A thread about building two Neumann U87 clones

Cirrus

Roadie
Messages
277
This is hardly breaking new ground, it's a pretty well trodden DIY route, but I'm building a couple of mics.

It might be this isn't the best place for a thread about a project like this, it might be that I forget to update it, and it might be that not many people really care. But... I'm making two condenser mics.

So far, I've got almost all the bits including "donor" cheap mics in the form of two ADK 51s, which together cost £130 and will provide the bodies, metal frame, xlr sockets, headbasket and capsule mounts. I do appreciate that it's totally nonsense to build two condenser mics by buying two working (for now)cheaper condenser mics. :LOL:

PXL_20230221_220007035.jpg


A bottle of Acetone and a lot of scrubbing later, the gold decals have been removed and the mic bodies are a mysterious black, like an SR-71 Blackbird.



And everything is now sat in a box, waiting for an evening of inspiration and motivation to get started. I've already began selecting certain components, using the absolutely shamefully jury-rigged testing device I was able to build out of random bits from my bag of cast-off electrical parts from previous projects:

- There are two resistors that have to be matched to, according to the 1980 Neumann schematic, 0.4%. That's beyond the power of my multimeter, so I made a "Wheatstone Bridge", which lets you cycle through a bunch of different resistors, measuring the difference between them and gradually closing in on the closest-matched pair. Out of a batch of 20 resistors, I eventually got two matched pairs, one for each mic, that were 0.05% matched

PXL_20230319_215246097.jpg



- Testing the FETs. I'm not going to pretend I understand FETS. But apparently in the U87 circuit, it's good to test your FETs for a characteristic called IDSS, and you do that by putting voltage across them and seeing what current they draw when you short two of the legs together. Lower is better, and a close match will be good to try to get a matched pair of mics, though it's not the most critical thing there - that's on the capsule.

PXL_20230319_204847042.jpg
 
There has already been a tragedy - most of the components (Circuit boards, Transformers, caps and resistors etc) arrived a fortnight ago. The capsules were coming from America, and got stuck in customs here in the UK for almost two weeks. Uh-oh, I thought... what are they doing to my uber-delicate babies?

It turns out they were getting their fingers all over them, pulling and twisting at wires, and have damaged the thin membrane of one side of each. See the wrinkles near the middle of the capsule here? They shouldn't be there.

PXL_20230314_183135733 (1).jpg


Luckily, the maker is sending me, free of charge, two replacements with a faster shipping service. I've really got my fingers crossed that there won't be any problems with these when they get here..
 
Awesome project. I've never thought about building mics before, I'll be watching this really closely. Are you following any kind of build guide or how are planning/organising this?
 
PCBs cleaned in pure alcohol, which inspired me to break out the Whiskey Fudge (obviously needed to finish it off to use the tin as my project box).

And... half the resistors soldered in. I go slow, checking with a multimeter before I solder, then again after in case I've damaged anything with heat. And I'm making two of these, which each have two boards, so after 90 minutes I feel like I've hardly done anything. Ah well. Maybe a glass of whisky* will make me feel better.

PXL_20230320_223328814.jpg



*Whisky. Scottish. Proper.
 
Do you test from the pads on the bottom or on the top? Or do you test the legs of the resistors from the top?
Where possible I test from a different pad that's connected to the pad I just soldered, so I'm testing the join as well. And that works the other way too - by placing the probes on the legs of two different components from the top of the pcb, I can test for continuity between those points which will include at least two solder joins. It's probably overkill but though I've done endless guitar wiring, made about a dozen pedals and an amp and modded several more over the years, I still feel like I've definitely screwed it up every time i pick up the iron. And I guess it saves troubleshooting time later if I know that I gave every single join my undivided attention and tested it afterwords.
 
I tend to fit one part at a time and measure all my resistors before fitting to minimise mistakes but I don't test the joints after I've made them apart from a visual inspection.
 
I've nearly got all the resistors on, am just wondering how to double check the 1g and 68m ones. My multimeter reads up to 60 megaohms, i guess I can measure two 68s in parallel and look for 34, but the 1 giga-ohm ones? all 6 in parallel will still be ~160m. I guess I could use the wheatstone bridge again and just assume that if there's no outliers, they're probably all fine.
 
I've nearly got all the resistors on, am just wondering how to double check the 1g and 68m ones. My multimeter reads up to 60 megaohms, i guess I can measure two 68s in parallel and look for 34, but the 1 giga-ohm ones? all 6 in parallel will still be ~160m. I guess I could use the wheatstone bridge again and just assume that if there's no outliers, they're probably all fine.

I'd say you're probably safe to risk it. I don't check my resistors to make sure they're fine but to double check the values so I don't have to read the codes. I've never had a dodgy resistor before.
 
Anyone ever learn this mnemonic?


accurate ‘nuf for learning ‘yor values.

Dad was a EE

“Bad Boys Raped Our Young Girls But Violet Gives Willingly.”
 
Last edited:
Damn, this is awesome!!! A U87 for a couple hundred bucks and some time doing something cool? Hell yeah!
I wish I coulda kept the cost that low! It's about £850 in parts for the two, that's including the donor mics, transformers, capsules, PCBs, electrical components, and a bit of customs duty sting. Could have done the pair for about 500 if I'd went for cheaper capsules, maybe only with one polar pattern, and cheaper transformers. But my thinking is use the best stuff I can get, then I don't feel like they're a "poor man's U87" or something. I felt the same when I built my tweed clone. Kind of, "what's the point unless I reckon it could stand shoulder to shoulder with the official thing?"

Saying all that, I'll probably be the first chump to build mics that actually catch fire at high SPLs or something :LOL:
 
I wish I coulda kept the cost that low! It's about £850 in parts for the two, that's including the donor mics, transformers, capsules, PCBs, electrical components, and a bit of customs duty sting. Could have done the pair for about 500 if I'd went for cheaper capsules, maybe only with one polar pattern, and cheaper transformers. But my thinking is use the best stuff I can get, then I don't feel like they're a "poor man's U87" or something. I felt the same when I built my tweed clone. Kind of, "what's the point unless I reckon it could stand shoulder to shoulder with the official thing?"

Saying all that, I'll probably be the first chump to build mics that actually catch fire at high SPLs or something :LOL:

Ahhhhh ok, yeah that changes things a bit. :rofl

Still badass you’re doing this!!!
 
Listings like this make it feel all worthwhile.

£11k for a pair of 1975 U87s!

Interestingly, the old ones tend to go for more in some circles because they're perceived to be better than the presently available model, the U87ai. The difference, other than the intangibles like "do old capsules have magic new ones don't" or "were the transformers somehow better?" is that the original version is relatively low output - the capsule is only polarised to around 46volts volts, as that's all 48v phantom power can provide. And the FET amplifier built into the body is quite low output, so it's apparently roughly on a par with a typical dynamic mic - maybe slightly hotter than an SM7b, and people are forever complaining about low output on those.

The modern "ai" version has a voltage booster to raise the capsule polarisation up to 60 volts, and that means more output, as much as 10db, so that's more than double, but it also changes the response of the capsule because the extra charge physically stiffens it up some - there's more attractive force between the diaphragm and the back plate.

These are copies of the older, lower output version, because that's what the PCB designer put together and it's less components for an older, ostensibly more desirable model. The differences with mine will be - is the capsule actually the same as what Neumann used, is the transformer voiced the same, and the mesh headbaskets of my mics are a little smaller than the u87 headbasket. Though, they do have the same basic geometric shape, with a flat section in front and back of the capsule. Headbasket dimensions do affect the sound, as they create resonances in the internal space. So assuming they work, I'll be listening to make sure I'm happy with their overall voicing and worst case is I'll need to buy different mic bodies to rehouse them.


Anyway. PCBs almost done, just need to select and solder in the FETs.

original_ec90f014-0afa-4a37-a630-91c4d212d028_PXL_20230322_214258497.jpg
 
And both Capsules are in place. I handled those babies like they were live nuclear detonators, rehearsing my moves before getting the screws in to make sure I didn't unthinkingly touch a diaphragm like an ignorant customs officer. Everything's been done wearing gloves to keep greasy body extrusions away.


NjrIVLivUzPxQxDdPgQk126QDDzvh3It3wm99WC5JJmmNSI5R2UQsIkGQpH-8Wos48p8FvzGJgNBA4DWXQ4c9rXDzk-gMJ5_DOc9kAzohQ7Y-nsJbDt8bx0xtTaGBKACqPPjOjhdb0w-ZHqZSnm4InPQgmptscQvBfev8x-NOjIw3h2ARZtCOUo-SIgH-A4yNNsLoK8emhLszNPtotglnlhqf9hGIFbxhPpSsgSTe4Lxcq6-SL6ceeoTiKBeo1hZlL9bqgyBmRc9DP3KxM6QOzTo8BIPkyZqn264d3uCyo8gRe0baTCRexrC4lJ2VvERqG2L5r5q4BE6p_fUa-MS-UdRiXBzYwGwAbnHIPj5Yy-aZHUDyEWSqp8q4sTkjcktmQCDIhjpltqIsBwraSJfPeaDHe14GA6RCrABRUBpJSHQkmDUEUeOk0k5E9IWq8UcuKwPQvOFNNmDfI76PN82D-rYBkKb0oduoLR3OrwAsW0hs5Et-cC7gEThqU-0mBOFiBZFaZgvuyvbn17lXPulxkurPimYIUULmvnJS7nvGORJFytKyyTNDWRosYJrwZddvDGuK1Hvl4Tw4e3M5jZIKMbJ4dCfAqy1LO2rzeHNNNuzgqafE0sF3ZevaIRyz6mrp6jLQgqS9pnBQ95ZlOr5O91LYIIxsI_AhVa-mDWDjUOsKntYLrkSZ1wcic8_FDgVkRPT3jqjWGNsaIXbRcNRsQdGXojOkCAoOTun1AL7hltkOHaqAu_94uV7lSkNJWx9ZEA1_nT5o5QeI7a3x8NewqsIJW1nuFAxXWV8vKKNPdk0HCw_JuGA5ZDdL0eONx-7JWiHe184d33ZSfSkbXfgikgC4PCOcV4l2_3gi0nNAYEEEl3-Q4yxp1WGJtb1dX6mQxg5kI9e9t6oVMFvBkg9qd7fXGsFullM3GBMQUNS_0V-Cgr6IQWOYgpKLh0SKjztfRoBX3l4nhxOHqewcKQ=w653-h1161-s-no
 
Back
Top