Neumann Vintage U87 Clone : Build Thread.

[wave]

Dany,
AWESOME!. Eric and I were just talking about you doing this project and him offering a capsule for it. Are your ears burning? Haha!

Keep up the great work brother Dany!

p.s. I love the fact that I have 3 more mics to do right now as well as a couple rack units in the queue and I'm already mentally committing to this project before it's even off the drawing board.
In other words, put me down for 4 of the initial run.

Dave

 

[kpearsall]

FANTASTIC!!! I'm in the same position. 8 channels of API, 20 S800 EQ, barry porter EQ, 1084's, refurb 4 Ampex 351's, fix an SWR bass amp, and tons more but you know what? I've been waiting so damn long for the 67, EVERYTHING will get pushed back when it's ready. I remember falling in love with it before I even got to use one when I found out most of my fave vocals were recorded with it. T. Rex, Bowie, Nilsson, and I believe even Rufus Wainwright. After using one there was just nothing else I'd use instead if I had the choice Thank you Dan. You just became my hero.

 

[geekmacdaddy]

all of the vocals as well as a few other tracks were recorded with my u87 and my capsule.

Vocals sound great. Real nice drum sounds too, btw.

 

[kante1603]

Hi Matador,

I've seen at the bottom of the specs page1 that there are 3 types of the 2sk170 existing.

Note: IDSS classification
GR: 2.6~6.5 mA, BL: 6.0~12 mA, V: 10~20 mA

Any preferences?

Cheers,

Udo.

 

[dandeurloo]

Put your DMM in current measuring mode.  Take a Jfet, short gate and source together, and connect to negative terminal of a 9V battery.  Connect one DMM probe to the drain, the other to the positive 9V battery terminal.  You should be able to read the IDSS on the meter.

Matador or anyone else.  I did this method to all 3 of my Jfets.  They all came out at 6.65v.  Does that seem correct?  I had my meter set to read DCV.  I am assuming that is correct.

Thanks guys.  I am close now that I have all my parts.  NOS transformers and capsules!

 

[kante1603]

Put your DMM in current measuring mode.  Take a Jfet, short gate and source together, and connect to negative terminal of a 9V battery.  Connect one DMM probe to the drain, the other to the positive 9V battery terminal.  You should be able to read the IDSS on the meter.

Matador or anyone else.  I did this method to all 3 of my Jfets.  They all came out at 6.65v.  Does that seem correct?  I had my meter set to read DCV.  I am assuming that is correct.

Thanks guys.  I am close now that I have all my parts.  NOS transformers and capsules!

Hi Dan,

No,seems you've measured dc voltage.We want dc current to get the idss.So you have to set the dmm to measure "mAs or As"and plug the probe(s) to the corresponding jacks for current.
Black probe-normally called "Com" goes to drain,red one to the positive terminal of the battery.

Hope to have helped,

Udo.

 

[dandeurloo]

Ok, that is what I was wondering/thinking.  My DMM must not do that.  Any other way to do it or else I can try and get over to a friends place who has much better test equipment.

 

[spase]

we are waiting for that ))

 

[Matador]

Hi Matador,

I've seen at the bottom of the specs page1 that there are 3 types of the 2sk170 existing.

Note: IDSS classification
GR: 2.6~6.5 mA, BL: 6.0~12 mA, V: 10~20 mA

Any preferences?

Cheers,

Udo.

GR's tend to be difficult to find.  The BL's are common and work fine.  Remember that the 2N3819's are spec'd from 2mA to 20mA which covers the same range.

Since this circuit hand-selects source resistances, and runs such a low quiescent current, it really doesn't matter which IDSS value you select in practice.  If the circuit was running higher current, a higher IDSS JFET would give you a bit more signal swing before the device hit's saturation but it's academic in this circuit.

Ok, that is what I was wondering/thinking.  My DMM must not do that.  Any other way to do it or else I can try and get over to a friends place who has much better test equipment.

Place a small resistor between the drain and the positive supply (like 10ohms).  You can then measure the voltage drop across it and calculate the current.

 

[kante1603]

GR's tend to be difficult to find.  The BL's are common and work fine.  Remember that the 2N3819's are spec'd from 2mA to 20mA which covers the same range.

Since this circuit hand-selects source resistances, and runs such a low quiescent current, it really doesn't matter which IDSS value you select in practice.  If the circuit was running higher current, a higher IDSS JFET would give you a bit more signal swing before the device hit's saturation but it's academic in this circuit.

Thank you!
That's what I was thinking.
Just stumbled over GR's on evil bay from a guy who sold them as matched quads.
O.K.,waiting for BL types now.

See you,

Udo.

 

[HellfireStudios]

Just a thought... If Nuemann could increase the gain significantly by replacing a single FET, don't you think they might have already done so? It seems strange they went to all the trouble to increase capsule voltage in the U87a without trying easier/cheaper methods of increasing gain. I'm definitely not against experimentation. There could be a FET that could better the performance of one that's design is decades old, though. If it works out, let us know.

 

[kante1603]

Just a thought... If Nuemann could increase the gain significantly by replacing a single FET, don't you think they might have already done so? It seems strange they went to all the trouble to increase capsule voltage in the U87a without trying easier/cheaper methods of increasing gain. I'm definitely not against experimentation. There could be a FET that could better the performance of one that's design is decades old, though. If it works out, let us know.

I see this different.
It's not a question of why Neumann didn't change the fet.We're talking about a part that is not available anymore,and we look for a part that might fit better because we're definetely off the specs
Neumann wrote in the 70's.Seen from my results I'm about 6dB off from what was to be expected.
That's twice the level.
Now thinking we're transporting an old piece of gear to this digital age lowish levels can be a big issue since one might not be able to drive the a/d converters hot enough without getting to noise trouble.
So every dB helps.
And please remember all designs in the pre digital aera where made for totally different reference levels,e.g. German Broadcast was +6dBm etc.Nowadays were talking about a 0dBFs point that equals something 14 to18 dB prior to hard digital clipping at a refence level of normally +4dBu.
So we are looking at least for a solution that will give us the spec'd level back,right now we're off from that.
As said the mic sounds beautiful,it has just a too low output level.

Udo.

 

[mrcase]

So if anybody of you "B2" guys is interested,I have some nice hints and solutions for you to get this baby fully working like the real thing.Things like how to mount the trafo,grounding improvement and so on.Tell me and I´ll post it.

My setup is all-styrene caps except of c3,Peluso capsule and T13 transformer.

And now:
Dan,my friend:You´ve done an awesome job on this!Thank you for all the support,always being friendly and the unbelievable patience with us-I definetely owe you one!!!
Also a big thank you matador,dave  and all of you who contributed so much here,it made the build doable with excellent results!

Cheers to everybody and a big kick in the ***** of all you guys at the customs who make our life in the diy world so hard!

Udo

I would be very interested in that kind of information. I will be building this mic in a b2 body as soon as I have more time!


P.S.:Will take the mic to our studios these days and do some comparisons with our real U87As.

Editeleted pics for bigger ones in my next post,these have been too small.

 

[kante1603]

Hi mrcase,

is that you Martin?
What´s the question?
Maybe you´ve overseen it,I already have posted my hints & tricks for the B2Po,please see the replies #s 750/753/754.

Cheers,

Udo.

 

[mrcase]

hehe, yes - too impatient again! have found the info! THX

cheers!
martin

 

[poctop]

Just a thought... If Nuemann could increase the gain significantly by replacing a single FET, don't you think they might have already done so? It seems strange they went to all the trouble to increase capsule voltage in the U87a without trying easier/cheaper methods of increasing gain. I'm definitely not against experimentation. There could be a FET that could better the performance of one that's design is decades old, though. If it works out, let us know.

I see this different.
It's not a question of why Neumann didn't change the fet.We're talking about a part that is not available anymore,and we look for a part that might fit better because we're definetely off the specs
Neumann wrote in the 70's.Seen from my results I'm about 6dB off from what was to be expected.
That's twice the level.
Now thinking we're transporting an old piece of gear to this digital age lowish levels can be a big issue since one might not be able to drive the a/d converters hot enough without getting to noise trouble.
So every dB helps.
And please remember all designs in the pre digital aera where made for totally different reference levels,e.g. German Broadcast was +6dBm etc.Nowadays were talking about a 0dBFs point that equals something 14 to18 dB prior to hard digital clipping at a refence level of normally +4dBu.
So we are looking at least for a solution that will give us the spec'd level back,right now we're off from that.
As said the mic sounds beautiful,it has just a too low output level.

Udo.

Hi Udo, here is some information regarding the different FET used in the vintage and Modern Version
http://repforums.prosoundweb.com/index.php/topic,11630.0.html
,  this is where  i decided to go with the 2N3819,  however the sk170 is the one in my orginal and the gain is very comparable i really hope you find the source of your problem

Best
DAn,

 

[Corusco]

You should strongly consider purchasing on of Eric's (tskguy) capsules. They are great. The Rk-87 is a great capsule for the money but you def need to go with the T-13 and all vintage caps with that capsule. I used that combo in my first mic and there are some samples I posted of it earlier in the thread.

I have been asked by a few other people to build mics for them and I'm pushing Eric's capsules to them as well.

Dave

Hi Dave,

Based on your story at GC and my personal preference for Funkenwork iron, I had planned on using the RK-87 capsule with styrene caps.  Is your recommendation for tskguy's capsules based on (aside from their apparent high quality) their capability with the Cinemags and modern components?  Do you plan on trying Eric's capsule in the T-13/styrene circuit to see how the capsules sound comparatively?

Thanks!

 

[Marc Duchesne]

Just finished my built, works great and sounds very vintage like. RK-87 capsule with styrene caps, AMI transformer. Excellent combination. I still have some leftovers styrene caps (470pf and 10pf) that I could ship, it will make you save some dollars plus they are hard to get by...
Just let me know,
-marc

 

[0dbfs]

Ok, that is what I was wondering/thinking.  My DMM must not do that.  Any other way to do it or else I can try and get over to a friends place who has much better test equipment.

Dan,
You could insert a 1k resistor between drain and + on the battery and measure voltage drop across the R then solve for current through the resistor (I=E/R). Check also with a 10k and/or a 100 ohm resistor to confirm that the 1k resistor is not limiting the current.

Cheers,
jb

 

[Matador]

Just a thought... If Nuemann could increase the gain significantly by replacing a single FET, don't you think they might have already done so? It seems strange they went to all the trouble to increase capsule voltage in the U87a without trying easier/cheaper methods of increasing gain. I'm definitely not against experimentation. There could be a FET that could better the performance of one that's design is decades old, though. If it works out, let us know.

It all depends on what you want to optimize for:  for example, gain and headroom are often at the expense of one another.  If you record loud transient sources up close you might not need much gain.  Distance mike'ing of quiet sources has different requirements.

I think the design is a good balance of the two, probably tipped in favor of headroom.  Keep in mind, once the circuit clips you are toast on one loud transient:  compression later won't help you.  However most preamps can give you 6-10dB more gain without problem so it seems like a reasoned circuit design decision.

I toyed with the idea of making the source resistor a constant current sink using a regular NPN BJT and few diodes.  Gain increases from 9dB up to 30dB, and the circuit becomes less susceptible to FET variation.  But input headroom falls by an order of magnitude, which means feedback is likely necessary and the circuit won't sound the same.

But this is the beautiful thing about DIY:  you can tailor the design to your needs!