OPEN SOURCE DIY Mic Project - ORS 87 - Stripped Down u87

[OneRoomStudio]

And the last thought of the day... Couldn't you build a more "u89-inspired" mic by not installing the de-emphasis and using Arienne's flat 47 capsule?
I know it's a different capsule, but her flat 47 is just so nice sounding on its own. And those smaller 87-style bodies on Ali are about u89 size.

I think that would end up just being the "FET847" project, as mentioned before.

As for the SMD version - sure, why not? But I guess the better question is why? The whole point of this was an easy-to-make 87-style mic using parts you might already have laying around.

An SMD U87 circuit has been done by AKG (and lots of other manufacturers) for years. They didn't always do it 100% faithfully....but there are only so many ways to skin a cat. Once you start bringing a pcb and surface mount components into the picture, you're better off just doing the Perception 220 to U87 mod.

 

[Wordsushi]

I think that would end up just being the "FET847" project, as mentioned before.

As for the SMD version - sure, why not? But I guess the better question is why? The whole point of this was an easy-to-make 87-style mic using parts you might already have laying around.

An SMD U87 circuit has been done by AKG (and lots of other manufacturers) for years. They didn't always do it 100% faithfully....but there are only so many ways to skin a cat. Once you start bringing a pcb and surface mount components into the picture, you're better off just doing the Perception 220 to U87 mod.

Perfect. I'm a big u89 fan so it'll be fun adding another branch to the overall project tree in my mind.

As far as SMD goes, I'm definitely more in the why not camp. For me, personally, this project has evolved very quickly past the theoretical, can it be done stage to how many ways can it be done? Given how tiny the Neutrik transformer is, can I make the circuit footprint really small? What would it sound like? What compromises are involved? I don't need a modded Perception, but I do feel the need to find out what's on the other side of this question.

So, it still begs the question of which components would be the most (I suppose the word here would be) agnostic, if SMD?

Right off the bat, I'm thinking of trying to SMD everything except C5,C6, C7 and R7. There's a 2n3819 SMD FET, which inspired this idea. I'd just love to hear some opinions from anyone who has worked with SMD in mics.

 

[peterc]

Hi Mark

>>> There's a 2n3819 SMD FET, which inspired this idea.

What part might that be? I searched for a 2n3819 smd equivalent, but was not succesful.

Thanks

 

[Wordsushi]

Hi Mark

>>> There's a 2n3819 SMD FET, which inspired this idea.

What part might that be? I searched for a 2n3819 smd equivalent, but was not succesful.

Thanks

Thanks for the heads up.
InterFET put out a datasheet with both their TO-92 and SOT23 smd versions of the 2n3819. I hadn't searched for its availability until now and it appears not to be in physical existence.

So, now I'll have to choose between another SOT23 fet or stick with through hole for the FET.

 

[peterc]

Ok thanks for clearing that up, such a pity....

 

[soliloqueen]

Thanks for the heads up.
InterFET put out a datasheet with both their TO-92 and SOT23 smd versions of the 2n3819. I hadn't searched for its availability until now and it appears not to be in physical existence.

So, now I'll have to choose between another SOT23 fet or stick with through hole for the FET.

linear produces a sot23 process 50ish jfet meant to replace the 3819 in medical instrumentation, the ls846:
https://www.digikey.com/en/products...ted-systems-inc/LS846-SOT-23-3L-ROHS/14312162
but being a small-batch process done entirely in the US, it is stupidly expensive. It would work great in this application, though. Its superior breakdown specs mean that there would be little selection that would need to be done from the factory.

 

[Wordsushi]

linear produces a sot23 process 50ish jfet meant to replace the 3819 in medical instrumentation, the ls846:
https://www.digikey.com/en/products...ted-systems-inc/LS846-SOT-23-3L-ROHS/14312162
but being a small-batch process done entirely in the US, it is stupidly expensive. It would work great in this application, though. Its superior breakdown specs mean that there would be little selection that would need to be done from the factory.

Thanks for the lead. If for whatever reason, the ls846 isn't available, is there a backup FET option you would recommend?

 

[soliloqueen]

Thanks for the lead. If for whatever reason, the ls846 isn't available, is there a backup FET option you would recommend?

the LSK189 is the same part with lower noise.

 

[OneRoomStudio]

How are you planning on making the high impedance connection to the gate of a surface mount FET (and 1G resistor)?

 

[Khron]

How are you planning on making the high impedance connection to the gate of a surface mount FET (and 1G resistor)?

There are ways...

 

[Migs 31]

There are ways...

@Khron could you expand on this mysterious sentence ?
Allways thought that 2n3819 had to be selected (for low noise+Idss).
Have a bunch of Fairchild ones from a reliable supplier in smallish quantity.

- BF256b (BF556 ?)
Worked really well and I know of at least one commercial U87 "copy" on the market that uses it too.
Though it is not pin compatible.

M

 

[Khron]

@Khron could you expand on this mysterious sentence ?

I could, but it's nothing that can't be found even here in this very subforum. And i'm pretty sure there's at least one EEVblog video on this very topic.

I'm more on the "teach a person to fish" side of things, and tend to share in @kingkorg 's cynicism regarding sharing "too" much information publicly here. But i can readily tell you it's no secret, it's a near-trivial design detail, that's already used in several commercial mics

 

[Migs 31]

Ah ok, I thought this was a novel/new groundbreaking/Allien given only to you, Technology.
If its in GDIY then no worries, I know how to fish

@Khron : "and tend to share in @kingkorg 's cynicism regarding sharing "too" much"
Yes that is absolutely true. Love you both cynics for that.

Back to U87 FET's.

M

 

[Jarbcd]

No, but measure the DC resistance of the transformer that is in there to determine primary and secondary, and then solder in the UTM trafo accordingly. You can download a spec sheet from their website.

Hey sorry to go back to this, gave it a go, but I ran into some serious noise issues. Do you have any more details/images of how you wired this up??

 

[Khron]

Hey sorry to go back to this, gave it a go, but I ran into some serious noise issues. Do you have any more details/images of how you wired this up??

What sort of noise issues?

 

[micolas]

Hey sorry to go back to this, gave it a go, but I ran into some serious noise issues. Do you have any more details/images of how you wired this up??

UTM 0587
Connect the wires like this:
Red at C7
Black and Yellow to GND
Purple and Green at R18, R19
If the microphone is not in phase, then reverse Green with Violet.
And of course the magnetic/metallic core to GND

 

[Wordsushi]

How are you planning on making the high impedance connection to the gate of a surface mount FET (and 1G resistor)?

I was thinking the easiest way is to keep the 1g through hole like the SMD Alice board does.

 

[OneRoomStudio]

I was thinking the easiest way is to keep the 1g through hole like the SMD Alice board does.

Are you talking about this one? https://www.jlielectronics.com/diy-accessories/opa-alice-single-channel/

That board is not setup for high impedance. It's actually a major flaw in that design.

 

[Wordsushi]

That is a consideration for sure. What would you suggest? I was thinking I either float it to a pin and use a through hole FET with it, or the other side of that is that I have a number of mics that don't have a high impedance capsule connection, including my Bouchard D87s and some Alice mics and it hasn't prevented them from being work mics, so I'm still in the "figuring that out" stage. I've had a PCB designer tell me in the past that it wouldn't make enough of a difference to matter, but he's not a user of mics So yes, any and all input is welcome.

 

[Wordsushi]

Homero was very kind enough to surprise me with some of his ORS87 PCB's. I wish he hadn't put my unsername on it and certainly not above OneRoomStudio, who deserves sole credit. They're nice and small. Gonna have to order some more transformers, but I'm really looking forward to trying it out with Khron's revised BOM.

Someone else has contacted me as well with another DIY ORS 87 pcb in the works, with an optional daughterboard for a pad and possibly also switchable high end EQ. So, exciting times ahead.