401/Clone1stTweak.aif
 
Geek,
You made the right move. It's totally sounding better and I don't hear any more distortion. What do you think?

Dave
 
I did think I heard something towards the end...back down to 10v....I needed to do this to match the real U87

ScreenShot2012-08-09at74304AM.png
 
That's strange, I really don't hear the bump in the 1k that you do.
I also liked the way your tweak sounded better than the original its smoother. And as for biasing  the scope its really the best way to be sure!
 
Hi Guys,
Ok so I have not used my Mic in a bit and when I took it out today I noticed that the capsule was leaning to the left quite a bit because of the way it was in the case and how the case was stored. 

My question is what kind of mounts are you guys using in the aurycle body?  Obviously this plastic mount that comes with it bend easily and doesn't hold its form.

Would I be able to get one of Peluso's mounts in there?

What are you guys using?

-Scott

 
Hi Scott,
I'm using the stock Aurycle mount but I noticed mine wasn't molded correctly. One half of the mold was a bit higher than the other half so I took my Dremel to it with a sanding drum. After I sanded it flush, the mount holds the capsule correctly.

Dave
 
Can a noob ask for a little troubleshooting help? I got the mic all put together but I'm just getting a faint sound. I have to turn up the gain all the way before I hear anything. And then it's just handling noise. Or I can tap on the head basket and hear the tapping sound. I checked all of the voltages before I hooked up the trafo and capsule and everything looked great. I had trouble biasing the Fet since I had never done it before but after several tries I finally got even clipping on the scope. So now I don't know where to take it. Any ideas?
 
Cgaines said:
Can a noob ask for a little troubleshooting help? I got the mic all put together but I'm just getting a faint sound. I have to turn up the gain all the way before I hear anything. And then it's just handling noise. Or I can tap on the head basket and hear the tapping sound. I checked all of the voltages before I hooked up the trafo and capsule and everything looked great. I had trouble biasing the Fet since I had never done it before but after several tries I finally got even clipping on the scope. So now I don't know where to take it. Any ideas?

What capsule are you using? Does it have the isolated dual backplates?
 
Here is some data I took on a few FET's.  The plan is that I'll mock up the front end of this circuit so we can see how these parameters play into the operation of the circuit.

I started by breadboarding up a simple test rig:

test_setup.jpg


You first dial in the pot (a multi-turn precision type works best for this so you can dial in voltages accurately) to show 0V on the voltmeter.  This should happen roughly half way through the range of travel.  You can then connect the JFET and read the IDSS on the ammeter directly.  Note this value down.

Now we want to figure out VP:  dial the pot so that the voltage becomes negative.  As you make the voltage more and more negative, you'll see the current drop as the device becomes more and more pinched off.  What you want to do is lower the voltage until the current drops to 10% of it's IDSS value.  Note this negative VGS value down as well.  Repeat these two measurements for all the devices you want to measure.  Here are results for the first three JFET's I had in my stash:

measurements.JPG


Now VP is difficult to measure exactly, as it's defined as the gate-source bias that results in zero drain-source current.  It's difficult to measure no current, but we can use the characteristic equation of the JFET to get very close:

ID = IDSS * [1 - VGS/VP]2

If you solve for VP, you get:

VP = VGS/(1 - SQRT(ID/IDSS))

So you have all the above information from your measurements, which results in the following data for these three devices:

Code:
IDSS	ID	VGS	VP
-----------------------------------------------------
10.79	1.079	-3	-4.387425887
13.51	1.351	-3.41	-4.987040758
9.33	0.933	-2.45	-3.583064474

This way of calculating VP is not perfect but I've found it gets pretty close.  In general, IDSS and VP track together very well, meaning that the higher the current, the more negative the pinch-off voltage.

So I'm going to pick through 20 or so 2N3819's and 20 or so 2SK170BL's, and find one of each at the high and low ranges of these parameters, then I'll construct the front end of the circuit so we can see what the gain and input ranges look like.

As a primer:  gm (or transconductance) is defined as the delta change in ID with a delta change in VGS.  Mathematically, this is just the derivative of the ID equation above with respect to VGS.  The result is this equation for gm:

gm = | 2 * IDSS/VP * [1 - VGS/VP]

The first interesting result is that gm is a linear function of VGS (and the other parameters as well, which are generally constants).  People tend to think of transconductance as a constant but it isn't in this case!  It also means that the transconductance is maximum the closer VGS is to zero.  These equations can be found in any standard text on amplifier circuits BTW.

For the U87, the basic gain equation for the front end (assuming bypassed source resistor) is:

AV = - gm * RD

So gain is maximized when gm is maximized, which means we should keep VGS as high as possible and operate the JFET as close to IDSS as possible.
 
Hey Mike , this is awesome ,  i have to say i do like the Math and derivative, 
can't wait to see Gm versus Gain and go back in the lab with this ,
Thanks For taking the times to reveals such a wonderfull door as this one ,  8)
Best ,
Dan,
 
I agree with Dany that this is awesome. Honestly though it's a bit beyond me. Is there a simple explanation of what we want to
look for? That's some complicated math for me  :)

Thanks for doing this! Hopefully the penny will drop for me at some point.

Dave
 
HellfireStudios said:
Cgaines said:
Can a noob ask for a little troubleshooting help? I got the mic all put together but I'm just getting a faint sound. I have to turn up the gain all the way before I hear anything. And then it's just handling noise. Or I can tap on the head basket and hear the tapping sound. I checked all of the voltages before I hooked up the trafo and capsule and everything looked great. I had trouble biasing the Fet since I had never done it before but after several tries I finally got even clipping on the scope. So now I don't know where to take it. Any ideas?

What capsule are you using? Does it have the isolated dual backplates?


I'm using Dave Thomas' AK67 capsule
 
If your using a 67 style capsule it will only work in this mic as cardiod, How do you have it wired?

Eric
 
tskguy said:
If your using a 67 style capsule it will only work in this mic as cardiod, How do you have it wired?

Eric


Well I actually thought I could wire it as a multi pattern so that's where my stupidity lies... I added a fourth wire and wired it like the multi pattern should've been. Capsule wires in the appropriate places and the body wires in the appropriate places but reversed for the figure 8 pattern. How would I wire it for cardioid only?
 
Back
Top